Policy review

Urban Studies1–20� Urban Studies Journal Limited 2015Reprints and permissions:sagepub.co.uk/journalsPermissions.navDOI: 10.1177/0042098015571055usj.sagepub.com

Assessing residential satisfactionamong low income households inmulti-habited dwellings in selectedlow income communities in Accra

Irene Appeaning AddoUniversity of Ghana, Ghana

AbstractMulti-habitation is the predominant housing strategy adopted by low income households toaddress their housing needs in urban areas in Ghana. The recent housing policy draft in Ghanarecommends multi-habitation as an urban low income housing strategy. However, a couple ofresearches indicate that households living in multi-habited houses are faced with a myriad of chal-lenges. One such challenge is conflict over inadequate shared facilities. Using both quantitativeand qualitative methods of data collection, this paper examines multi-habited households’ residen-tial satisfaction with a holistic approach where the characteristics of the dwelling unit, the socialnetworks and neighbourhood facilities are all considered in accessing household satisfaction. Fivelow income communities in the Greater Accra Metropolitan Area were studied. The researchfound out that the dwelling characteristics have a negative influence on the residential satisfactionof the respondents. Households derived the most satisfaction from community support but weremoderately satisfied with their neighbourhood characteristics. The significance of this finding onmulti-habited housing development and planning in Ghana reflects in the design, location andmaintenance of such dwellings.

KeywordsAccra, dwelling characteristics, Ghana, low income, multi-habited housing, neighbourhood charac-teristics, residential satisfaction, social networks

Received November 2013; accepted December 2014

Introduction

Globally, urban low income housing unitsare often described as inadequate, lackingthe basic amenities and often found in poorneighbourhoods. These neighbourhoods areovercrowded and characterised with run-down facilities, poor quality buildings, poorenvironmental facilities and poor sanitation

(Addo, 2013; Arku, 2009; Yeboah, 2005).They are often classified as slums and gov-ernments often target these neighbourhoods

Corresponding author:

Irene Appeaning Addo, Regional Institute for Population

Studies, University of Ghana P.O. Box LG 96, Accra, P.O.

Box AF 3229, Ghana.

Email: ireneappeaning@yahoo.com

for urban renewal programmes. Althoughthese neighbourhood characteristics are notcontested, residents’ perception and use ofthese neighbourhoods may be contrary towhat the policy-maker thinks. Research hasshown that households staying in multi-habited houses in the low income commu-nities exhibit very little residential mobilityin their housing decisions (Addo, 2013;Ardayfio-Schandorf, 2012; Rain et al.,2011). Adriaanse (2007) suggests that overtime people develop a sense of ‘dwelling’ or‘being in place’ and that their habitual rou-tines build up a cognitive awareness of theresidential environment to the point thatthey become psychologically fused with it (p.290).

This paper seeks to explore the residentialsatisfaction of multi-habited low income res-idents in Accra in terms of dwelling charac-teristics, social networks and neighbourhoodcharacteristics. Multi-habitation is a situa-tion where a number of related or unrelatedhouseholds share a common compound andamenities. Given the housing deficit experi-enced in Ghana and currently estimated at1.7 million units in 2014, the low incomehouseholds often seek affordable accommo-dation from these multi-habited dwellings.The 2010 population and housing censusestimates that about 51.5% of the dwellingunits in Accra are multi-habited (GhanaStatistical Service (GSS), 2012). The govern-ment of Ghana seeks to address this housingdeficit through the provision of multi-habited dwelling units or compound houses(MWRWH (Ministry of Water Resources,Works and Housing), 2009). It is imperativeto understand the residential satisfactionexperienced by households living in multi-habited dwellings to give a participatoryapproach in urban low income housingdevelopment (Ogu, 2002). According to Liand Wu (2013), there is a need to study thesesettlements from the residents’ perspectivesince current studies have not paid adequate

attention to the feelings of inhabitants abouttheir living places.

Studies on residential housing satisfactionin Ghana observed housing satisfaction as acomposite of the physical and the socialdynamics (Baiden et al., 2011; Sinai, 2001).Sinai (2001) studies on residential housingsatisfaction in Kumasi observed housingsatisfaction as a composite of the physicalcharacteristics (number and size of occupiedrooms, materials that the house was madefrom, availability of water and electricity,availability of cooking and bathing facilities,a veranda, type of toilet) and the socialdynamics (number of households sharingcooking and bathing facilities). Baiden et al.(2011) also measured residential satisfactionin Accra conceptualising dimensions ofhousing to be the physical characteristics ofthe building (house type, housing condition)and the social dynamics within the house (interms of the demand or control the house-hold has, the meaning one attaches to wherehe or she lives and socio-demographic fac-tors). The neighbourhood characteristicswere not actually studied, however neigh-bourhood characteristics play a significantrole in residential satisfaction and householdmobility patterns (Rohe et al., 2013). Hence,this paper will measure residential satisfac-tion as a composite of the physical character-istics of the building in terms of the dwellingcharacteristics, the social dynamics and theneighbourhood characteristics.

Three objectives were set out for thispaper. The first objective is to present thenature of multi-habitation in the study areasand the second objective is to measure thelevel of residential satisfaction among lowincome households in these study areas. Thethird objective is to identify the specific vari-ables that influence residential satisfactionamong multi-habited households. It is antici-pated that findings from this research caninform Ghana’s national housing policy as itseeks to adopt both compound and family

2 Urban Studies

houses as a strategy to address urban lowincome housing needs. Second, it is antici-pated that the findings will contribute to thediscourse on residential satisfaction in lowincome multi-habited dwellings in Ghanaand in other developing countries.

Multi-habited housingdevelopment in Ghana

With housing policies shifting away fromdirect public supply of affordable lowincome houses, individual households havefilled the gap by building low income housesand cashing in on the limited supply ofaffordable housing in the urban centres.Petty landlordism is the provision of incre-mentally built affordable housing by individ-ual developers in the informal sectortargeting urban low income households. It isassociated with urban low income landlordswho, through this means, supplementtheir meagre urban incomes (Datta, 1995;Kumar, 1996). When small landlordism pre-dominates, landlord and tenant have similarsocial-economic backgrounds and a quitecordial relationship. Owners prefer to adver-tise through word of mouth, and they oftenlet out to kin and friends. Landlordism canbe explained on the level of scale in terms ofthe number of rooms occupied by the land-lord in a compound house and it takes theform of multi-habitation of households(Kumar, 1996). As the fortunes of urban lowincome households improve, they undertakehousing projects including building addi-tional rooms which are rented to other lowincome households seeking rental facilities(Aina, 1990; Kumar, 1996). Schlyter (2003)identified that low income groups extendtheir house space constructing ‘illegal’ out-buildings which are let out to lodgers as anincome-generating venture. Verandas areenclosed and used as additional rooms.

