ACCRA GHANA by chenmeixiu


									                                               ACCRA | GHANA
   African Union for Housing Finance

                             Ashna Mathema
                            October 15, 2006
I. Introduction
We selected a stratified sample across various urban poor neighbourhoods in Accra to conduct our interviews.
These are part of the qualitative analysis undertaken for the AUHF study, to provide an in-depth picture of the
living conditions and priorities of the urban poor in Accra. Although the data collected is not statistically
significant, they do present an in-depth understanding of the situation, and help draw some broad conclusions
with regard to the characteristics of these households.

A quantitative socio-economic survey is currently underway. Once complete, it will be used to validate, verify
and/ or substantiate the findings from this qualitative analysis, and vice versa.

Field Work
Household Interviews. Twenty-four household interviews were conducted in five settlements in Accra:
Jamestown, Nima, and Mamponsa, Neo Plan Station (Circle) railroad settlement, and Sodom and Gomorrah.
One additional interview was conducted in Kotobabi. These interviews were 2-3 hours long, and included
questions related to household structure, relationship with other plot residents, migration history, access to and
expenditure on housing and infrastructure, property title/ tenure, income and employment, access to finance,
and development priorities.

The first three settlements—Jamestown, Nima, and Mamponsa—may be categorized as “Legal/ de facto

                                                                                                                                                         General Findings
tenure” settlements (to differentiate them from “squatters”), meaning that most people in these areas were
allocated the land or given permission to occupy the property by the “owner.” To further clarify, most people in
these settlements do not have titles, and the property is not registered; it is de facto ownership, and recognized
by the local leadership.

The other two—NeoPlan Circle railroad settlement, and Sodom and Gomorrah—primarily comprise of
squatters, although there are some “Legal/ de facto tenure” occupants of those areas as well.

Interviews with Market Women. In addition to these household interviews, additional interviews were carried
out with market women (and 1 man) who had signed up for the HFC micro-lending program. The structure of
these interviews was slightly different in that they were geared more at extracting their experiences with the
program, and to draw lessons from it to inform the design of future programs aimed at increasing access to
finance for the urban poor.

Interviews with Micro-lenders. We also conducted interviews with the three major lenders: HFC, Sinapi, and
Pro-Credit to understand their programs and target groups.

                                                                                                                     African Union for Housing Finance
Data Analysis
Most areas showed a mix of “legal” and “illegal” residents (for example, in Mamponsa, while most plots were
allocated, one of our interviewees occupied a plot 100 mts from the beach, which whether or not it was
allocated, cannot be deemed fit for habitation), or people who worked in the area and lived somewhere else (as
was the case in Nima with an electronics workshop owner, and in Sodom and Gomorrah, where one person ran
a bar, and another one a truck repair workshop, but both lived elsewhere). Due to these overlaps, and the
inability to clearly define each of these areas as “legal” or “illegal”, the analysis has been based on data
disaggregated by tenure, legal status, and income levels of individual households.

                                               Renters                      N=5
                        1. By TENURE
                                               Owners                       N=19

                                               Squatters                    N=10
                        2. By LEGAL STATUS
                                               Legal/ de facto tenure       N=14

                                               < =325K pci                  N=12
                        3. By INCOME
                                               >325K pci                    N=12

                                               Overall                      N=24

Data gathered from the interviews with the market women has been analyzed separately, to allow a “before”
and “after” comparison of access to finance. In other words, the findings that are presented on informal

                                                                                                                                                     General Findings
settlements (above) do not include the cases of the market women.

It must be emphasized again that this data is based on a small stratified sample, not a statistical
sample, but it has enough breath and depth to provide a clear overall picture. The intent of the analysis
here, in other words, is to give more context to the findings from the field research, to enable the reader
to quantify the otherwise subjective and qualitative conclusions, with numbers drawn from real cases
and real people. It is anticipated that the statistical limitations will be more than compensated by the
depth of the stories of the people.

The next section summarizes the key findings, and is followed by a more detailed discussion based on the data.
Section IV presents the interviews in detail.

                                                                                                                 African Union for Housing Finance
II. Key Highlights
1. Family Houses in Old Accra

-   Most of the families residing in family houses in Old Accra are poor. Their vulnerability is made worse by
    the fact that these households are also large in size, often women-headed, with several extended
    family members and one or more single mothers. The poorest typically have the largest families, and
    the most number of dependents, and unemployed persons.

-   Living conditions in many of these settlements are deplorable. Water and sanitation facilities are practically
    non-existent within plots. People mostly buy water, at 2-3 times the actual cost, and use public toilets which
    are expensive, and often in short supply, and ill-maintained. Household expense on services is15-20
    percent of total household expenditure, and this includes zero expense for housing.

-   Places like Jamestown already have the trunk infrastructure in place. Still, there are very few houses that
    have tapped into it to build individual toilets, or to get a water connection. This, they say, is because there
    are too many “owners” to the property, and it is almost impossible to come to an agreement on any

    issue related to investment. Also, some households earn less than the others, and cannot pay, which puts
    an uneven share of the burden on those who can.

