Content uploaded by Charlotte Wrigley-Asante
Author content
All content in this area was uploaded by Charlotte Wrigley-Asante on Sep 28, 2022
Content may be subject to copyright.
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=rafg20
Download by: [University of Ghana] Date: 29 June 2017, At: 08:06
African Geographical Review
ISSN: 1937-6812 (Print) 2163-2642 (Online) Journal homepage: http://www.tandfonline.com/loi/rafg20
Gender dimensions of climate change adaptation
practices: the experiences of smallholder crop
farmers in the transition zone of Ghana
Charlotte Wrigley-Asante, Kwadwo Owusu, Irene S. Egyir & Tom Mboya
Owiyo
To cite this article: Charlotte Wrigley-Asante, Kwadwo Owusu, Irene S. Egyir & Tom Mboya
Owiyo (2017): Gender dimensions of climate change adaptation practices: the experiences of
smallholder crop farmers in the transition zone of Ghana, African Geographical Review, DOI:
10.1080/19376812.2017.1340168
To link to this article: http://dx.doi.org/10.1080/19376812.2017.1340168
Published online: 22 Jun 2017.
Submit your article to this journal
Article views: 5
View related articles
View Crossmark data
AFRICAN GEOGRAPHICAL REVIEW, 2017
https://doi.org/10.1080/19376812.2017.1340168
Gender dimensions of climate change adaptation practices: the
experiences of smallholder crop farmers in the transition
zone of Ghana
Charlotte Wrigley-Asantea, Kwadwo Owusua, Irene S. Egyirb and Tom Mboya Owiyoc
aDepartment of Geography and Resource Development, University of Ghana, Legon, Ghana; bDepartment of
Agricultural Economics and Agribusiness, University of Ghana, Legon, Ghana; cAfrican Development Bank, Immeuble
du Centre de commerce International d’Abidjan, Abidjan, Cote d’Ivoire
ABSTRACT
This paper discusses the gender dimensions of climate change adaptation
strategies among small holder crop farmers in the transition zone of Ghana.
A total of 612 farmers (328 females and 284 males) were interviewed
using purposive sampling technique. Our results indicate that adaptation
strategies are gendered with men mostly resorting to on-farm agronomic
practices such as the use of articial fertilizers and also moving into new
cash crops. Female farmers also use similar on-farm agronomic practices
particularly articial fertilizers to boost crop production but most importantly
resorted to petty trading in agricultural and consumable goods, an o-
farm strategy. This shows women’s resilience which has also subsequently
improved their decision-making role at the household level, which is an
indicator of empowerment. We recommend that institutions that support
climate change adaptation initiatives at the local level must take gender
dierences into consideration and support particularly women to strengthen
their resilience and consolidate their empowerment.
Introduction
Agriculture continues to play a major role in the socioeconomic and livelihood development in most
rural communities in Ghana (Institute of Social Statistical & Economic Research, [ISSER], 2014) but
the sector is increasingly being subjected to climatic changes. With Ghana’s agricultural sector largely
dependent on rainfall, the eect of climate change is seriously impacting on the livelihoods of small-
holder farmers. As a result, both male and female small-holder farmers are adapting to the vagaries
of climate change (Arku, 2013; Codjoe & Owusu, 2011). e Transition Zone of Ghana occupies an
ecotone between the tropical forest to the southwest and the semi-arid areas of the north. e zone
remains the nation’s breadbasket, making it crucial to the food security of Ghana. Yet the rainfall has
changed as a result of global climate change (Owusu & Waylen, 2012, 2013), and subsequently aected
food production. It has also been revealed that the short dry spell (July/August) and the minor rainy
season (September/October) have undergone signicant changes with negative implications for rain-
fed crop production. Whereas the rainfall regime during the major rainy season (March-June) is fairly
reliable for rain-fed crop production, the minor rainy season has been associated with high risks of
© 2017 The African Specialty Group of the American Association of Geographers
KEYWORDS
Gender; climate change;
adaptation; Ghana; women;
men
ARTICLE HISTORY
Received 28 September 2016
Accepted3 June 2017
CONTACT Charlotte Wrigley-Asante cwrigley-asante@ug.edu.gh Department of Geography and Resource Development,
University of Ghana, P.O. Box LG 59, Legon, Ghana.