Critics of petty landlordism contend thatpetty landlordism is a sign of ineffective

public housing policies and inefficient hous-ing markets (Kumar, 1996). Petty landlord-ism is a reaction to the unaffordable andinaccessible ‘low income’ single family hous-ing that governments have over the yearsattempted to build to address the housingsituation in developing countries. However,these housing interventions have not suc-cessfully addressed the dire housing needs ofthe urban low income households (Aina,1990; Kumar, 1996). Again, the informalityassociated with petty landlordism has pre-vented any enforcement of rent control mea-sures and the quality of such dwelling unitsleaves much to be desired (Datta, 1995).

Renting is the predominant tenurearrangement in such multi-habited dwell-ings. Peil (1994) has suggested that it is dueto the low income households’ inability toown formal housing that has led to the spateof renting among the urban low incomehouseholds. However, UN Habitat (2003)suggest that rental housing policy may offermore opportunities for improving livingstandards of the poor considering the chal-lenges associated with the provision ofowner-occupied housing by governments.Peil (1994) identified that rents charged werea minor percentage of 10–15% of the house-hold income, making it affordable for theurban poor. In Ghana a collection ordemand for an upfront downpayment oftwo to three years rent makes renting unaf-fordable to the urban low income house-hold. However, there are some householdsliving rent-free in multi-habited dwellings.According to the GSS (2008), about 25% ofthe population in Accra lives rent-free andthis form of tenure provide accommodationfor the low income household.

Measuring residential satisfaction

Residential satisfaction gives an indicationof the quality of life of residents and a reflec-tion of the degree to which residents feel that

Appeaning Addo 3

their housing is helping them to achieve theirlivelihood. Residential satisfaction is oftenconceptualised as the difference betweenhouseholds’ actual and expected housingand neighbourhood conditions (Galster andHesser, 1981). Residential satisfaction hasbeen conceptualised in many different waysbut is generally analysed by associatingsatisfaction with internal dwelling design,including facilities, size, cost and design(Buys and Miller, 2012) or the housing unit,neighbourhood and homeownership (Roheet al., 2013).

In measuring residential satisfaction,Amerigo and Aragones (1997) consideredaspects of the neighbourhood and the house.The physical and social environment as wellas the subjective and objective evaluation ofthe housing attributes including residents’opinion of their neighbourhoods offeredimportant insights on which aspects of the set-ting have a greater impact on overall house-holds’ residential satisfaction (Adriaanse,2007; Amerigo and Aragones, 1997). Gibsonet al. (2011) suggested that very specificaspects of the physical or the built environ-ment can influence residential satisfaction asthey impact on mental wellbeing and qualityof life. Neighbourhood facilities also influenceresidential satisfaction in many ways. Theposition of the housing area with respect towork place and other facilities such as dis-tances to town centre, school, police station,hospital, market, shopping centres, publiclibrary, religious buildings, bus and taxi sta-tions are all factors that will influence a resi-dent’s satisfaction (Mohit et al., 2010). Inaddition, green spaces, environmental healthor pollution, upkeep and cleanliness, pace oflife, as well as the social milieu are equallyimportant neighbourhood predictors in resi-dential satisfaction (Rioux and Werner, 2011).Other community factors that are likely topredict housing satisfaction include variablessuch as noise, crime, accidents, security andcommunity relations (Mohit et al., 2010).

Place attachment influences residentialsatisfaction. For example, strong attachmentto an area predicts greater maintenance of arespondent’s house (Galster and Hesser,1981). Inversely, households who are notattached to a place or who consider them-selves as tenants hardly invest in their dwell-ings and are quick to express residentialdissatisfaction and are quick to move.Again, petty landlords who treat multi-habited houses only as an investment with-out much attachment may not be necessarilyconcerned with residential satisfaction.

Saegert and Evans (2003) draw a directlink between poverty, housing and health,suggesting that residing in resource-richhousing environments positions residents toimprove their life circumstances, includinghealth and housing by employing the humanand social capital they accumulate. Thus, liv-ing in inadequate housing situated in neigh-bourhoods lacking social capital and basicinfrastructure may not bring total satisfac-tion (Saegert and Evans, 2003). Hiscock etal. (2001), studying the ontological securityand psycho-social benefits using housingtenure as the measure, found that residentialsatisfaction has to do with having wealth,living in a nice area, living in a larger andbetter quality dwelling and being settled inrelationships and work.

A conceptual framework developed byWeidemann and Anderson (1985) viewedhousing as a home representing the social-physical environment and multifaceted incharacter. Hence, residential satisfactionfrom the occupants’ evaluation revolvesaround three responses to their home: affec-tive, cognitive and behavioural. The affectiveresponse is the feeling of pride and satisfactiona person derives from specific aspects of thehome including the physical environment, thedwelling or the house and the social-physicalenvironment. The cognitive response is thejudgement formed that influences residentialsatisfaction of the individual household.

4 Urban Studies

According to Lu (1999), residential satisfac-tion is a complex cognitive construct that hasbeen approached from different angles by sev-eral researchers but the theoretical underpin-nings have been similar. However, Jayanti(1995) concluded that it is not enough tostudy consumer satisfaction from only thecognitive perspective but also from the affec-tive dimension. Both cognitive evaluationsand affective responses play complimentaryroles in residential satisfaction judgements.Amerigo and Aragones (1997) hypothesisedthat favourable behavourial attitudes such asgood relations with the neighbours, participa-tion in neighbourhood activities and mainte-nance of the house will affect residentialsatisfaction. Together, the cognitive, affectiveand behavioural responses of householdswill determine residential satisfaction amongurban low income households in Ghana.

Study area

The study communities lie within theGreater Accra Metropolitan Area (GAMA)in the Greater Accra Region of Ghana asshown in Figure 1.

Accra, being the capital city, is the hub ofpolitical, economic and migrant activities. Itholds the seat of government and has thehead of various ministries and judiciaryinstitutions located within it. The cosmopoli-tan nature and economic importance ofGAMA have made it a destination for bothinter-regional and intra-regional migration.GAMA has a total population of 3,756,423out of the total regional population of4,010,054 (GSS, 2012). Approximately16.3% of the national population lives inGAMA (GSS, 2012). The population den-sity for the Greater Accra Region is 1236persons per km2 (GSS, 2012). Bordered bythe Gulf of Guinea at the southern side, thespatial expansion of the city is limited toonly three directions, which are the north,east and west.