                                                                                                                                                          General Findings
-   The quality of construction is mixed: very poor in some places, alright in others. However, in general, the
    floor-area-ratio is extremely low, with most structures being one or two stories. At the same time, room
    occupancy, and plot densities are very high. Many of the old family houses are overcrowded, with an
    average of 4 persons to a room among the poor families. We met several with 15-20 persons sharing a
    room: the room is used for storing the goods; the people sleep outside. In Jamestown, for example, there
    was a household with 16 family members, living in one room of a family house with 8 rooms, with 7 other
    households (a total of 40 residents, of which 16 share 1 room). In several cases, particularly among the
    poorer families, families split up in different houses due to lack of space. Husbands live separately from
    their wives, the children split up, or the children are handed over to the in-laws or grandparents to look
    after, because of lack of space.

-   Most household heads run their own businesses, e.g. small scale food vending, water sale, and so on.
    But most have no access to loans to expand their businesses. As a result, profits continue to be low, and
    the poorest often do not have capital to take advantage of savings from bulk purchases etc. They express
    interest in business loans, but not in a housing loan. They say they will build the houses from the money
    they make from the business.

                                                                                                                      African Union for Housing Finance
-   Remittance income is not very substantial, but does constitute about 10 percent of the income of the
    poorest families. Rental income, if any, in the family houses does not really benefit the residents, because it
    goes to those owners who are currently not occupying the property (unlike other rental housing income in
    places like Nima).

-   Despite all of the inherent problems with family house, culture of cross-family support lends itself as a
    social safety mechanism. The poorest in the family at least have a roof on their heads, however modest
    and/ or inadequate, and do not have to incur additional expenses on housing. Ownership disputes are not
    as common as one would expect.

2. Landlords and Renters
-   Renters often pay an advance of up to 6 years of rent. Three years is more common for residential
    property, 6 years for commercial property.

-   Rent paid in advance affects people’s mobility. It also affects the housing market negatively by potentially
    locking up the stock for an extended period of time, and in turn, in increasing the rents (due to limited
    supply at any given time).

-   Tenants claim to have very limited rights with respect to their tenancy contracts. Evictions are common,
    and the tenant often has no real recourse, especially since the courts are already lagging far behind in

    terms of property dispute cases.

                                                                                                                                                          General Findings
-   In neighbourhoods like Nima, most original residents have moved out, and converted their old family
    houses into rental housing. These are essentially absentee landlords, and do not provide toilets or water
    as part of the accommodation. The tenants possibly take it simply due to lack of a better or affordable
    alternative. Physical conditions are similar to the old family houses discussed above; in some cases, the
    tenants undertake basic improvements with their own funds.

-   For the landlords, this is a low-risk, high return investment. For the tenants, it is the only affordable
    housing available.

-   Renters are typically either single (unmarried single men or young women), or recent migrants (single or
    married), which makes them economically better off in terms of per capita income. That is also another
    reason they are able to afford rent, which constitutes 10-14 percent of the total expenditure. Squatters,
    on the other hand, also with similar household characteristics, typically have lower incomes than renters.

3. Sanitation Facilities
-   Lack of toilet facilities are the result of a host of factors: poverty, complications emanating from multiple
    ownership of family houses, absentee landlords who are not interested in making any improvements to the

                                                                                                                      African Union for Housing Finance
    rental housing, and possibly the most critical demand-related issue, which might suggest that people simply
    do not prioritise a private toilet.

-   Public toilets are very expensive: the price can vary between 400-1000 cedis per use. Hence, people
    use toilets only when they really have to. The consequence: children defecating in the beaches or the bush,
    or even in the open, adults using any available corner. It is, hence, of little wonder that every wall in Accra
    more than 3 mts tall bears a warning against urinating

-   Expenditure on toilets is directly proportional to the number of household members, and particularly high for
    larger families which are also poorer. Five to ten percent of the total household expenditure goes to
    public toilets. In comparison, the economically better off families who have toilets in their plots typically
    spend in the range of 2 percent of total expenditure on sanitation.

4. HFC Micro-lending program
-   Market women who signed up for the HFC savings and micro-lending program had significantly higher
    incomes, greater savings, and much better living conditions. They had toilets, shared in some cases
    (not public toilets, as was the case with the vast majority of the others), but mostly individual. And most had
    built new houses after signing up with the savings scheme: they had taken loans to expand their
    businesses, and using the profit from the larger business (and larger savings, because of larger capital that
    allows them to buy stuff cheaper in bulk, for example), were able to build their own houses.

                                                                                                                                                          General Findings
                                                                                                                      African Union for Housing Finance
III. Analysis based on Tenure, Legality
and Income
      Household Size
The average size of households in informal settlements was found to be 6, which in comparison with other
African countries is not unusually large. However, the characteristics of these households—with a large
number of single parents, dependents, and unemployed persons—makes them more vulnerable. This is
particularly the case with poorer households, with per capita income (pci) below 325 cedis1, where we found
92 percent of the households headed by women, and 83 percent of the households with at least 1 single parent.
Also the average household size is 8, bigger than the overall average, due to more extended family members
(typically dependents).