2 C. WRIGLEYASANTE ET AL.
crop failure even though it has the most potential to reduce post-harvest losses (through sun drying
of grains) and provide households with food stock during the long lean season from December to
May (Owusu & Waylen, 2013).
e eects of these negative implications on small-holder farmers are gendered with some stud-
ies highlighting the vulnerability of women vis-a-vis men due to the challenges associated with
climate change (Codjoe, Atidoh, & Burkett, 2011; Yaro et al., 2010). As a result, men and women
have dierent ways and respond dierently to challenges associated with climatic hazards due to
social processes such as inheritance rules, land tenure systems and lack of support from formal
institutions (World Bank, 2009; Yaro et al., 2010). Some studies in Ghana have highlighted the
gendered dimension of climate change focusing mainly on perception and adaptation strategies
(Abubakari, Lawson, Mensah, Gordon, & Padgham, 2016; Arku, 2013; Codjoe et al., 2011). But
as argued by Arku (2013, p. 420) although research on local people’s adaptation to climate change
is increasing, more research is needed to enable the sharing of dierent coping strategies adopted
by farmers at dierent places. is study put in perspective the way small-holder farmers perceive
climate variability and change, the way their activities are vulnerable to climate change, and what
they are doing to decrease their vulnerability. Our paper contributes to the existing literature on
climate change by highlighting small-holder men and women farmers’ knowledge and experiences
on changes in climatic conditions as well as the dierent adaptation strategies, in the the Ejura-
Sekyedumase Municipality in the Ashanti region and the Wenchi Municipality in the Brong-Ahafo
region. Specically, the paper seeks to:
(1) Examine the gendered perceptions of climate change and its eect at the household level;
(2) Assess the dierent adaptation strategies of men and women and its implications on food
security at the household level;
(3) Discuss how adaptation strategies inuence the decision-making role of particularly females
at the household level.
e paper is divided into ve sections. Aer this introduction, the next section discusses the
conceptual issues in relation to gender and climate change. is is followed by the study area and
methodology. Subsequently, the ndings have been discussed and then the conclusion and policy
implications.
Gender and climate change vulnerability
e success of a society or individual’s eort to adapt to climate variability has been linked mainly to
its perception and awareness of the risks and hazards associated with its vulnerability (Adger, 2000).
Vulnerability in the context of climate change refers to ‘the degree, to which a system is susceptible to,
or unable to cope with adverse impacts’ of climate change, including climate variability and extremes
(Intergovernmental Panel on Climate Change, [IPCC], 2007, p. 781). Vulnerability to climatic risks
has been associated with inadequate nancial and institutional support (Beg et al., 2002), as well as
the availability of natural resources which many societies still rely on for rural livelihoods (Denton,
Sokona, & omas, 2000). In Ghana, the forest-savanna transition zone has the next highest social
vulnerability to climate change following the savanna zones (Yaro et al., 2010). It has been observed
that farmers are vulnerable to shocks resulting from extreme events such as ooding, seasonal var-
iation particularly the timing and amount of rainfall and long-term trends such as increased mean
temperature (Acquah & Onumah, 2011). Crop farmers thus express heightened concerns about erratic
rainfall patterns as these increase uncertainties about planting regimes and may induce diseases and
pests leading to mass crop failures (Yaro et al., 2010). In fact, Kemausuor, Dwamena, Bart-Plange,
and Kyei-Baour (2011) note that farmers’ vulnerability to climate risks in Ghana arise from abrupt
changes in season, droughts, reduced rainfall, increased temperature and oods, for both crops and
livestock. Kalame et al. (2011) have summed up the climatic risk to agriculture in the transition zone
AFRICAN GEOGRAPHICAL REVIEW 3
as droughts, erratic and late rains, temperature rise, windstorms, and bushres. Erratic and late rains
result in poor crop growth and poor tree seedling regeneration in some cases. Studies that examined
farmers’ perception of climate change suggest that most farmers perceive changes in various aspects
or elements of the weather over time (see for instance Codjoe et al., 2011). Similar ndings reported
by earlier studies include reduced ow in streams and rivers, prolonged rainfall shortages and drought
that led to crop failures, negatively impacting livelihoods, and shortage in water resources (see for
instance Gyampoh, Idinoba, & Amisah, 2008).