GAMA displays diverse residential settle-ments ranging from low income settlementsto high income residential areas. Each resi-dential settlement is a mix of migrants andindigenous people living in single and multi-family dwellings including family housesand compound houses. Housing supply inGAMA is mainly from five sources; namelygovernment-sector supply, the corporate sec-tor supply, the not-for-profit supply, individ-ual household incremental supply andinformal housing. However, the low incomehousing supply is mainly achieved throughmiddle to low income individual householdswho build incrementally and rent out roomsas petty landlords but hardly from the formalhousing market segment. The government-sector supplied houses include bungalowsthat are built by individual organisationsand mainly let out to government officialsin managerial positions. Examples of suchpublic houses are those built for the civilservice, police, nurses, fire service and themilitary. These houses are often locatedwithin prime residential areas such asRoman Ridge, Cantonments and Labonein the AMA. Another form of housing sup-ply is by the real estate developers. Thesehouses are classified as high income to mid-dle income houses.

The lower income housing supply ismainly from non-governmental organisa-tions, individual petty landlords and infor-mal housing. These houses have beendescribed as inadequate and MelaraArguello et al. (2013) found that the envi-ronmental and sanitation state of such set-tlements in GAMA is deplorable. However,the share of low income housing supply inAccra is estimated to be 80% of the existinghousing stock (Intsiful, 2004). In GAMA,family houses are the major form of urbanlow income housing provision since this typeof housing provides accommodation for lowincome indigenous households who live rent-free and migrant low income households

Appeaning Addo 5

seeking affordable rental housing within thecity centres.

The five study communities are JamesTown, Tema Manhean, Madina, Accra NewTown and Ashaiman (Figure 2). These com-munities are described as low income com-munities in the GAMA. Madina, Accra New

Town and Ashaiman are predominantlymigrant communities that have expandedover the years. However, the old part ofthese communities displays a lot of multi-habited compound housing. James Townand Tema Manhean are predominantly indi-genous communities occupied by the Ga1

Figure 1. Location map of Greater Accra Region showing GAMA.Source: Adapted from the Survey and Mapping Division of the Lands Commission in Ghana.

6 Urban Studies

tribe who live in family houses. However,unrelated households, in some cases, tend torent rooms from the family houses. Suchhouseholds are expected to conform to thenorms of conduct prevailing in the houses.Generally, the patterns of social relationshipsin family houses are closer than in compoundhouses. The family head in family houses actsas the symbol of authority ensuring thatthere is some continuity to the mutual obliga-tion of giving and receiving assistance fromextended family members. The family head issaddled with the responsibility of efficientlymanaging all social affairs pertaining to theextended family members.2

On the other hand, compound houses, pre-dominantly occupied by migrants, show moreethnic diversity. In these houses one can cer-tainly not choose his/her immediate neigh-bour since only available rooms are given outto prospective tenants. Consequently, mostneighbours are seldom related and display amix of different ethnic background or social

status. A relationship is developed out ofkeeping ‘good terms’ with each other. Themigrant household experiences a social systemwhich is characterised by conflicts and theabsence of traditional forms of security.However, mutual assistance is provided tomembers of the house with each memberrecognising that their positions can alternatebetween a helper and needing help. Thisrelationship is primarily directed towardsachieving households’ livelihood outcome. Inmigrant compound houses, the landlord orlandlady is usually the head of the compoundexercising the power of administration andoverseeing the smooth running of activities inthe house to ensure good cooperation amongthe households by instituting norms to governthe operations in the house. However, socialactivities within each household are con-ducted by the household head. If social activi-ties need to take place within the compound,then permission is sought from the landlord/landlady. This is contrary to what pertains in

Figure 2. Map showing the study areas.Source: Adapted from the Survey and Mapping Division of the Lands Commission in Ghana.

Appeaning Addo 7

the predominantly indigenous family houseswhere the family head has more control overthe social activities held in the house.3

Survey and methodology

Using a cross-sectional approach, a multi-stage cluster random sampling method wasused in the sampling design (United NationsStatistics Division (UNSD), 2005). The sur-vey collected information from approxi-mately 290 respondents in five communitiesin GAMA. Of the total respondents 26.6%were female while 74.4% were male. Theadministered questionnaire captured infor-mation on the social and economic profilesof households by purposefully selecting thehousehold head. In his (her) absence, anadult representative was selected as therespondent. The household head was definedbased on the spatial (within the housingunit), functional (as the main provider forthe household) and structural terms (thesocietal definition of the household head,and this is often the male adult if present) ofa household (Yaro, 2004).

Although the selected communities weregenerally classified as low class residentialareas in1999 (Songsore et al., 2006), somecommunities within Madina and Ashaimanhave developed into middle–high class resi-dential areas. Hence, the study was limitedto the old settlements that had mainly multi-habited compound housing and the ques-tionnaires administered by trained researchassistants. The household survey was con-ducted in the selected communities fromNovember 2009 to September 2010. Ininstances where the respondents were notfluent in the English language the questionswere translated into two predominantGhanaian languages (Ga or Twi) spoken inGAMA. Using proportional sampling of thepopulation in the communities, 34 house-holds were each sampled from James Townand Tema Manhean, while 74 households

were each sampled from Madina, AccraNew Town and Ashaiman. In all a total of290 households were interviewed.

In addition to the quantitative survey, aqualitative data collection method was con-ducted to gather information to confirm, dis-prove, support, ascertain and explain trendsand results obtained from the householdquestionnaires. In-depth interview guidesand focus group discussions were employedduring the data collection. These interviewswere conducted outside the homes to givethe respondents the free will to express them-selves without any fear of being overheardby the landlord and ejected from the house.The perceptions and acceptability of multi-habitation in urban low income housing pro-vision was considered by looking at theinteractions and informal social networksunder multi-habitation, sharing of space andconflicts under multi-habitation, the level ofsatisfaction of households living in multi-habited dwellings and the acceptability ofmulti-habitation as a housing provisionstrategy in the 21st century.

The household respondents were asked toindicate their level of satisfaction of thesevariables on a Likert scale which rangesfrom 1 (very dissatisfied) to 5 (very satisfied).The satisfaction indices were computed byadding up all scores for each variable by arespondent. The total of the scores obtainedfrom the Likert scale is divided by the maxi-mum possible total score and the result isthen multiplied by 100 to obtain the satisfac-tion index of the respondent as a percentage(Ogu, 2002). The satisfaction index for aparticular residential component was calcu-lated according to Ogu (2002):

SIc=

PNi= 1 yi

PNi= 1 yi

3 100 ð1Þ

where SIc is the satisfaction index of arespondent with the component c, of theresidential environment, N is the number of

8 Urban Studies

variables being scaled under c, yi is theactual score by a respondent on the ith vari-able and Yi is the maximum possible scorethat i could have on the scale used.

After calculating for the satisfactionindex for the individual variables, these aresummed to achieve the residential satisfac-tion index for a component. The satisfactionindices were computed for dwelling unit,social networks and the neighbourhoodfacilities. The indices achieved were equallygrouped into quartiles of very low (0–25%),low (25.1–50%), moderate (50.1–75%) andhigh (75.1–100%) satisfaction levels follow-ing Ogu (2002) and Mohit et al. (2010).