                                                                                                         Average % HHs
                             # immediate family   # extended family   Total # members in

                                                                                                                          parents in HHs
                                                                                                                          Avg # of single
                                                                                                         with at east 1
                                 members              members                 HH

                                                                                           % of women-

                                                                                                         single parent


                                                                                                                          with single


                                                                                                                                                                                General Findings
                             Average    Median    Average   Median    Average    Median

    Renters                     2         1         0          0         2         1          20             0               0
    Owners                      4         3         3          2         7         6          63            58               2

    Squatters                   4         3         1          0         5         4          60            40               2
    Legal/ de facto tenure      3         3         4          2         7         6          50            50               2

    < =325K pci                 4         3         5          3         8         7          92            83               2
    >325K pci                   3         3         0          0         3         3          17             8               2

    Overall                     3         3         3          1         6         5          54            46               2

 325K cedis is the median per capita income per month recorded from our sample of 24 households across 5 settlements in

                                                                                                                                            African Union for Housing Finance
The following section presents the numbers by disaggregating the data in three ways: by tenure, legal status,                Box 1. Margaret Amoah: Jamestown
and income levels.

Renters versus Owners
Renters typically seem to have smaller families. The average household size in our sample was 2, the median
was 1.0. This may be attributed to the fact that many of them are young people, still single, who have recently
moved out of their parents homes and established their own businesses. Others are recent migrants who have
come into the city with their—relatively young and small—families.

Owners, on the other hand, mostly comprise families that have lived in these settlements for generations, and
have expanded in size over generations. The households are typically large, with the average recorded at 7
persons per household (with 4 members in the core/ immediate family, and 3 dependents or extended family

All of the “owner” households interviewed in Nima and Jamestown, for example, had extended family members
(nephews/nieces, parents, or siblings). The average number of extended family members was 5.8 (range 2-14;
median of 3), and the average household size was 8.9 (range 3-16; median 7). See boxes 1 and 2 for illustrative

Squatters versus Legal/ de facto Tenure Occupants
Like the renters, most of the households in squatter settlements (Sodom and Gomorrah, and the railroad

settlement in our sample) are single families, and hence have a small household size, the average being 5.

                                                                                                                                                                                              General Findings
The average household size recorded in the “legal/ de facto tenure” category was 7. Since this includes both
owners and renters, and given that the renters are substantially smaller households than owners, it does
indicate that there are many owner households in the so-called “legal” settlements that are substantially larger.
Examples shown in boxes 1 and 2 substantiate this.
                                                                                                                      Female-headed household
                                                                                                                      4 single mothers, 2 mothers married but living
Households below versus those above median per capita income of 325K cedis/month                                      separately
Poorer households are clearly larger in size. Households with per capita income (pci) less than 325K                  Household income: 1100K cedis/ month
cedis/month have an average household size of 8, nearly three times as large as those with per capita income
greater than 325K.                                                                                                  Margaret Amoah lives in Jamestown. Her household comprises her
                                                                                                                    mother and two children, but she shares her accommodation—one
                                                                                                                    room in a family house—with her 4 sisters and their children, totalling
Single parents (with children) are also clearly more prevalent in poor households: 83 percent of the households     16 persons. Margaret is married, but her husband lives in his own
with per capita income less than 325K had at least one single parent (an average of 2 single parents per            family house in Jamestown for lack of space.
household, mostly mothers). Ninety two percent of the “poor” households were headed by women. In contrast,
                                                                                                                    Only 3 members of this 16-member household are males, and all of
we saw single parents in only 8 percent of the households with per capita income higher than 325K. In               the 7 other households on this plot—Margaret’s cousins and their
Jamestown, in particular, every household interviewed had single mothers, ranging from 1 to 5 in number (with       families—are women-headed. In all, on the plot, there is a total of 9
an average of 2.5), and all of these were women-headed households. Once again, this is illustrated by the           males (among whom only 2 are older than 18 years), and 31
examples in Box 1 and Box 2.

                                                                                                                                African Union for Housing Finance
    Room Occupancy and Plot Density                                                                                                               Box 2. Manti Ayikai: Jamestown
Plot density is substantially higher in legal settlements where people have titles or de facto tenure, with an
average of 7 households and 24 residents per plot. The same applies to poorer households, where the average
number of persons per plot was recorded at 18. Poorer households also have the highest room occupancy,
with an average of 4 persons to a room, double that of the others.

                                          # persons per room           # HHs per plot          # residents per plot
                                         Average       Median      Average       Median       Average      Median
              Renters                        2            2            8            9            12           10
              Owners                         3            3            3            1            15            8

              Squatters                      3            3            1            1            4             4
              Legal/ de facto tenure         3            2            7            6            24           23

              < =325K pci                    4            4            4            3            18           11
              >325K pci                      2            2            5            5            12           10

              Overall                        3            3            4            4            15           11

The following section presents the numbers by disaggregating the data by tenure, legal status, and income
                                                                                                                                        Female headed household