Within the African and specically the Ghanaian context, women and men’s dierential access
to resources is one of the key dimensions of gender inequality. Although women depend on natural
resources for their livelihood, they have very limited control over resources such as forest products
and land (African Development Fund – Human Resource Development Department, [ADF-OSHD],
2008). As a result, there are gender dimensions to climate change vulnerability in the sense that there
are dierences in the adaptive capacity to the eects of climate change and these dierences determine
how well men and women respond to climate change. ese dierences are due to social norms, tra-
ditional roles and dierent power structures and therefore more women than men are likely to lose
their livelihoods to climate change and extreme weather events due to their lack of resources (Jost
et al., 2016; Wright & Chandani, 2014).
Whilst smallholder farmers draw on ecosystem services in many instances to support their liveli-
hoods, women tend to have limited access to resources creating a situation of vulnerability to climate
change (Jost et al., 2016). ere are therefore gender dierences and preferences in agronomic practices
with respect to climate change (Arku, 2013; Codjoe et al., 2011). e conclusion oen drawn is that
women deserve to be targeted with climate change adaptation strategies and institutional support (see
for instance World Bank, 2009).
Again, because women are sometimes not part of the household and community decision-making
processes that aect their lives, they are oen excluded and underrepresented in decision-making and
policy processes regarding climate change (World Bank, 2009). Indeed, cultural biases prevent women
from participating actively in group training and extension meetings in certain communities and most
oen capital intensive technology does not necessarily benet women (CARE International, 2013).
Moreover, rural women farmers are reluctant to accept technological advances whose risks particularly
in terms of household food security are not known (Opare & Wrigley-Asante, 2008).
However, it is also argued that women are active agents who have developed locally adapted, appro-
priate and sustainable coping strategies and responses to climatic shocks. Indeed, their knowledge
can inuence adaptive mechanisms in most vulnerable areas (Egyir, Owusu, Jatoe, & Wrigley-Asante,
2014; Nhemachena & Hassan, 2007). is implies that women could pursue adaptive mechanisms
to limit vulnerability and strengthen livelihoods (World Bank, 2009). In view of this, the study will
explore the adaptation measures smallholder farmers in the transition zone of Ghana are using to
address their vulnerability and strengthen their livelihoods. e implications of this at the household
level in terms of decision-making will also be discussed.
The study area and methodology
is paper is part of a broader study conducted in two administrative municipalities that fall within
the Forest-Savannah Transition Zone of mid-Ghana (see Figure 1). ese are the Ejura-Sekyedumase
Municipality in the Ashanti region and the Wenchi Municipality in the Brong-Ahafo region. ese
two municipalities were selected based on preliminary interactions with the key stakeholders which
revealed that the two are major food crop producers in their respective regions. ese districts were
therefore purposively selected by the United States Agency for International Development (USAID)
funded project1 to implement the climate action for agricultural development programme. ree
communities were selected from each municipality, using purposive sampling technique and this was
also based on discussions with District Agricultural Extension ocials. In the Ejura-Sekyedumase
Municipality, Anyinaso, Babasso and Sekyedumase were selected to cover the geographical areas of the
4 C. WRIGLEYASANTE ET AL.
municipality; Anyinasu in the north, Babasso in the south and Sekyedumase in the middle portion of
the municipality. Again, Sekyedumase is considered the highest maize producing area whilst Babasso is
the least in the municipality. Awisa, Nkonsia and Akrobi, were selected from the Wenchi Municipality,
due to the fact that Nkonsia is closer to the Techiman market (one of the largest markets in Ghana);
Awisa is closer to Wenchi (the municipal capital) and Akrobi (which is closer to Awisa) was chosen due
to the presence of irrigation facility in the community. Awisa and Akrobi are geographically located
in the south of the municipality whilst Nkonsia is located in the north.
Information at households and institutional levels was gathered using desk review of literature,
individual surveys and focus group discussions (FGDs). Semi-structured questionnaires were used to
guide the interviewing of individual farmers and representatives of institutions. An interview guide
was developed for the FGDs. A total of 12 eld assistants and four technical team members were on the
eld. e eld assistants participated in community resource mapping by observation and individual
farmer survey while the technical team members participated in community resource mapping, FGDs
and key informant interviews.
Among the six communities, Sekyedumase has the highest population and Babasso and Anyinaso
have the lowest. e project agreed to have a total sample size of about 600. us because each com-
munity had more than a 1000 households, the sample size for each community was targeted at 80
or more. Proportionate sampling size was therefore considered in the sense that about 80 or more
respondents were targeted in each community. Each community was divided into four based on the
north-south-east-west frame to derive a quadrant (Abrahamson & Weaver, 2008). Every 10th house
in the quadrant was therefore visited so as to cover the quadrant. In other words, the project agreed
that each section and household in the community should have an equal chance of being sampled.