The habitability index is computed bysumming the individual scores for eachrespondent for a variable. The sum total isdivided by the total respondents. The indicesare then calculated as a percentage of maxi-mum possible score.

HIx=

PNi= 1 ayix

PNi= 1 Ayix

3 100 ð2Þ

where HIx is the habitability index of a vari-able x, N is the number of respondents, a isthe actual score by a respondent on the ithvariable and A is the maximum possiblescore that i could have on the scale used.

The habitability indices were computedfor the dwelling unit, social networks andthe neighbourhood facilities. These indiceswere grouped into negative (0–49%) moder-ately positive (50–69%) and positive habit-ability (70–100%), following Ogu (2002).

Results

The nature of multi-habitation

Out of a total of 290 respondents, 72%(208) of the households live in multi-habitedcompound and family houses. Althoughthere were generally more male-headedhouseholds, there was no marked difference

in the distribution of male (73%) and female(77%) household heads in James Town.About 32% of the households living in thesehouses had only one room while about 35%had two rooms. Just about 14% of thehouseholds had more than five rooms. Morelow income households in migrant commu-nities occupied single rooms compared withthose in the indigenous communities. Forexample, about 53% of the households inAshaiman stayed in one room. This gives anindication of the extent of overcrowding inthese houses.

A greater proportion of the houses hadmore than three households and only 21%of the houses had less than two households.About 58% of all the households surveyedwere tenants staying in compound houseswhile 38% of the households stayed in fam-ily houses. Just about 4% lived in temporaryshelters. About 50% of the respondents hadstayed in their residential areas for morethan 20 years. It was only in James Townthat about 70% had stayed in the commu-nity for more than 20 years. Only 17% of therespondents in Ashaiman had stayed in theirresidential community for less than 5 years.There were more rent-paying tenants (54%)than rent-free tenants (36%). However,about 63% of the households in JamesTown were not paying rent.

As can be seen in Table 1, bathrooms andelectricity supply are the most shared ame-nities in all the communities. This was fol-lowed by toilet facilities, waste collection andpipe-borne water in four of the communitiesexcept James Town. A majority of the com-pounds in James Town did not have toiletfacilities and a proper waste collection sys-tem. Kitchen and lobby are the least sharedamenities in the house. In Ashaiman, only7% had lobbies and verandas. These spaceshave been enclosed into sleeping rooms, pre-venting adequate ventilation into the innerrooms. In some instances, the verandas had

Appeaning Addo 9

Tab

le1.

Am

enitie

ssh

ared

inm

ulti-hab

ited

com

pound

and

fam

ilyhouse

sby

com

munity.

Loca

lity

Faci

litie

sSh

ared

(%)

Not

shar

ed(%

)D

onot

hav

e(%

)

Ind

igen

ou

sco

mm

un

itie

sJa

mes

Tow

nN

=24

Bat

hro

om

/show

er45.8

12.5

41.7

Ele

ctri

city

supply

62.5

20.8

16.7

Toile

t8.3

091.7

Was

teco

llect

ion

8.3

12.5

79.2

Pip

e-born

ew

ater

12.5

087.5

Kitch

en12.5

12.5

75.0

Lobby

16.7

8.3

75.0

Tem

aM

anhea

nN

=22

Bat

hro

om

/show

er90.9

4.5

4.5

Ele

ctri

city

supply

54.5

45.5

0To

ilet

50.0

9.1

40.9

Was

teco

llect

ion

45.5

45.5

9.0

Pip

e-born

ew

ater

36.4

13.6

50.0

Kitch

en33.3

52.4

14.3

Lobby

18.2

36.4

45.5

Mad

ina

N=

56

Bat

hro

om

/show

er91.1

7.1

1.8

Ele

ctri

city

supply

94.6

3.6

1.8

Toile

t87.5

1.8

10.7

Was

teco

llect

ion

82.1

17.9

0Pip

e-born

ew

ater

60.7

039.3

Kitch

en57.1

19.6

23.2

Lobby

44.6

28.6

26.8

Mig

ran

tco

mm

un

itie

sA

ccra

New

Tow

nN

=51

Bat

hro

om

/show

er92.2

7.8

0Ele

ctri

city

supply

86.3

9.8

3.9

Toile

t78.4

5.9

15.7

Was

teco

llect

ion

40.8

55.1

4.1

Pip

e-born

ew

ater

58.8

7.8

33.3

Kitch

en43.1

27.5

29.4

Lobby

47.1

29.4

23.5

Ash

aim

anN

=55

Bat

hro

om

/show

er72.7

25.5

1.8

Ele

ctri

city

supply

47.3

52.7

0To

ilet

25.5

23.6

50.9

Was

teco

llect

ion

21.8

74.5

3.6

Pip

e-born

ew

ater

18.2

20.0

61.8

Kitch

en38.9

46.3

14.8

Lobby

7.3

38.2

54.5

Not

e:T

he

surv

eydat

ais

now

five

year

sold

and

som

ediff

eren

ces

are

antici

pat

edin

findin

gsin

2014

afte

rth

een

forc

emen

tofby-

law

sto

intr

oduce

dom

estic

toile

tsin

all

house

sin

Acc

ra.

Sour

ce:Fi

eld

surv

ey,200

9.

10 Urban Studies

been enclosed and converted into additionalliving rooms either by the landlord or by theoccupants. One respondent commented:

There are no kitchens, toilets and verandas inour [family] house and many other [family]houses in James Town because all of themhave been converted to sleeping rooms. Theissue is the land we occupy is insufficient for usto build kitchens and toilets. How can we havethese facilities if the rooms in the family houseare not adequate to cater for the expandingfamily members.

Respondents who did not have pipe-bornewater in their houses fetched water fromcommunity stand pipes or from wells. Therewere no bathrooms specifically built for onlymales or females. They shared these facilitiestogether. Sometimes the heads of the houseor family heads have their own locked facili-ties, which they do not share with the rest ofthe household.

While the migrant communities per-formed better in terms of provision of facili-ties, most of the compounds in theindigenous communities lacked these ame-nities. However, James Town was the worstcommunity in terms of toilet provision andwaste collection while Madina had the mostamenities in the compound houses.

Sharing of household items is usuallyreciprocal in nature. A member of the focusgroup living in James Town commented onthe reciprocal nature of sharing in multi-habited family houses:

I live with my sisters. Sometimes we contributeand buy food ingredients and we cooktogether and eat. I sometimes leave my chil-dren with my sister and she also does.

Where the reciprocity is lost, conflicts occurand the person at the receiving end isdescribed as a beggar. This creates a sourrelationship among the households.