                                                                                                                                                                                                                General Findings
                                                                                                                                        2 single mothers
                                                                                                                                        Household income: 750K cedis / month
Renters versus owners
Most of the renters interviewed were in the “legal/ de facto tenure” settlements (i.e. not in the squatter                            Manti (right, in the photo) is a widow who lives with her 85-year old
settlements), occupying rooms in plots with absentee landlords, and 10-20 rooms per plot. As discussed earlier,                       mother, Rebecca, and 14 children: 7 of her own, and 7 from her
                                                                                                                                      deceased sister. The husband of the deceased sister is a fisherman
they are mostly small families, with an average of 2 persons per room. As a result, although they recorded the                        and lives in Jamestown; he provides occasional child support for his
highest number of households per plot (an average of 8, and a median of 9), their plot density is a mere 12                           7 children. With so many members in the household, and only 2
persons per plot, lower than the overall average of 15.                                                                               rooms, several of the children go to the Chief’s palace next doors to
                                                                                                                                      sleep at night. The plot is shared with 4 other families.
The owner category in the table includes residents of squatter settlements, which are typically small nuclear                         Manti runs a small business with her mother: they buy pigs in the
families living in temporary houses in undefined lots, thus brining the overall average down. The next sections                       rural areas, bring them to the city for slaughter, and cook the meat
presents a more realistic picture of the densities in some of these legal settlements with de facto tenure.                           and sell it locally. They do not have space for a refrigerator, or the
                                                                                                                                      money, so they typically store the extra meat in a neighbour’s fridge
                                                                                                                                      at 5K cedis per day. From this business, they earn some 300K per
Squatters versus “Legal/ de facto tenure” occupants                                                                                   month. In addition, they sell water from their private standpipe, which
                                                                                                                                      fetches them some 150K per month in profit.

  The plot density for squatters is largely irrelevant because there is no real defined plot. The walls of the structures—typically
temporary houses or kiosks—define the plot.

                                                                                                                                                  African Union for Housing Finance
The average number of persons per plot in “legal” neighborhoods was recorded at 24. This is attributed largely
to neighborhoods like Jamestown (with a large number of “owner” families) and Nima (with a large number of
renter families). The average number of households per plot ranges from 1 to 13, and the number of residents
on the plot from 8 to 40; in both cases, the poorer families are the one that record the larger numbers.

Below and above median per capita income of 325K cedis/month
As one would expect, the plot density and room occupancy is much higher among poorer families: the plot
density for those households with per capita income less than 325K per month is 18 persons per plot, while that
for the others is 12. Similarly, room occupancy is 4 persons to a room for the former, and 2.3 for the latter. Note
that this is despite the relatively smaller average for the number of households on the plot, simply because they
have larger households.

In several cases, particularly among the poorer families, in Jamestown and other similar neighbourhoods,
families split up in different houses due to lack of space. Husbands live separately from their wives, the children
split up, or are handed over to the in-laws or grandparents to look after, because of lack of space. And this is
almost always related to poverty. In Jamestown, there was a household with 16 family members, living in one
room of a family house with 8 rooms, with 7 other households (a total of 40 residents, of which 16 share 1

       Income and Poverty
Given the variations in household characteristics in these settlements, it is more appropriate to use income per

                                                                                                                                                          General Findings
capita—rather than household income—to rate vulnerability. The overall median household monthly income
for the sample is 1450K cedis, and the median per capita monthly income (calculated by dividing HH income
by HH size) is 325K cedis.

Our interviews indicate that the “owner” families are the worst off, with a median per capita income of 300K
cedis/month. Renters, in contrast, are the best off with a per capita income of 505K cedis/month. Squatters fall
in the middle ground between owners and renters: the disadvantage of their lower income is slightly offset by
the fact that they pay nothing for housing (which is also the case with owners), but not enough to compensate
for the family size, which even though much smaller than owner families, is not as small as renter families.3

The following section presents the numbers by disaggregating the data by tenure, legal status, and income

    The average HH size was observed to be 1 for renters, 4 for squatters, and 7 for owners.

                                                                                                                      African Union for Housing Finance
                                            HH income per month         Per capita income per
                                                 (‘000 cedis)           month (‘000 cedis)
                                          Average          Median         Average        Median
                Renters                     1210            1300            835            505
                Owners                      2022            1500            361            300

                Squatters                   2220            1400            519            300                          Box 3. Steven Amanfo: Sodom and Gomorrah
                Legal/ de facto tenure      1678            1450            477            325

                < =325K pci                 1369            1425            207            247
                >325K pci                   2438            1675            782            579

                Overall                     1904            1450            494            325

Renters versus Owners
Although the median monthly household income for renters was found to be lower than that for owners—1300K
versus 1500K—their per capita income is actually substantially higher, with a median of 505K versus 300K
cedis per month. The average we recorded was still higher, at 835K cedis. This is largely due to the relatively
small household size and other household characteristics discussed earlier.

Owner families are diametrically opposite: relatively higher median household income, bus substantially lower
per capita income (300K cedis per month) due to the larger household size.

                                                                                                                                                                                             General Findings
                                                                                                                        Male-headed household
Squatters versus “Legal/ de facto tenure” occupants                                                                     Household income: 9600K cedis / month
Squatters and renters present similar household characteristics and residential mobility, but are significantly
different in terms of income levels, the squatters being much poorer with a median per capita income of 300K          Steven operates a toilet/ shower business, and lives with his wife
cedis per month. This could make for the argument that those can afford to get “legal” housing do so, the others      and 6 children in Sodom and Gomorrah. His income from the toilet
                                                                                                                      business is substantial: 250-300K cedis per day. His wife is a
simply squat. It is a rational economic choice for multiple reasons: one, housing is “free”; two, it may be poor in   vendor, who sells drinking water sachets; from this, she makes on
quality but is not much worse off than the living conditions in most rental housing in “legal” settlements where      average 20K per day. Four of his children attend school: one is in a
they would have to pay; and three, it is close to the workplace (most squatters operate their businesses from         technical school training in construction; the youngest three attend
their homes).                                                                                                         private school. Although not an elected leader per se, he commands
                                                                                                                      respects in the settlement because of his “entrepreneurship”, he
The relatively high average per capita income for squatters is explained by a few individuals who have “made
it,” as they say, as is the case with Steven, a toilet operator in Sodom and Gomorrah, with an income of about
9600K cedis per month (See Box 3).

