Inadequate housing in Ghana

Two themes are evident in housing research in Ghana. One involves the study of how to increase the number of dwellings to correct the overall housing deficit, and the other focuses on how to improve housing for slum dwellers. Between these two extremes, there is relatively little research on why the existing buildings are poorly maintained. This paper is based on a review of existing studies on inadequate housing. It synthesises the evidence on the possible reasons for this neglect, makes a case for


Introduction
Housing "inadequacy" in Ghana is usually interpreted in two ways. One interpretation emphasises inadequate quantity: the gap between housing supply and housing demand. For instance, when Deputy Works and Housing Minister Hannah Bissiw was asked how the government would solve the inadequate housing the country faces, she said the "government will construct over 200,000 housing units over the next 5 years . . . will put up some 300 units for Members of Parliament (MPs), ministers and other officials, while looking at providing another 1,000 units for the police, fire service, immigration officers and health workers in the country" (Koomson, 2009). Generally, the housing policies pursued by successive governments in Ghana have all tried to close the gap between the needed and existing stock (Arku, 2009). The other interpretation of inadequate housing emphasises slum housing. A recent United Nations Human Settlements Programme (UN-HABITAT) report on improving housing conditions in Ghana suggested that inadequate housing implies slum housing when it noted that "[e]ven though housing is a basic necessity of life, more than half of the population in Ghana live in poor houses where they have no access to adequate sanitary facilities, water and warmth to meet their daily physical needs" (UN-HABITAT, 2010: 4). As is the case in many other countries around the world (Gilbert, 2007), this interpretation usually serves as the official reason to evict or resettle people that dwell in these inadequate buildings. As with past governments, the current government of Ghana has already carried out a number of evictions (Ghana Institute of Planners, 2010).
One aspect of inadequate housing that receives little attention in official reports and government housing policies is inadequately "maintained" housing. Successive governments in Ghana have been much more interested in delivering new housing units rather than maintaining existing ones that have fallen into disrepair, prompting Prince Cobbinah (2010) to advocate a national housing maintenance policy. Building maintenance has many benefits. It ensures that expensive repairs are avoided in the future and saves countries from eventually losing their housing stock through decay (Harding et al., 2007). Generally, well-maintained buildings have higher market value (sometimes 13% more) than poorly maintained ones (Wilhemsson, 2008). A higher market value may mean little to groups that place sentimental value, rather than market value, on their properties. However, such groups also benefit from maintenance because it can help protect their buildings from collapse. In addition, the government stands to gain more revenue from property rates or taxation, especially on improved capital values (Appiah et al., 2000;Inanga & Osei-Wusu, 2004).
There are other ways in which maintenance impacts individuals and groups and their housing situation. A recent book by Sarah Glynn (2009) contains a wealth of international evidence of the health-related and educational benefits of better-maintained buildings (see also Zavadskas et al., 1998). When people have better-maintained buildings, their health and the educational performance of their children improve through a reduction in housing-related health problems such as respiratory infections, asthma and injuries (Krieger & Higgins, 2002). There is a direct relationship between poor-quality housing and poor mental health, developmental delay, heart disease and short stature (Bashir, 2002). Given these potential benefits of better quality housing for Ghana, it is useful to analyse why buildings are poorly maintained there and to discuss how the problem could be remedied. This paper makes a modest contribution. It reviews and synthesises the scattered pieces of evidence from more detailed studies about inadequately maintained housing. The three sections of this paper describe the concept of inadequate housing, identify its main causes and effects in Ghana, and offer suggestions for remedying the problem.