The results indicate that only 16% of therespondents were landlords and 84% wereboth rent-paying and rent-free tenants.About 61% of the total households sharedthe responsibility of maintaining the housewith other family members, tenants and theirlandlord/landlady. About 16.3% of therespondents claimed that the maintenance ofthe compound house is done by a residentlandlord and about 6.3% responded that itis done by a non-residing landlord. Justabout 11.2% of the respondents claimed thatmaintenance of the house is done solely bythe tenants. About 5.4% of the respondentsclaimed that the houses they live in have notbeen renovated, repaired or maintained. Itwas only in James Town that 62.5% of therespondents claimed that only tenants main-tained the family houses. Generally, housesthat had resident landlords were better main-tained than houses with absentee landlords.

Cleaning of compounds and bathroomsare usually shared among the female mem-bers of the house while the male memberstake up the responsibility of maintaining thehouse. The women and children clean thecommunal areas while men buy the cleaningagents. Cleaning is done in turns by house-holds or collectively. The study found outthat about 66% of the households rotateddaily in cleaning the house. About 29% ofthe households collectively clean the houseand this sort of arrangement is prevalent inthe family houses. Others (5%) employedother means of cleaning the house usinghired labour and professional cleaners.Waste disposal in both the family housesand the compound houses are done throughthe hiring of the services of waste manage-ment companies.

The major sources of conflict stem fromthe mode of sharing, the mode of cleaningand the interpersonal relationships withinthe house. These conflicts were, in particu-lar, accentuated in family houses in James

Appeaning Addo 11

Town compared with compound houses inMadina and Accra New Town. This situa-tion is reported by a respondent in a com-pound house:

Many problems associated with compoundhouse living is that other members do notclean the house very well especially the toilets.You have to do the cleaning always becauseyou are particular with the cleanliness of thebathroom and toilet and you do not want toget sick.

However, when questions were asked aboutconflicts during the household survey, amajority of the households responded thatthey were not experiencing any form of con-flict living together and sharing facilities inthe house. This was contrary to the responsereceived during focus group discussions andin-depth interviews held with respondentsoutside their homes. The difference inresponses maybe attributed to fear of ejec-tion if overheard by the landlord or land-lady. The interviews revealed that conflictoften occurs in the mornings when house-holds have to rush to work and need to usethe only bathroom on the compound, and atthe end of the month when the electricityand water bills have to be shared for the

households. The disagreements occur overthe method of sharing the total bill. Some ofthe arguments were that some householdsown more electrical equipment than othersand hence they should not share the electric-ity bill equally. This has led to the installa-tion of separate meters for each room uniton some compounds. Conflicts can occurwhen there are petty jealousies and badblood between households. The lack ofmaintenance of the compound and disagree-ment over payment of electricity bill exhib-ited by the tenants in Box 1 resulted in aconflict that had to be partially resolved inthe rent control department.

Household residential satisfaction undermulti-habitation

Household residential satisfaction is tied tothe dwelling characteristics, the kind of rela-tionship existing between households andaccess to neighbourhood facilities. In thisstudy 16 residential satisfaction variablesgrouped under three key components wereused (Table 2).

Figure 3 shows the satisfaction differ-ences in dwelling characteristics, social net-works and neighbourhood characteristics

Box 1: The complaint of a male landlord

I have 16 tenants living in my house in Agbogba, Madina. They do not maintain thecompound of the house by not weeding or cleaning the compound. When the electricitybill is brought, they do not pay and they do not pay for refuse collection. The house isunkempt. The Madina Municipal office fined me and took me to court so I came to tellthem to maintain the house. Together they hauled me to the rent control department(in Accra) that they will not continue to stay in the house and so I have to refund theirmonies (rent advance) to them. They were given one month to stay till they vacated thepremise. When they were leaving, some owed me four to six months electricity bill.They have left without paying.

Source: Ofori (2011) TV3.

12 Urban Studies

for the five communities. Over 40% of thehouseholds in Madina (MA) and AccraNew Town (ANT) expressed high

satisfaction with the dwelling characteristicscompared with households in Ashaiman(AS) and James Town (JT). In Ashaiman

Figure 3. Proportion of household indicating level of satisfaction for all the components.

Table 2. Variables used in measuring residential satisfaction in the three components.

Dwelling (physical)characteristics (D)

Social networks (S) Neighbourhood facilities (N)

Number of roomsRoom sizeCost of housingUtilities availablePrivacy in the compound

Community supportCommunity safetyProximity to friendsand relations

Proximity to Central Business District (CBD)Proximity to the marketProximity to workplacesProximity to community standpipeProximity to public toiletProximity to bus stationProximity to recreational facilities.Environmental cleanliness

Appeaning Addo 13

and James Town, almost half of the house-holds expressed moderate satisfaction withthe dwelling characteristics. Generally, mostof the households in all the five communitieswere moderately satisfied with the existingsocial networks but James Town, TemaManhean (TM) and Ashaiman had over60% of the households expressing moderatesatisfaction as compared with Madina (MA)and Accra New Town (ANT). Similarly,most of the households were moderately sat-isfied with their neighbourhood characteris-tics but James Town and Ashaiman hadover 80% of the households expressing mod-erate satisfaction. However, less than 5% ofthe households in James Town, TemaManhean and Ashaiman expressed highsatisfaction with the neighbourhoodcharacteristics.

Overall, the results showed that 78.5% ofthe respondents were moderately satisfiedwith their residential areas. However, 77%of the respondents were moderately satisfiedwith the neighbourhood characteristics ascompared with 46.4% who were moderatelysatisfied with the dwelling characteristics.This was followed by 62% who were

moderately satisfied with the existing socialnetworks.

The results indicate that respondents’ dissa-tisfaction was pronounced in the existing socialnetworks and the dwelling characteristics.Very few respondents expressed dissatisfactionwith the neighbourhood characteristics, andespecially in Madina. Overall, 38.4% of therespondents were highly satisfied with theirdwelling characteristics.

Habitability index

From Table 3, the habitability indices forthe variables of the dwelling unit rangedfrom 47.4 (number of rooms) to 59.5 (utili-ties available). The habitability index forsocial networks ranged from 73.4 (commu-nity support) to 47.4 (proximity to friendsand relations). The habitability index forneighbourhood characteristics ranged from68.7 (proximity to recreational facilities) to41.0 (proximity to market).

Community support received positivehabitability index of 73.4%. This was fol-lowed by proximity to recreational facilities(68.7%) and proximity to public toilet

Table 3. Rank ordering of habitability indices for the various components.

Component Description of variables Habitability indices

S Community support 73.4N Proximity to recreational facility 68.7N Proximity to public toilet 62.5D Utilities available in the house 59.5N Environmental cleanliness 59.2S Community safety 58.1N Proximity to community pipe 57.3N Proximity to central business district 55.1D Privacy in the compound 53.1D Cost of housing 51.9D Room size 47.8S Proximity to friends and relations 47.4D Number of rooms 47.3N Proximity to workplace 44.4N Proximity to bus station 43.3N Proximity to market 41.0

Notes: S: social networks; N: neighbourhood facilities; D: dwelling facilities.