                                                                                                                                 African Union for Housing Finance
Below and above median per capita income of 325K cedis/month
Based on the numbers, our sample may be split into the following per capita income quartiles:
        1st Quartile             x > 579K
        2nd Quartile             325K< x <579K
        3rd Quartile             247K< x < 325K
        4 Quartile               x < 247K

    Income Sources
Most households in informal settlements are actively engaged in at least one informal sector business activity.
Two thirds of the households in our sample got all of their income from business activity.

Remittance is directly related to the extended family structure: those who are functioning as nuclear families
rarely receive any remittance or family assistance. This includes most of the squatters and the renters. Poorer
families are more reliant on remittance and family assistance, which constitutes a larger share of their
household income.

                                                                                  Average %
                                               Average %
                                                                  Average %      income from

                                            income from self-
                                                                 income from      remittance/
                                                                wages (formal)      family

                                                                                                                                                        General Findings
                   Renters                         69                14                0
                   Owners                          81                 9              10

                   Squatters                      100                 0               0
                   Legal/ de facto tenure          68                18              14

                   < =325K pci                     78                 8              14
                   >325K pci                       86                12               2

                   Overall                         82                10               8

Two thirds of the households got all of their income from business activity. The squatter households, in
particular, were all self-employed, deriving all of their income from informal sector businesses. Only a fifth of
those living in “legal” settlements worked in the formal sector.

                                                                                                                    African Union for Housing Finance
Renters versus Owners
None of the renters interviewed reported any remittance income. On the other hand, owners, in the legal
settlements only (see below), reported a median remittance equal to 10 percent of household income.                               Box 4. Susu savings accounts

Squatters versus “Legal/ de facto tenure” Occupants
Fifty percent of the households in Nima, Jamestown, and Mamponsa (the three “legal/ de facto tenure”
settlements) reported remittance income. None of the interviewees in the 2 other settlements reported any
remittance. The average remittance as a percentage of total household income for “legal” residents was found
to be 14 percent.

Below and above median per capita income of 325K cedis/month
For the poorer households, average remittance was 14 percent of total income, compared with 2 percent for the
others. This higher ratio may be attributed to the low basic income of the poorer families to begin with, but also
goes to show that they are much more dependent on family assistance.

                                                                                                                     William Paakwesi , Mamponsa      Marianna Tetteh, Jamestown:
                                                                                                                     Water vendor/ public shower              Food vendor
                                                                                                                                operator               Saves 5K cedis daily in susu
    Savings                                                                                                           Saves 20K cedis daily in susu        (100K cedis/ month)
In general, more of the economically better off households (with per capita income higher than 325K cedis per              (600K cedis/ month)

month) have savings compared to the poorer households; they also save more, and save in formal sector
schemes. Poorer households, on the other hand, have smaller savings, and tend to save in informal sector
schemes (mostly susu collectors).

                                                                                                                                                                                       General Findings
74 percent of the households interviewed had savings in one form or another. 44 percent of those who save do
so with susu collectors, 38 percent in formal sector accounts, 13 percent in both, and 6 percent at home.
Median savings amount to about 18 percent of the household income, which translates into 175 cedis per

There is not a significant variation in the savings between renters and owners, and between squatters and
“legal” residents. There is, however, a clear difference in the savings patterns between those below and above
the per capita monthly income of 325K cedis.

                                               Savings as % of income       Median HH         Median per capita        Manti Ayikai, Jamestown:       Akua Kobi, Railroad squatter:
                          Median savings/
                                                                           monthly income      income/ month            Food and water vendor                 Food vendor
                         month (‘000 cedis)   Average         Median
                                                                            (‘000 cedis)         (‘000 cedis)          Saves 5K cedis daily in susu    Saves 20K cedis daily in susu
    Renters                     150              22              16            1300                  505                  (150K cedis/ month)              (400K cedis/ month)

    Owners                      300              15              18             1500                300

    Squatters                   300              21              19             1400                300

                                                                                                                                 African Union for Housing Finance
    Legal/ de facto tenure         150             15            13              1450               325                   Box 5. Susu and formal savings accounts

    < =325K pci                    175             20            20              1425               247
    >325K pci                      300             15            18              1675               579

    Overall                        250             17            18              1450               325

Below and above median per capita income of 325K cedis/month
Fifty eight percent of those with per capita income less than 325K cedis per month save on a regular basis.
Among these families who save, the 57 percent save only in susu accounts, 14 percent save only in formal
sector accounts (HFC, Sinapi, or a bank), and 29 percent save in both. The median saving as a percentage of
income is 20 percent, or 175K.