Inadequate housing and types of dwellings in Ghana
At its sixth session in 1991, the Committee on Economic, Social and Cultural Rights (CESCR) adopted a detailed general comment on Article 11(1) of the covenant dealing with the right to adequate housing. According to paragraph 8, "the concept of adequacy is particularly significant in relation to the right to housing since it serves to underline a number of factors which must be taken into account in determining whether particular forms of shelter can be considered to constitute 'adequate housing'" (Office of the High Commissioner for Human Rights, 1991). The CESCR identified the following as essential components of adequate housing: legal security of tenure; availability of services, materials, facilities and infrastructure; affordability; habitability; accessibility; location; and cultural adequacy.
Settlements that lack these features are usually labelled by UN-HABITAT as "slums" (UN-HABITAT, 2003). By definition, slum dwellings are also inadequate. The interest in this paper, however, is in buildings that may be deemed inadequate on the habitability criterion. Subparagraph 8c of Article 11(1) of the covenant dealing with the right to adequate housing states that "[a]dequate housing must be habitable, in terms of providing the inhabitants with adequate space and protecting them from cold, damp, heat, rain, wind or other threats to health, structural hazards and disease vectors. The physical safety of occupants must be guaranteed as well" (Office of the High Commissioner for Human Rights, 1991). This framework excludes several dimensions of inadequate housing (Tipple & Speak, 2005.

Inadequate housing in Ghana
The focus of this paper is on "poorly maintained dwellings", a description that does not fit neatly into any of the categories in the global framework. Inadequate housing, in this context, is not slums, and the people in these buildings are not, as Western housing terminology would suggest, homeless (Tipple & Speak, 2006). These are "in-between" buildings: they do not have the characteristics of buildings in gated communities (Grant, 2009) or the attributes of slums (UN-HABITAT, 2003). In-between buildings are those that are poorly or never maintained, but whose structural and aesthetic features could be enhanced by various forms of maintenance.
The particular country of interest is Ghana. Residential buildings in this country fall into one of five types: "compound houses", which consist of several households around a kraal for livestock (46%), separate or detached homes (25%), semi-detached homes (15%), flats (4%), huts and structures in compound houses (4%), and tents, kiosks, containers or improvised housing (2%; Bank of Ghana, 2007). Many of these buildings are not maintained. Indeed, any structure with a roof is officially regarded as a building (Tipple & Speak, 2009). The Bank of Ghana (2007) claims that 56% of the dwelling units in Ghana are not fit to be called homes. "Mother Ghana", according to one observer, "has indeed moved from where she was, after her independence, into a very promising nation . . . However, one of the major problems, confronting the development of this country, is the poor attitude of its stakeholders towards rehabilitation, and maintenance of its structures and facilities" (Mensah, 2008).
More systematic studies confirm the pervasiveness of the problem. Franklin Obeng-Odoom's (2008) study of the Habitat for Humanity Housing Scheme revealed that over 50% of the dwellings constructed are never maintained. In addition, Bernard Baiden et al. (2005) found that 53% of a sample of timber-constructed homes predominantly over 40 years old had never been repaired or maintained, although the buildings were rundown. The housing supply in Ghana is therefore not only inadequate in quantity but also in quality. It is important to go beyond a description of the problem and identify causes, consequences and some strategies for possible remedy.

Causes of inadequate housing in Ghana
Evidence from earlier studies shows that seven factors are particularly responsible for inadequate housing in Ghana: a poor maintenance attitude, lack of estate management expertise, insecure tenure, lack of rental payments, rent control, expensive building materials and poverty.

Poor maintenance attitude
Poor building maintenance is widely regarded as the result of people's apathetic attitude, a poor "maintenance culture" (Ghana News Agency, 2001). Here, culture is to be interpreted as a set of attitudes. According to the United Nations Educational, Scientific and Cultural Organization (UNESCO;1997), the problem of maintenance plagues Africa as a wholethe "African Poor Maintenance Culture". Perhaps this is why, in a BBC discussion titled "Is Africa's architecture dying?" many of those that commented on the topic felt that the problem with Africa was not its architecture but its poor maintenance culture (BBC 2006). In this view, dwellings in Ghana are in poor condition because the occupants do not have a responsible attitude toward maintaining their buildings.
This apathetic attitude toward proper maintenance could have its roots in some traditional cultural beliefs. Deborah Pellow (1988Pellow ( , 2001 has found that maintenance is of secondary concern to some ethnic groups in Ghana that believe that homes should have sentimental value rather than market value. As such, the condition or state of repair of buildings does not matter. What matters is the set of social relations that a dwelling embodies. The family house (Twi abusua fie), for example, is not for sale, and hence maintenance -which enhances market value, but not necessarily sentimental value -is not usually a major concern (Tipple, 1987;Willis & Tipple, 1991;Geest, 1998). Explained this way, it would seem that the "culture of poor maintenance" is endemic (for a description of this tendency in Osu, Accra, see Grant, 2009).
However, not all buildings in Ghana are equally poorly maintained. Cobbinah (2010) found that the homes of Ghana Health Service staff are better maintained than those of the Ghana Police Service. Given that occupants of these homes are also Ghanaians, there may be other reasons beyond attitude and culture why many buildings are in a state of disrepair. This leads to a viewpoint held by housing professionals, particularly those of an estate-management orientation, that lack of professional estate-management units is a major cause of inadequate housing (Tudzi, 2006;Lai, 2011).