Individual farmers were selected from the households and only adults (above 40years) who own
farms and were major decision-makers of that farm were marked for selection. e 40years and above
criteria was used because the key objective was for farmers to be able to compare the weather changes
they have experienced in the last ve years with previous two decades. us upon entering a house,
Figure 1.Map showing the study sites. Source: Geographic Information Systems Lab, University of Ghana.
AFRICAN GEOGRAPHICAL REVIEW 5
the following questions were posed: ‘do you own a farm’? ‘If yes, how old are you’? Responses to these
two questions aided the selection of respondents. e study targeted individual farmers rather than
households. us in each household, a famer, who could be the head, spouse or ward (but should
be 40years and above) was targeted. But there were few instances where both husband and wife
were sampled. is was so because in certain cases women also had their own farms and were major
decision-makers of these farms and therefore qualied as respondents. Such women were considered
in the sample since gender sensitivity was a central theme of the study and therefore was to ensure
participation by women farmers.
Altogether, a total of 612 farmers were sampled comprising 328 female and 284 male farmers for
the survey (see Table 1). e bias toward women was not planned even though the project avoided
the situation where females would be excluded by targeting individual farmers rather than household
heads, who were likely to be men and major decision-makers (Opare & Wrigley-Asante, 2008). Another
possible reason for more females than males in the sample was as a result of male-outmigration in the
communities as a coping strategy to climatic changes. In such instances, females have been le-behind
to take care of agricultural work (Hughes, 2012; Wrigley-Asante & Agandin, 2015).
A total of 10 FGDs were carried out with an average of 14 persons in each FGD. Two were mixed
groups and the other eight involved four male only and four female only groups. First, the separate
sex groups were conducted and then the mixed groups were done in two communities (Babasso and
Akrobi) to validate some of the responses derived from the single sex separate groups. For the focus
group, any adult farmer (full-time or part-time) willing to participate in the discussion was welcomed.
e community representative (Assembly person) assisted in organizing the members of the focus
group. In addition, some institutions were also interviewed to solicit more views on institutional
support for small-holder farmers.
Key questions used in this paper included farmers’ perceptions of climatic changes, options small-
holder crop producers use to adapt to climate change impacts and the gender dynamics at the house-
hold level in terms of gender roles and relations. ese major themes guided the analyses of the
paper. Data was analyzed using SPSS version 20 to generate individual variables and cross tabulations
required. is was supported with qualitative quotes derived through the FGDs.
e general characteristics of respondents showed quite a signicant variation in personal as well as
farming background. e dominant groups were: female farmers (54%), those aged less than 61years
(80%), those without formal education (56%), the married (70%), Christians (81%), natives (67%),
full-time farmers (93%), farmers with experience of one decade or more (88%), commercially-oriented
farmers, selling more than one-half of harvested produce (69%), farmers cultivating less than 4 hectares
(80%), land owners of plots (54%) and solely dependent on farm income (55%).
Table 1.Communities visited and sample size.
Note: M = Male; F = Female.
Region District Community Sample size for survey
Ashanti Ejura Anyinasu M = 30
F = 51
Babasso M = 40
F = 41
Sekyedumase M = 47
F = 83
Brong-Ahafo Wenchi Akrobi M = 61
F = 45
Awisa M = 46
F = 57
Nkonsia M = 60
F = 51
Total 612
6 C. WRIGLEYASANTE ET AL.
Results and discussion
Extreme weather eects at household level
As noted, farmers’ perception of changes in various aspects or elements of the weather over time have
been used as evidence of climatic changes (see for instance Arku, 2013; Djagbletey et al., 2012). In
this study, both male and female farmers reported that there have been signicant changes in weather
patterns in the last 5years. Late onset of the rainy season, early cessation of the rainy season and
decrease rainfall amounts during the rainy season, appeared to be the three key observations among
both male and female farmers as shown in Table 2.