Seventy five percent of the households with per capita income above 325K cedis per month save money on a
regular business. Fifty six percent of these households save in formal sector accounts, 33 percent in susu
accounts, and 11 percent at home. The median saving as a percentage of income is 18 percent, or 300K.
                                                                                                                     Top: Obayaa Aseiduaa is a petty trader who lives in the railroad
                                                         % HHs in      % HHs in both     % HHs at                    squatter settlement. She saves 20K cedis daily in susu (600K cedis/
                                           % HHs in
                HHs with savings                          formal         susu and         home                       month) which she deposits into her formal bank account at the end
                                         susu accounts
                                                         schemes          formal                                     of the month

                < =325K pci                   57            14              29               0
                                                                                                                     Below: Dorothy Mensah is a cloth merchant in Makola Market. She
                >325K pci                     33            56               0              11                       is a member of the HFC savings scheme, and also has one susu
                                                                                                                     account, where she contributes 10K cedis per day. She uses the

                                                                                                                                                                                            General Findings
                                                                                                                     money in her susu account to pay the utility bills for the shop, and
Susu Savings Scheme                                                                                                  any other business-related expenses. The susu essentially works
The vast majority of those earning less than the median income save in susu accounts. Susu collectors are            like a current account, and HFC as a savings account.
private groups that organize agents to collect individual savings on a daily (or weekly) basis. A service fee
usually equivalent to the average contribution of a day is levied. At the end of each month, the money is
returned to the individuals after deducting the fee. In simple terms, the susu scheme works like a safety deposit
box, with membership renewable every month.

Among our interviewees, most people who had such accounts contributed anywhere between 2K and 200K
cedis per day (typically 5 days a week, but could be more or less depending on the arrangement). They “put the
money away”, they say, mostly from their daily earnings, “otherwise it gets spent.” Typically, people use these
savings at the end of the month, to buy stock for their businesses or bulk supplies for the house. So essentially,
it serves as a short-term safety deposit box. To cite some examples, a woman in Makola market has an account
with the HFC as well as a susu collector. She uses the susu for he business expenses at the end of the month,
whereas the HFC account is purely a savings account where she deposits her profits. Another woman in
railroad settlement, deposits her savings into the susu scheme, and at the end of the month, takes it to
Metropolitan Bank (see Box 5). In her case, the susu account works as a safety deposit box where she stores
her savings till she can take them to her bank.

                                                                                                                                African Union for Housing Finance
Several interviewees expressed suspicion and scepticism with regard to susu schemes. Several complained of                             Box 6. HFC Micro-lending
being cheated, where “the collector ran off with the money”, or having heard of others being duped.

Formal Sector Savings and Loans Schemes
There are many such schemes in Ghana, although telling from the case of Accra, there is still a lot of untapped
potential in terms of clients for these schemes. We met a few people in Jamestown and the railroad settlement
who had enrolled in the Sinapi program, for instance, and got assess to quick loans. Some people in Sodom
and Gomorrah were aware of the savings scheme offered by the Homeless People’s Federation, and so on.
However, most of the people were unaware of any such programs, and when we discussed these with them,
they listened and expressed keen interest.

HFC Savings and Micro-lending Program
The HFC program for market women is a clear example of how savings schemes can really work for the poor. It
is less the concept, however, and more the design of the specific program that makes it tick. In our interviews
with market women, it was clear that many of them had started off in conditions comparable or far worse than          Top: Gladis Appiah started out as a hawker in Kanishi selling
some of our interviewees in the informal settlements, and today have substantial savings, and have been able          tomatoes, walking around with a basket on her head. From her
to incrementally expand their businesses, and even build homes. More details on programs offered by Sinapi,           savings over the years, she has expanded her business and now
                                                                                                                      trades in provisions (cans, spices, etc.) on a table-top in
HFC, and ProCredit are discussed in Annex A. The interviewees range across the income spectrum – from a               Agbogbloshie market. Gladis opened an account with HFC in 2003.
vendor with a stall who was previously a tomato seller selling tomatoes on a basket on her head and now owns          Her first deposit was 5K cedis. She now contributes 15-30K cedis
a table-top stall, to a slipper wholesaler who started out with a table-top stall and now rents a real shop to        daily, and has got a balance of 1.4 million cedis. She estimates that

conduct business, to a cloth merchant who started with a small stall and now owns a cloth business worth some         her stock today is worth some 5-7 million cedis.
1 billion cedis (see Box 6).                                                                                          Below: Ophelia Wilson started out her business in a stall on the
                                                                                                                      side-walk, with a monthly revenue of some 1-1.2 million. Through the

                                                                                                                                                                                               General Findings
The table below compares the economic situation of these market women to the other interviewees of informal           HFC program, she was able to save, and reinvest into her business.
                                                                                                                      She bought this shop in Makola market, and increased her stock
settlements. Their per capita income too is nearly 3 times that of the other interviewees from the informal           over the years. Today, her shop’s net worth is about 1 billion cedis
settlements. Their savings, in real terms and as a percentage of income, are substantially higher than the typical    (300 million for the shop, and 700 million for the goods). Her current
informal settlement dweller. The median current savings balance recorded was 13.6 million cedis and the               balance in the savings account is 45 million cedis.
average 27.6 million cedis.