Lack of estate-management expertise
Is the limited number of estate-management professionals in Ghana the source of the problem? Some may look at this diagnosis as a means by which some housing professionals want to create jobs for themselves. However, the need for estate management deserves some careful consideration (Arnison, 1988;Li, 1997) because, to use Brian Wood's (2005) journalistic language, maintenance is more than "Cinderella going to the ball and meeting Prince Charming -by accident".
Professional estate management, on the other hand, may be regarded as the set of activities needed to keep a property in the state in which it was when it was new and, possibly, to improve upon it. Estate managers generally provide services such as maintenance, security and cleaning, marketing and financial management. However, it is difficult to manage buildings without records. The dictum "one cannot manage what one does not know" is applicable to estate management. For this reason, records management is an integral part of estate management (Ferry & Brandon, 1991;Ratcliffe & Stubbs, 1996;Wood, 2005;Martin, 2006). The lack of sufficient estate management professionals in Ghana could therefore mean that buildings lack "professional care".

Insecure tenure
Insecure tenure is a major problem in Ghana, where many land transactions are not covered by proper documentation (Abdulai, 2006(Abdulai, , 2010. In this context, tenure, or the rights of an individual or group over land, is secure if it ensures protection against illegal removal, eviction or harassment by government or individuals, corporations or other legal persons (UN-HAB- ITAT, 2003). It is generally believed that when people do not feel secure they under-invest in their homes: people unsure of how long they can keep their homes generally will not invest in maintaining them (UN-HABITAT, 2003). Given that insecure tenure is a major problem in Ghana, it would be fair to name it as one of the reasons for poor building maintenance, although insecure tenure fails to explain why some landlords, even those whose properties have secure tenure, do not carry out proper maintenance of their buildings (Arku & Luginaah, 2010). Other factors must therefore also be considered.

Lack of rental payments
The willingness and ability of landlords to maintain housing also depend on the revenue streams from renting. An analytical framework for understanding this is the Filtering Model. This model stipulates that buildings depreciate with age. The rate of housing depreciation is, however, contingent on the quantity and quality of maintenance. The amount and effort homeowners or landlords spend on maintenance depends on: (1) how much they expect to receive in rents and (2) how much the homeowners want to realise from selling their homes. In addition, tenants will usually pay more for a dwelling that is better managed because a building of that nature offers higher-quality services (O'Sullivan, 2003). Therefore, in theory, when the forces of demand and supply raise the amount of rents for higher-quality housing, landlords spend more on maintenance to improve the quality of their properties in order to make them "filter up" on the rental market (Somerville & Mayer, 2003).
It could be argued that free residential housing provided to the staff of some public institutions in Ghana, such as the Ghana Police Service, may be one of the reasons for their poor maintenance conditions. Police quarters can be used as an example. Some police officers, such as those below the rank of assistant superintendent, usually occupy their bungalows without paying rent. For this reason, the landlord (the Ghana Police Service) has no incentive to maintain the buildings. The tenants, on the other hand, are rarely ejected or surcharged for poorly maintaining their homes. There is therefore no disincentive not to allow their buildings to fall into disrepair (Cobbinah, 2010). Looked at another way, because some police officers do not pay rent, they do not usually feel entitled to demand better housing conditions. A variant of this view is that the rent paid by the Ghana Police Service to private landlords is so low that landlords find it difficult to maintain the rented homes (Public Accounts Committee, 2009). Closely related to this view is the issue of rent control.