14 Urban Studies

(62.5%), both described as moderately posi-tive. In all three components some of thevariables were described as moderately posi-tive and others as having negative habitabil-ity index. No particular component wascompletely described as having only positiveor negative habitability indices.

Discussion

The results show that there is a direct corre-lation between the facilities available in thehouse and level of residential satisfaction.Most of the households in Madina andAccra New Town expressed high residentialsatisfaction with the dwelling characteristics.This is explained by the fact that compoundhouses in these two communities had betterutilities and amenities in the house, althoughthey were shared.

With multi-habited dwellings being pro-posed as a key strategy to address housingsupply in Ghana and increase accessibilityto affordable housing, it is important toidentify and understand the key variablesthat contribute to residential satisfaction inmulti-habited dwellings especially for com-pound houses. This is necessary since provi-sion of housing is mainly through the privatereal estate developers and individual pettylandlords. Public sector-led housing supplyis just about 7% (GSS, 2012). Although theGhana government is beginning to intervenein the housing supply, it is targeted at thepublic sector and not directed to the privatesector.

The gap in affordable housing supply isfilled by petty landlords who incrementallybuild multi-habited compound housing.Urban low income households become pettylandlords supplying other low incomehouseholds with affordable housing throughthe renting out of single rooms in thesehouses. As migration increases, these infor-mal housing developments become the mainchannels to supply affordable housing onto

the market. Formal housing supplies arelimited, expensive and inaccessible to thelower income groups. Multi-habited familyhouses are also an alternative means thathouseholds find accommodation and liverent-free. However, the expanding familymembers living in family houses result inovercrowding and conflict in family houses.

This assertion was confirmed by the studyshowing that there is overcrowding in multi-habited houses in the study area given thenumber of rooms available to each house-hold and the household size. Most of thefamily houses lack spaces for expansion butaccommodate about three generations ofhouseholds. Lack of adequate resources hasprevented a number of the low incomehouseholds from developing and moving outof the family houses or the compoundhouses. The study also indicates that theshared facilities in multi-habited dwellingsare inadequate and lack regular maintenancealthough houses with resident landlords werebetter maintained. These landlords ensurethat there is regular collection of monthly oroccasional fees for maintenance especiallyduring the Homowo4 festival and other cele-brations. Inadequate and undefined mainte-nance schedules of shared facilities oftenresults in conflicts. In such situations, resi-dential satisfaction of households decreases,leading to dissatisfaction among householdssince there is a direct relationship betweenresidential satisfaction and dwelling facilities.The dissatisfaction is pronounced in JamesTown, an indigenous community andAshaiman, a migrant community.

Again, the limited number of roomsavailable to a household (one room and anadditional room created by enclosing theveranda) reduced satisfaction among house-holds. The petty landlords build small unitsaffordable to the low income household andalso maximise the use of the land. However,they lack capital for multi-level develop-ment. This situation greatly reduces a

Appeaning Addo 15

household’s habitability index regardingroom size and the number of rooms avail-able to each household. This dissatisfactionis buttressed by a respondent in a compoundhouse:

If the government is planning to intervene thenhe should be thinking of building a hall and achamber compound house (instead of singleroomed compound house) so that the unit ismore spacious. This will also be affordable tothe very poor (compared to the cost of con-structing two-bedroom bungalow), and we cantake advantage of the inter-relations existingin the compound house. We can receive helpfrom each other when we are managing oureveryday needs.

The other variable associated with dwellingcharacteristics is the shared utilities availableon the compound. They are inadequate andconflicts arise over their use. In some cases,as observed by Arku et al. (2012) and docu-mented by AMA (2011), some of the housesin Accra lack toilet facilities (although this isnow illegal). The situation poses problemsfor the local government as sanitation andwaste management issues increase throughindiscriminate dumping of refuse (MelaraArguello et al., 2013). The situation is notdifferent for indigenous and migrant com-munities, although compound houses inmigrant communities such as Madina andAccra New Town seem to have better facili-ties. Hence proximity to a public toilet isone of the key parameters for residentialsatisfaction. According to GSS (2008) onlyone out of every ten households in Ghana(22.2% in urban areas and 1.1% in ruralareas) has access to a flush toilet. Ideally putby a respondent in a compound house:

Each family should have its own toilet, bath-room and kitchen. This is the only way tosolve the problem (of conflicts). This is becauseevery human interaction from different fami-lies having a wife and children in addition tosharing facilities would definitely lead to

conflicts. You will want to live peacefully butanother person might want to pick a quarrelwith you.

The situation has affected satisfaction levelsof residents in the low income communities.From the study, residents living in com-pound houses were moderately satisfied withtheir dwelling characteristics, neighbour-hood characteristics and existing socialnetworks. However, the proportion ofrespondents moderately satisfied with theirneighbourhood characteristics is higher thanthe other two components. Several studieshave documented the deplorable neighbour-hoods in which urban low income house-holds stay which affect their residentialsatisfaction. Although households in thisstudy are not any different (Arku et al.,2011; Baiden et al., 2011), their response(moderately satisfied) could be attributed totheir long residency in the various commu-nities. The respondents may exhibit hopeless-ness and despair knowing that their situationmay never change and multi-habited housesmay be the only option to access affordablehousing in Accra. Others hoped they couldone day build and move on and so they justadapt to the existing situation. This findingsupports what was observed in Nairobislums, where slum dwellers claimed that theywere satisfied with their living condition(Mudege and Zulu, 2011). Another reasonmay be attributed to the ‘sense of place’households feel after living in such neigh-bourhoods for a long time (Adriaanse,2007), as one respondent in a family houseputs it:

I have lived here all my life. This house belongsto my grandparents and it was bequeathed tomy parents. I now occupy the room that myparents had as their share of the inheritance.

Increased satisfaction of urban low incomehouseholds is highly dependent on commu-nity support received. This is because

16 Urban Studies

households often have low incomes and tendto depend on each other for support. Thesocial net of the traditional extended familiesin the rural areas are lost in the urban centresas a result of weak family systems in theurban centres and the nuclearisation of theurban family. The monetisation of the urbaneconomy also takes a toll on the migranthousehold in terms of paying to access themost basic necessities such as housing whichin the rural areas would have been free. Thefinding supports earlier research of Addo(2013) where it was established that house-holds living in multi-habited dwellingsdepend greatly on the existing informalsocial networks. The results suggest thathouseholds in the indigenous communitieswere better satisfied with the existing socialnetworks as compared with households inthe migrant communities. The indigenouscommunities have their relatives stayingclose by. According to Ardayfio-Schandorf(2012), these support systems act as socialcapital for the low income households asthey strive to achieve their livelihood out-comes. The social support systems act asmechanisms for pooling income, sharingconsumption and household relations. Thistogether act as ‘critical safety nets’ and‘shock absorbers’ by reducing the vulnerabil-ity of its members (Moser, 1998). Accordingto a respondent in a compound house:

If someone is undergoing any difficult timesthe person can receive help from co-membersof the house as compared to those living inself-contained houses.