                                                Savings as % of income       Median HH        Median per capita
                          Median savings/
                                                                            monthly income     income/ month
                         month (‘000 cedis)    Average         Median
                                                                             (‘000 cedis)        (‘000 cedis)
    HFC market women
                               2500              45               45            6250                1000
    Other interviewees
                                175              18               17            1450                325

Interestingly, most of the women interviewed had toilets in their plots if not within the house. Most had built new
houses after signing up with the savings scheme. Most had taken loand to expand their businesses, and using
the profit from the larger business (and larger savings, because of larger capital that allows them to buy stuff
cheaper in bulk, for example), were able to build their own houses.

                                                                                                                                 African Union for Housing Finance
    Cost of Housing and Services
Nearly a quarter of the monthly expenditure of households in informal settlements is on housing and services.
This is despite the fact that most do not spend anything on housing. The excessive expenditure on services is
mainly attributed to lack of water and sanitation facilities in the vast majority of these houses.

The table below shows household expenditure on housing and basic services (water, sanitation, electricity).

                                      % HH expenditure on              % HH               Total % HH
                                        services (water,           expenditure on        expenditure on
                                      sanitation, electricity)        housing         housing and services
                                      Average        Median      Average     Median   Average      Median
            Renters                      17            17          11          12       27           27
                                                                                                                               Box 7. Public toilets as a business
            Owners                       18            19           3          0        20          19

            Squatters                    20            19           2          0        22          19
            Legal / de facto tenure      17            17           7          0        24          21

            < =325K pci                  18            19           5          0        23          22
            >325K pci                    17            17           5          0        22          19

            Overall                      18            17           4          0        22          21

                                                                                                                                                                                           General Findings
There is a general consistency in the spending patterns, in that they all spend an average of 17-20 percent on
services. The only stark difference is in the housing expenditure for renters. As expected, they pay rent, which
averages about 11 percent of monthly household expenditure, thus making their total expenditure on housing
and services about 27 percent.

The high cost of services is attributed largely to the extremely poor, and often non-existent, systems for           Our interviewee Steven Amoah started his public toilets business a
sanitation and water supply. These are discussed below.                                                              few years ago, and today is a rich man in his neighbourhood, Sodom
                                                                                                                     and Gomorrah. His income is 9600K cedis per month. In our sample
                                                                                                                     across 5 settlements, the next highest income is less than a fourth
Sanitation                                                                                                           of Steven’s.
Whether in the squatter settlements or the so-called “legal” settlements, there are hardly any private toilets or
even shared toilets within the residential plots. Even in neighbourhoods like Jamestown, where the trunk
infrastructure for sewage lines already exists, there are very few private toilets. Only 20 percent of our entire
sample of interviewees had any form of toilet facility on the plot, mostly shared with several other households.
Most people use public toilets, either provided by the AMA, or built privately by local businessmen. Incidentally,
the toilet industry is a real money-maker (see Box 7), and several entrepreneurs who have the resources for a
water connection and the ability to get the necessary approvals, have built public showers and toilets.

                                                                                                                                African Union for Housing Finance
Apart from the issues of poor hygiene, and limited supply of toilets, they are expensive. The table below shows                    Box 8. Lack of sanitation facilities
the expenses incurred by households on toilet facilities. These numbers are conservative estimates, assuming
at most 1 use per person per day, unless specified otherwise by the individuals.

                                                                                      % HH expenditure on
                                         Cost of toilet (per    % HH expenditure on
                                                                                       water, sanitation,
                                          use) in cedis               toilets

                                        Average      Median     Average    Median     Average     Median

           Renters                       358           325         8          4         17          17
           Owners                        647           700         9          7         18          19

           Squatters                     720           500         11         8         20          19
           Legal/ de facto tenure        518           600         7          7         17          17
                                                                                                                        Top: The beach front in Mamponsa, used for dumping garbage and
           < =325K pci                   575           600         8          7         18          19                  defecating.
           >325K pci                     629           600         5          4         17          17                  Below: A wall in Jamestown, with graffiti: “Don’t urinate. Pig urinate
                                                                                                                        here. Dog urinate here.”.
           Overall                       602           600         6          5         18          18

Depending on the availability and the locality, the price can vary between 400-1000 cedis per use. This is a
costly affair by any standards: where people barely have a dollar a day at their disposal, it is difficult to imagine

                                                                                                                                                                                                 General Findings
how they can part with 5-10 cents each time they use the facilities. As a result, people use toilets only when
they really have to. The consequence: children defecating in the beaches or the bush, or even in the open,
adults using any available corner. It is, hence, of little wonder that every wall in Accra more than 3 mts tall bears
a warning against urinating (see Box 8).

Households with toilets. In contrast, those house owners who do have the facilities within the compound spend
significantly less (see table below): an average of 2 percent on toilets, a third of the overall average.

                               % HH expenditure        % HH expenditure     % HH expenditure    Total expenditure on
                                  on toilets             on services           on housing       housing and services
                              Average     Median      Average    Median    Average    Median    Average     Median
 Owners with toilets                2        3            6         6         0         0          6           6
 Overall                            6        5           18        18         5         0         23          21

                                                                                                                                    African Union for Housing Finance
Water Supply
As in the case of toilets, the problem of lack of water supply is consistent across all categories. Only about30
percent of our interviewees had a water connection on the plot. Regardless of whether the person is a squatter
or a “legal” occupier of the property, an owner or a renter, someone extremely poor or relatively well off, most
households do not have water connections in their homes. On average, they spend about 9 percent of their
household income on water, which is about half of their total service cost (excluding housing).