Rent control
When rent is controlled, landlords tend to become disinterested in maintaining their properties because their profit margin is reduced. These tendencies create inadequate housing, as noted in a different national context by Andrew Baum (2004 (1990) and Willis and Graham Tipple (1991) found that, while years of rent control had succeeded in making housing inexpensive, rent in indigenous buildings could be as low as 43% of the market rents or 52% of the market rent in the case of tenement housing. They also found that rent control discouraged landlords from spending on the maintenance of the rental units in order to increase the capital return on their investment. They argued that rent control constrained the supply of housing, even imposing welfare costs on society as a whole, and generating a disincentive to landlords that was greater than the benefit to tenants. They therefore recommended that the Rent Act of Ghana be abolished.
Although the Rent Act of Ghana has not been abolished, it is not systematically enforced. This makes it possible for landlords to charge whatever amount of rent they desire, even demanding two to three years of rent in advance (Obeng-Odoom, 2009). Nonetheless, there has not been any dramatic improvement in the maintenance of their homes (Arku & Luginaah, 2010). To overcome this tendency, rent control may be implemented simultaneously with enforcing strict building regulations. That is, while the state controls rent, it may also compel landlords to invest in maintenance. It is evidently the case in Ghana where the Rent [Control] Act of Ghana of 1963 (Act 220) is supposed to be enforced alongside regulation 12(1) of the National Building Regulations (Legislative Instrument, no. 1630/1996. According to regulation 12(1): "A District Planning Authority may, in respect of any building which has in its opinion fallen into a state of disrepair or neglect, and constitutes a safety or health hazard to the public, or for aesthetic purposes serve notice in writing upon the owner of such building requiring him to carry out such reasonable repairs or painting as may be specified in the notice and within such time as may be stated in the notice." Sub-regulation 2 states: "Where a person notified under sub regulation (1) fails or refuses to carry out the repairs or painting within the stipulated period, the District Planning Authority may carry out the repairs or painting and may take legal action to recover the cost involved in the work." However, local governments in Ghana typically have poor human resources, weak structures and inadequate revenue to enforce building regulations (Mensah, 2005;Yeboah & Obeng-Odoom, 2010). In turn, private landlords usually pay little attention to housing maintenance.

Expensive building materials
It is widely believed that building materials in Ghana are expensive relative to the incomes of the majority of people. The home price/income ratio in Accra (14.0) is high compared to the average in African cities (12.5; UN-HABITAT, 1998). Out of 26 cities for which the UN-HABITAT had data in 2003, only five (Abidjan, Tanta, Monrovia, Maputo and Jinja) had a higher home price/annual income ratio than Accra, and only six cities (Abidjan, Tanta, Monrovia, Maputo, Jinja and Antananarivo) had a higher home price/income ratio than Kumasi (11.6), Ghana's second largest city (UN-HABITAT, 2003). Because building materials are expensive, landlords and homeowners are discouraged from maintaining their buildings.
The price of building materials has been increasing, often rising faster than incomes. One study by Kwadwo Konadu-Agyemang (2001)  It could therefore be argued that the high cost of building materials forces people to build with cheap and inferior materials, which may hasten the deterioration process. A study by Willis & Tipple (1991) found that cheap but poor-quality building materials, such as sun-dried local clay, quickens the pace of deterioration. However, according to Konadu-Agyemang (2001), the majority of dwellings in Accra, for example, are built of good-quality materials. In addition, Tipple (1999) has suggested that the problem with maintenance goes beyond the use of poor housing materials. According to Tipple (1999: 29): "A lack of maintenance tends to create the impression of poor quality construction. Leaking water pipes and drains, badly maintained roofs, unrepaired and unpainted woodwork, and unkempt and damaged grounds and access ways, all lead to accelerated decline and the poor appearance symptomatic of housing at the end of its economic life or housing in need of more than ordinary levels of maintenance." Evidently, other reasons exist for the poor maintenance of buildings in Ghana. At the root of the situation is the problem of low incomes.