Residents’ habitability index in multi-habited dwellings decreased when proximityto workplaces, marketplaces and bus sta-tions reduced. Most urban low incomehouseholds work in the informal job marketsfound in the market places. Unfortunately,the development of multi-habited dwellingsby individual petty landlords is dependent

on the available land, which is often in thesuburban areas, further away from theworkplaces and the market places and busstations. According to UN Habitat (2013),coordinating and integrating urban trans-port and land development is imperative forcreating sustainable urban futures. Notably,the design and layout of a city strongly influ-ences travel demand.

Conclusion

In conclusion, the research has addressedthe three objectives: first the nature of multi-habited dwellings has been presented bylooking at the gender distribution of thehousehold heads, the household size, thenumber of rooms in multi-habited dwellings,and the length of stay. The living arrange-ment pertaining to multi-habited dwellingsin terms of sharing and maintenance of facil-ities as well as the social dynamics prevailingin such form of living were all highlightedunder objective one. The second and thirdobjectives examined residential satisfactionlevels and factors contributing to residentialsatisfaction among households. The resultsshow that residential satisfaction of house-holds living in multi-habited compound andfamily houses is linked to the dwelling char-acteristics, neighbourhood characteristicsand the existing social networks.

However, the research found that inade-quate dwelling characteristics have a nega-tive impact on the residential satisfaction ofthe respondents. Generally, most of thehouseholds were dissatisfied with their dwell-ing characteristics. The habitability index formost neighbourhood facilities was low. Thesignificance of this finding on multi-habitedhousing development and planning inGhana reflects in the design, location andmaintenance of such dwellings. Multi-hab-ited dwellings will need to have adequaterooms and should be located in places whereresidents will have access to transport and

Appeaning Addo 17

economic facilities (market and work places).Efficient landlord maintenance schemes willneed to be developed in partnership withtenants, local government authorities andrent control departments. Although theaddition of extra rooms will lead to highrental costs, and may result in gentrificationof low income households (Rose, 2013) inthese communities, provision of housingsubsidies could curtail this phenomenon.Moreover, research has shown urban lowincome households’‘willingness-to-pay’ whentheir housing is improved (Ayoola andAmole, 2014; Melara Arguello et al., 2013).Therefore, ultimately income improvementstrategies will be the most effective tool forenhancing housing consumption.

Proximity to recreational areas alsoplayed a key role in resident’s satisfactionand that aspect should be promoted whenplanning for urban low income housing.Multi-habited houses, either compound orfamily houses, need to have adequate sharedfacilities such as toilets and baths. This willcurtail indiscriminate littering of the environ-ment with human waste and its associatedchallenges in sanitation (Melara Arguello etal., 2013) and promote good public health.There will also be fewer conflicts over sharedfacilities. Unfortunately, inner city lands areused up and hence low income housingdevelopment will need to be in the suburbanareas or the peripheries of the city. However,provision of efficient transport systems willenhance movement of households andgoods.

Given that urban housing development inGhana is not matched with infrastructuresupply, and left in the hands of private realestate developers (UN Habitat, 2011), thereshould be policies to drive urban infrastruc-ture expansion, enforcement of local govern-ment by-laws to ensure that each house hasadequate amenities and utilities such as bath-rooms, toilets, electricity and water supply.

Other places in Dar es Salaam and Kampalahave encouraged local development banks toinvest in sanitation infrastructure (Pieter vanDijk et al., 2014). Individual petty landlordsshould be encouraged to develop multi-levelhousing that will allow higher populationdensities but with less overcrowding. Thiscan be achieved through resourcing andfinancing individual and real estate develo-pers with housing finance loans. Putting inplace proper management strategies will helpmaintain the multi-habited dwellings andcurtail conflicts that arise out of shared man-agement. The issue of location in the devel-opment of multi-habited urban low incomehousing in cities should be further exploredas a means of addressing residential satisfac-tion of low income households in Accra andin other developing countries in terms ofaccessibility.

Funding

This research received no specific grant from anyfunding agency in the public, commercial, or not-for-profit sectors.

Notes

1. The Ga are an indigenous minority ethnicGroup from Greater Accra Region compris-ing just 7.4% of the national population(GSS, 2012).

2. Fieldwork by the author for PhD thesis.3. Observations made by the author during the

field survey.4. An annual festival of the Ga people.

References

Accra Metropolitan Assembly (AMA) (2011)

Participatory Slum Upgrading and Prevention

Millennium City of Accra, Ghana. Available at:

http://ama.gov.gh/resources/docs/PSUP%20-%

20COMPILED%20DOCUMENTATION%20

COMPONENT%20-%20edited%202.pdf (acc-

essed June 2014).Addo IA (2013) Perceptions and acceptability of

multihabitation as an urban low income

18 Urban Studies

housing strategy in Greater Accra Metropoli-

tan Area, Ghana. Urban Forum 24: 543–571.Adriaanse CCM (2007) Measuring residential

satisfaction: A residential environmental satis-

faction scale (RESS). Journal of Housing and

the Built Environment 22(3): 287–304.Aina TA (1990) Petty landlords and poor tenants

in a low-income urban settlement in Metropol-

itan Lagos, Nigeria. In: Amis P and Lloyd PJ

(eds) Housing Africa’s Urban Poor. Manche-

ster: Manchester University Press, pp. 87–102.Amerigo M and Aragones JI (1997) A theoretical

and methodological approach to the study of

residential satisfaction. Journal of Environmen-

tal Psychology 17: 47–57.Ardayfio-Schandorf E (2012) Urban families and

residential mobility in Accra. In: Ardayfio-

Schandorf E, Yankson PWK and Bertrand M

(eds) The Mobile City of Accra: Urban Fami-

lies, Housing and Residential Practices.

CODESRIA: Dakar, pp. 47–72. CODESRIA,

DakarArku G (2009) Housing policy changes in Ghana

in the 1990s: Policy review. Housing Studies

24(2): 261–272.Arku G, Luginaah I, Mkandawire P, et al. (2011)

Housing and health in three contrasting neigh-

bourhoods in Accra, Ghana. Journal of Social

Science and Medicine 72: 1864–1872.Arku G, Luginaah I and Mkandawire P (2012)

‘You either pay more advance rent or you

move out’: Landlords/ladies’ and tenants’ dilem-

mas in the low-income housing market in Accra,

Ghana. Urban Studies 49(14): 3177–3193.Ayoola A and Amole D (2014) The value of

housing among the poor in Ilesa, Osun State

Nigeria. Architecture Research 4(1A): 45–54.Baiden P, Arku G, Luginaah I, et al. (2011) An

assessment of residents’ housing satisfaction

and coping in Accra, Ghana. Journal of Public

Health 19(1): 29–37.Buys L and Miller E (2012) Residential satisfac-

tion in inner urban higher-density Brisbane,

Australia: Role of dwelling design, neighbour-

hood and neighbours. Journal of Environmen-

tal Planning and Management 55(3): 319–338.Datta (1995) Strategies for urban survival?