                                                                                                   % HH expenditure
                                               Cost of water (per    % HH expenditure on
                                                                                                  on water, sanitation,
                                                bucket) in cedis           water
                                              Average      Median    Average      Median          Average      Median
          Renters                               383         400         5           4                17            17
          Owners                               364           300           6            6             18         19                              Box 9. Wastewater disposal

          Squatters                            386           350           7            4             20         19
          Legal/ de facto tenure               384           300           6            5             17         17

          < =325K pci                          363           350           8            8             18         19
          >325K pci                            427           500        10              5             17         17

          Overall sample                       385           300           9            7             18         18

Households with water connections. A situation similar to the toilets exists with water supply. Those who have

                                                                                                                                                                                                         General Findings
piped water pay at the rate of 6.9 cedis/ liter, which in terms of buckets (25 liters) is about 175 cedis per bucket.
Those without water supply in their plots pay 2 to 2.5 times that. This, like toilets, contributes in large part to the
increased cost of services: households with piped water spend on average some 14 percent of their income on
housing and services, as compared to those without who spend nearly double that (27 percent of their income).
To some extent, this may be attributed to the fact that most people with piped water sell it to others as well; the
profit from the water sales more than offsets the cost of the water they use themselves. Even so, their base cost
of water is far lower than their selling price: 175 cedis per bucket versus 400 cedis (see table below).                           Open disposal of solid and liquid waste in Sodom and Gomorrah:
                                                                                                                                   note the residential structure built just over the open sewage, and
                                                                                                                                   the narrow foot-bridge (on the right of the picture).
                          HH expenditure              % HH              % HH                    % HH             % HH exp. on
                           on water per           expenditure on    expenditure on          expenditure on       housing and
                          bucket (in cedis)           water            services                housing             services
                         Average    Median       Average   Median   Average    Median       Average   Median   Average    Median
 HHs with piped
                           175        175            2        2       12        11            2            0      14       12
 water on plot
 HHs without piped
                           425        400            11       8       21        19            6            1      27       25
 water on plot
 Overall sample            385        300            9        7       18        18            5            0      23       21

                                                                                                                                              African Union for Housing Finance
There are also many areas where the water supply is poor in general, so it does not really help having a piped                       Box 10. Electricity connections
water connection. For example, in Nima we talked to an owner of a family house who has piped water, but says
he the water does not come for months at a time. They have to go to other neighbouring settlements by taxi to
fetch water, which adds to the already exorbitant cost. Another example is Andreas, a renter in Kotobiabi: he
spends 140K cedis per month on transporting water, nearly equal to the cost of the water itself, which is150K
cedis. In other words, he spends 21 percent of his 1400K cedis monthly income on water and transportation of

Most of the households we interviewed had electricity, all barring 4 (out of 24). This is interesting in light of the
fact that most of these households do not have water or toilets, but manage to get an electricity connection.
Very few are responsible for paying the bill to the electricity company, of course; they pay a flat amount by
usage or “points” (derived from number of sockets or gadgets being used) to the registered “consumer”. In other
words, much of the electricity is bought and sold illegally, without the necessary permits. This is particularly the
case in squatter settlements like Sodom and Gomorrah, where practically every unit has power, even the small
wooden kiosks (see Box 10).

However, the fact that so many people have electricity sheds light on the fact that there is something more than
“poverty” per se that hinders people from getting access to the other services like water and sanitation, possibly
high connection / capital costs etc, and/ or regulations pertaining to water and sanitation which have associated

health and environmental implications. A third possibility could be the high demand for electricity on part of the
consumers, and relative ease of supply and high profitability on part of the seemingly aggressive local

                                                                                                                                                                                                 General Findings
                                                                                                                        Vida Awarika is a 28-year old single mother of two young girls living
                                                                                                                        in a temporary wooden kiosk in Sodom and Gomorrah. She works
                                                                                                                        as a seamstress, with an income of 800K cedis per month.

                                                                                                                        Vda has no piped water or toilet, but does have an electricity
                                                                                                                        connection. When we asked her where they pay their bills, she
                                                                                                                        responded: “We don’t know. Someone comes and collects the
                                                                                                                        money. I don’t even know where the meter is. There are so many
                                                                                                                        wires that it is difficult to trace the path.” In other words, someone
                                                                                                                        has given her a connection, charges her exorbitantly for very little
                                                                                                                        consumption, and makes a fortune.

                                                                                                                                    African Union for Housing Finance
IV. Interviews
This section presents the interviews in detail in the following order:

1.   Jamestown (mostly traditional family houses, with “owner” families)
2.   Nima (mostly renters in family houses with absentee landlords)
3.   Kotobiabi
4.   Mamponsa (beach side settlement)
5.   Sodom and Gomorrah (squatter settlement near the lagoon, slated for resettlement)
6.   Railroad settlement (squatter settlement)
7.   HFC market women (from Makola Market and Agbogbloshie Market)

                                                                                                                             General Findings
                                                                                         African Union for Housing Finance

To top