Poverty
One would expect that, in a country where 28.5% of the population lives below the poverty line (Ghana Statistical Service, 2007) and a large section of the urban population works in the informal sector of the economy and earns low wages (International Confederation of Free Trade Unions, 2004), many people would not be able to set aside a reserve fund for the maintenance of their homes, as shown in Figure 2. Godwin Arku (2009) notes this point, but interprets it only in terms of inability to purchase "high class" housing in Ghana. However, inadequate housing is also a reflection of the level of poverty in a society (Barton, 1977). It is partly because of the link between poverty and housing quality that Cape Coast, one of the poorest cities in Ghana, also has such poor housing conditions, euphemistically called "heritage or historical housing" (Agyei- Mensah, 2006).
David Korboe (1992) notes that family housing in Ghana is one of the most poorly maintained types because its occupants are poor. The stigma of "the poor family house occupant" and the pressure to bear a disproportionate share of living cost in the family house forces those members whose economic position improves over time (and could contribute to maintaining the building) to leave such dwellings for private accommodation. Of course, other factors, such as overcrowding, also quicken the rate of deterioration in family houses (Willis & Tipple, 1991;Garrod et al., 1995;Sinai, 2002). One practice that reduces the rate of deterioration of family houses in Ghana is the funeral ceremonies that are organised when members of the family pass away. During this time, members of the family that are better off and might have left for private accommodation contribute money to be used to "beautify" the family house because it is the place for the dead to be viewed by the public (Geest, 2000). This temporary beautification is necessary for a good name for the family. However, the end of the burial ceremonies also marks the end of maintenance, with the cycle beginning once death strikes again.
A major driver of this approach to maintenance is low income and poverty (Gyamfi-Yeboah & Boamah, 2003a, 2003bAsiedu, 2004;Asare & Whitehead, 2006). Poverty is therefore a major cause of inadequate housing in Ghana. The pervasiveness of inadequate housing in Ghana is recipe for several social economic problems, including ill health. Ill health itself may cause inadequate housing when it leads to poverty. In turn, the country may suffer in economic terms, as in labour time lost and loss of revenue through the treatment of ailing citizens as well as loss in terms of the cost of replacing the housing stock incurred by the state (British Medical Association, 2003). How can the situation be improved? As with the causes of inadequate housing, there are several suggestions for reform.