Women landlords in Gaborone, Botswana.

Habitat International 19(1): 1–12.

Galster GC and Hesser GW (1981) Residential

satisfaction compositional and contextual cor-

relates. Environment and Behavior 13(6):

735–758.Ghana Statistical Service (GSS) (2008) Ghana Liv-

ing Standards Survey Report of the Fifth Round

(GLSS 5). GSS.Ghana Statistical Service (GSS) (2012) 2010 Popula-

tion and Housing Census: Summary Report. GSS.Gibson M, Thomson H, Kearns A, et al. (2011)

Understanding the psychosocial impacts of

housing type: Qualitative evidence from a

housing and regeneration intervention. Hous-

ing Studies 26(4): 555–573.Hiscock R, Kearns A, MacIntyre S, et al. (2001)

Ontological security and psycho-social bene-

fits from the home: Qualitative evidence on

issues of tenure. Housing, Theory and Society

18(1–2): 50–66.Intsiful GWK (2004) Housing and Urban Develop-

ment Initiatives as Means Towards the Allevia-

tion of Poverty. RICS Foundation. Our

Common Estate Research Paper Series.Jayanti RK (1995) The relative influence of affec-

tive and cognitive factors in determining ser-

vice encounter satisfaction. Journal of

Consumer Satisfaction, Dissatisfaction & Com-

plaining 8: 147–154.Kumar S (1996) Lardlordism in Third World

urban low-income settlements: A case for fur-

ther research. Urban Studies 33(4-5): 753–782.Li Z and Wu F (2013) Residential satisfaction in

China’s informal settlements: A case study of

Beijing, Shanghai, and Guangzhou. Urban

Geography 34(7): 923–949.Lu M (1999) Determinants of residential satisfac-

tion: Ordered logit vs. regression models.

Growth and Change 30(2): 264–287.Melara Arguello JE, Grant R, Oteng-Ababio M,

et al. (2013) Downgrading – An overlooked

reality in African cities: Reflections from an

indigenous neighborhood of Accra, Ghana.

Applied Geography 36: 23–30.Ministry of Water Resources, Works and Hous-

ing (MWRWH) (2009) National Shelter Policy

(draft). MWRWH.Mohit MA, Ibrahim M and Rashid YR (2010)

Assessment of residential satisfaction in

newly designed public low-cost housing in

Appeaning Addo 19

Kuala Lumpur, Malaysia. Habitat Interna-

tional 34(10): 18–27.Moser CO (1998) The asset vulnerability frame-

work: Reassessing urban poverty reductionstrategies. World Development 26(1): 1–19.

Mudege NN and Zulu EM (2011) In their ownwords: Assessment of satisfaction with residen-tial location among migrants in Nairobi slums.Journal of Urban Health 88(2): 219–234.

Ofori O (2011) Issues on urban housing and rentcontrol. TV3 television Mission programme.Telecasted on 6 November 2011, during the 7

O’clock news.Ogu VI (2002) Urban residential satisfaction and

the planning implications in a developing worldcontext: The example of Benin City, Nigeria.International Planning Studies 7(1): 37–53.

Peil M (1994) Urban housing and services inAnglophone West Africa: Coping with aninadequate environment. In: Main H and Wil-liams SW (eds) Environment and Housing in the

Third World Cities. Chichester: John Wiley.Pieter van Dijk M, Etajak S, Mwalwega B, et al.

(2014) Financing sanitation and cost recoveryin the slums of Dar es Salaam and Kampala.Habitat International 43: 206–213.

Rain D, Engstrom R, Ludlow C, et al. (2011)Accra Ghana: A City Vulnerable to Flooding

and Drought-Induced Migration. UN-Habitat(ed.) Background paper.

Rioux L and Werner C (2011) Residential satis-faction among aging people living in place.Journal of Environmental Psychology 31(2):158–169.

Rohe WM, Van Zandt S and McCarthy G (2013)The social benefits and costs of homeowner-ship: A critical assessment of the research. In:Mueller EJ and Tighe JR (eds) Affordable

Housing Reader. Routledge: New York. pp.196–213.

Rose D (2013) Rethinking gentrification: Beyondthe uneven development of Marxist urban the-ory. In: Brown-Saracino J (ed.) The Gentrifica-tion Debates: A Reader. Routledge: UK.Chapter 16, pp. 195–210.

Saegert S and Evans GW (2003) Poverty, housing

niches, and health in the United States. Journal

of Social Issues 59(3): 569–589.

Schlyter A (2003) Multi-Habitation: Urban Hous-

ing and Everyday Life in Chitungwiza, Zim-

babwe. Research report no. 123. Uppsala:

Nordic Africa Institute.Sinai I (2001) Intraurban housing mobility in a

traditional West African city: Shelter or busi-

ness decision. Urban Studies 38(3): 535–540.Songsore J, Nabila JS, Yangyuoru Y, et al. (2006)

Environmental Health Watch and Disaster

Monitoring in the Greater Accra Metropolitan

Area (GAMA), Ghana. University Press: Uni-

versity of Ghana, Legon.UN Habitat (2003) The Challenge of the Slums:

Global Report on Human Settlements. United

Nations Centre for Human Settlements. Lon-

don: Earthscan.UN Habitat (2011) Ghana Housing Profile. Nair-

obi: United Nations Human Settlements

Programme.UN Habitat (2013) Planning and Design for Sus-

tainable Urban Mobility: Global Report on

Human Settlements 2013 – Policy Directions.

London and Sterling, VA: Earthscan Publica-

tions Ltd.United Nations Statistics Division (UNSD)

(2005). Household Sample Surveys in Develop-

ing and Transition Countries. United Nations

Statistics Division.Weidemann S and Anderson JR (1985) A concep-

tual framework for residential satisfaction. In:

Altman I et al., Home Environments. Springer:

New York, pp. 153–182.Yaro JA (2004) Combating food insecurity in

Northern Ghana: Food insecurity and rural live-

lihood strategies in Kajelo, Chiana and Korania.

PhD thesis, Department of Sociology and

Human Geography. Oslo: University of Oslo.Yeboah IEA (2005) Housing the urban poor in

twenty-first century Sub-Saharan Africa: Pol-

icy mismatch and a way forward for Ghana.

GeoJournal 62(1–2): 147–161.

20 Urban Studies