Policy implications and strategies to remedy the situation
Although the focus of this paper has been on Ghana, the problem of, and inattention to, inadequate housing is a problem in other African countries. South Africa is more globalised and has better infrastructure than many African countries (Otiso et al., 2011). However, even in South Africa, housing delivery was the main strategy for successive governments until August 2006, when, realising that the country was failing to maintain existing stock, the cabinet approved a National Infrastructure Maintenance Strategy (Wall, 2008). Inadequate housing is a problem in Morocco, Zimbabwe and Egypt (UN-HABITAT, 2008). A comparative analysis of inadequate housing in Africa is difficult to come by because it can be argued that there would be, as with homelessness (Tipple & Speak, 2006), a lack of consensus on what constitutes "poor maintenance". It is suggested here that there is no uniform "language" in housing and urban studies at the moment to describe this problem. Would this imply a need for such language? Grand theories come with their own problems too, and indigenous meanings and definitions continue to garner much interest in housing studies (Darcy, 2010).
However, even within the "indigenous Ghanaian understanding" of inadequate housing, there are no easy answers to how to resolve this problem. The seminal works by Malpezzi et al. (1990) and Willis and Tipple (1991) suggest that removing rent control would give landlords an incentive to provide maintenance. However, as has been shown, there are many dimensions to this problem, ranging from culture to lack of professionals. In any case, what is the guarantee that landlords would invest the extra money in maintenance and not in building more housing units for rental purposes? Similar questions were raised by Joseph Kironde (1992) when Malpezzi et al. (1990) and Willis and Tipple (1991) first published their work. Such concerns are legitimate, given the overwhelming evidence of "misappropriation" of, for example, aid and microcredit, both internationally and in the Ghanaian context (see, e.g., Kotir & Obeng-Odoom, 2009). Moreover, what is the guarantee that, even if landlords did invest in maintenance, tenants would pay for higher-quality housing? Again, Kironde (1992) raised this question when Malpezzi  (2011) shows that, faced with the choice of choosing between formally trained estate agents and informal agents, the latter are preferred because, among other reasons, they charge low fees. Thus, while extending market relations into rental housing could improve maintenance, it must be part of a larger programme and not considered on its own.
What about improving security of tenure in order to encourage people to invest in maintenance? The relationship between security of tenure and maintenance is not direct. Geoffrey Payne (2001) notes that there are cases in which the offer of titles to squatters has even been rejected because of high administrative charges. Other cases exist in which the offer of legal titles has not generated any improvement in housing. Ann Varley's (1987) pioneering work on the relationship between security of tenure and housing improvement shows that extending legal title to informal housing would not necessarily lead to improvement in housing. There are many other factors such as the provision of services that could also provide security of tenure. Work by Raymond Abdulai & Felix Hammond (2010) shows that in Ghana secure tenure is not a prerequisite for a loan; rather, it is a post-requisite.
Insecure tenure could even encourage better maintenance. Some people could develop and maintain their homes if they hoped that such behaviour could prevent the authorities from removing them from illegally occupied land. That would be the case if the principle of "estoppel by acquiescence" (that one should not be made to suffer because of another's misleading act) is invoked. Hence, if people without title to land are led, by the silence of the authorities, to believe that they could continue occupation, people's subsequent development of the land could give them some legal claim to it, especially if it is shown that the authorities would benefit by dispossessing those people (Woodman, 1996). A more direct objection to providing secure tenure, as a panacea for proper maintenance, is to ask whether all the people with secure tenure regularly maintain their homes. Evidence provided by Arku and Isaac Luginaah (2010) shows that in Ghana they do not.
The development of a secondary mortgage market could help the situation. However, although providing credit may temporarily provide relief to the poor to maintain their homes, there is now formidable evidence -not least of all from the disappointment with microcredit in India -that credit is not an effective way to ameliorate poverty in the long run. In a study by Julius Kotir & Obeng-Odoom (2009) in the northern region of Ghana, it was discovered that the majority of recipients of credit use the money for needs other than what the money was originally meant for. There is also the problem of default and subsequent indebtedness. A study of an affordable housing scheme run by Habitat for Humanity Ghana showed that 7 out of 10 people default on mortgage repayment (Obeng-Odoom, 2008). To end poverty-induced maintenance problems, a more sustainable approach could be for the government to provide jobs and to institute progressive forms of taxation that would eradicate social inequality, which remains pervasive in cities (Obeng-Odoom, 2009). In general, relying on the market to solve the problem of inadequate housing would cause more problems than it would solve.
State intervention could be considered a way to resolve some of these market failures. For example, assistance could be given to the institutions that train estate managers to produce more trained "property care" people. Care should, however, be taken not to reproduce the current stereotype: valuation-centric estate surveyors (Obeng-Odoom, 2011). In addition, there could be an attempt to subsidise building materials, although a more long-lasting measure would be to consider a revision of building codes, which would place more emphasis on locally produced building materials and values. The problem of "alien" building codes goes beyond Ghana, of course. It affects sub-Saharan Africa as a whole, if not the entire African continent. Ambe Njoh (2009) has shown that town planning policies and human settlement codes in the sub-region have been heavily influenced by indigenous elitism (during the pre-colonial era before 1884), European racism (the colonial era, 1884-1960s), modernism (1960-1980s) and globalism (1990s-present). Njoh (2009) calls on planners to "Africanise" such policies and purge them of their "alien" character.
This view could be extended, for example, by considering whether the provision and enforcement of building codes should be the sole preserve of the state. To obtain general acceptance of and respect for the codes, there is a need for a detailed analysis of the cherished values of the Ghanaian people. These values should be understood and enshrined in any building codes. People would inherently want to use their homes as they please. As such, it would be useful to pursue a two-stage strategy that (1) emphasises people's participation in reversing imperial housing codes handed down by the colonialists and (2) insists on drafting and enforcing new codes that reflect native values (Symons, 1980). The award of the 2009 Nobel Prize in Economics to Elinor Ostrom for her work in economic governance -which highlights how ordinary people can successfully manage common property (Royal Swedish Academy of Sciences, 2009) -would suggest that ordinary people could successfully manage their homes, especially if they take part in designing the standards or codes against which success is be measured. By empowering agencies like the Rent Control Department, to which all tenants could lodge complaints, the state could make some advances in ensuring improved maintenance. Whether these could be applied in Ghana needs more detailed consideration in the specific context, of course, but these conceptual and empirical contributions from international research suggest that it is at least an open question.
In addition, the state could directly invest in housing maintenance. The Kwame Nkrumah government (1957)(1958)(1959)(1960)(1961)(1962)(1963)(1964)(1965)(1966) tried the Roof Loans Scheme and many other state-led approaches to maintenance. However, because that government was toppled in the 1966 coup before the policies could be fully developed, it is hard to make a careful evaluation of those state-led policies (Aryee et al., 2005). Of course, a common problem of public housing is generally how to fund it. Taxation could be considered. A new tax could be instituted on land or an existing land tax could be increased. Alternatively, the incidence of taxation could be broadened. These suggestions are not without challenges. For example: would the state have the capacity to collect taxation? Would the people have the capacity to pay? Local taxation in Ghana does not have a very good record. Revenue from property taxes, for example, is low. In Kumasi, the city authorities collect only about a third of the full tax. In Kasena, it is less than that (Obeng-Odoom, 2010). Logistical constraints, poverty, corruption and embezzlement are all possible reasons why tax revenue is low (Mensah, 2005). Direct poverty reduction needs to underpin any housing policy. At present, housing is not integrated into the Ghana Poverty Reduction Strategy. It seems that government is approaching the housing question independent of the poverty question. However, as has been seen, the two are closely related.
There are other problems with suggesting that the state become involved in housing maintenance. Edward Symons (1980) argued some three decades ago that investing in maintenance could be politically difficult for governments because, unlike new construction that creates conspicuous signs of government achievement, maintenance is less visible. Given that governments would want to win elections, they are more likely to consider building new housing units as more politically efficacious. In addition, unions may feel that maintenance would not create as many jobs as building new units. These concerns are genuine, but they could be addressed if maintenance were seen as part of a holistic housing programme that emphasises both quantity and quality, in terms of slums and poorly maintained housing.
Overall, a broad state-led housing maintenance program is feasible, but would not come naturally to the Ghanaian people, who must embrace the discourse of housing maintenance and steer it from its current state of relative obscurity to its rightful place in the housing issue. As with the issue of housing quantity, the maintenance question should be made a political one, so that "grassroots" agitation on the issue could shape the "language" of opposition political parties. It is not clear how this grassroots interest could be aroused, directed and sustained. It is also not clear how the state structures would respond to this new demand for better housing maintenance. There is a need for greater research on these issues, but the recent display of popular power in Egypt and Tunisia offers ample hope.

Conclusion
It has become commonplace for policymakers in Ghana to posit that the solution to inadequate housing is to increase the overall housing stock. However, the preceding analysis has shown that this approach hides the pervasive problem of inadequately maintained housing and its attendant socioeconomic and health consequences. Overcoming this problem is difficult: partly because it is multidimensional, partly because it is endemic, partly because the posited solutions conflate symptoms with causes and partly because some policy makers benefit from the status quo. The state has the potential to remedy the current situation but its capacity to do so depends, among other things, on the extent to which it can be held accountable by the majority of its citizens.