In the Ashanti region, both males and females considered reduced rainfall amounts during the
rainy season, late onset of the rainy season and uncertain onset of the rainy season as the three key
observed changes. Similarly, in the Brong-Ahafo region, whilst men considered late onset of the rainy
season, persistent drought and early cessation of the rainy season, females considered late onset of the
rainy season, early cessation of the rainy season and erratic rain during the rainy season as the three
key observed changes. It thus appeared that both male and female farmers had similar notions of
what constitute climate change and that both were aected more by drought-related conditions than
oods. ese ndings resonates with Arku’s (2013) study which also revealed similar understanding
of climate change among men and women farmers in the Eastern region of Ghana.
e emphasis on drier conditions as observed changes is not surprising as about 50% and 32.5%
of female farmers in Ashanti and Brong-Ahafo regions, respectively, reported that droughts aected
them to large extent. Similarly, 33.2% and 30.3% male farmers in Ashanti and Brong-Ahafo regions,
respectively, also reported that droughts aected them to large extent. ese conditions are, however,
more pronounced in the Brong-Ahafo region than the Ashanti region where crop farmers expressed
heightened concerns about erratic rainfall patterns as this increases uncertainty about planting regimes
and as Acquah and Onumah (2011) have noted, it may also induce diseases and pests leading to mass
crop failures. Such uncertainties about planting regimes were expressed during FGDs as follows:
In the past, one could depend on rain to plant. e rainfall pattern was good but in the past few years, the pattern
has changed drastically. You are never sure of when to plant, sometimes we have to wait till May before planting
and when you plant during this time, you are not likely to get very good returns because the rains can stop and
all your crops can die (Male farmer in Nkonsia in the Brong-Ahafo region).
e rainfall pattern has changed. In the past we could plant 2 times in the year, nowadays it’s only once because
the rains are unpredictable. Sometimes it just rains for a short time and it stops. e weather then becomes hot
and hotter, so once you plant you can lose all your crops. e crops can also easily burn as a result of bush res.
(Female farmer in Akrobi in the Brong-Ahafo region)
ese statements imply that changes in rainfall pattern observed as reduced amounts of rainfall
during the rainy season, long dry spells and persistent drought subsequently aects agriculture par-
ticularly crop production. In all the six study communities, farmers reported of substantial decreases
Table 2.Observed weather changes.
Note: Male (248) Female (328).
Source: Fieldwork, 2013.
Observed Changes
Ashanti Brong-Ahafo Total
Male Female Male Female Male Female
Late onset of the rainy season 56.6 64.2 89.2 88.0 72.9 76.1
Early cessation of the rainy season 18.9 21.9 45.7 52.8 41.6 37.4
Uncertain onset of the rainy season 43.0 39.8 32.9 33.0 38.0 36.4
Erratic rainfall during the rainy season 29.7 23.2 66.7 65.3 48.2 44.4
Long dry spells during the rainy season 16.5 19.6 42.4 47.4 29.5 33.5
Less rain during the rainy season 58.0 53.3 48.2 47.6 53.1 50.5
More rain during the rainy season 2.2 0.7 – 1.3 1.1 1.0
Persistent drought 39.2 38.3 50.7 49.7 45.0 44.0
Rising day temperature 6.8 10.6 11.5 13.4 9.2 12.0
Rising nighttime temperatures 2.5 2.8 3.6 8.2 3.1 5.5
AFRICAN GEOGRAPHICAL REVIEW 7
in crop production with more females in both regions reporting that crop production has decreased
a lot in the last 5years as shown in Table 3.
Only few farmers reported that crop production has increased a lot with more males experienc-
ing increases than females. Again, the increases appear to be higher in the Ashanti region than the
Brong-Ahafo region. e gender dierences in increases in crop production was due to the fact that
men had more access to agricultural resources such as improved seedlings, fertilizers that they may
have used on their own as adaptation options to address the adverse impacts of the changed climatic
conditions. is has been discussed in details in the next section.
Decreases in crop production in the last 5years have aected food security at the household level.
In the Ashanti region, 24% males as against 37.3% females reported that there were months that they
experienced food shortages and these were April, May, June and July. Similarly, in the Brong-Ahafo
region, April, May, June, July, and August were the months in which both male and female farmers
experienced food shortages. Decreases in rainfall also had signicant eects on trading activities in
the sense that whilst crop production decreased, less agricultural goods were being sold. In all the
communities there were gender dierences with more females (65.2%) than males (59.4%), reporting
of experiencing decreases in trading in agricultural goods. is is not surprising since in Ghana trading
in agricultural goods is the domain of women. Indeed, it is noted that 95% of actors in agro-processing
and 85% in food distribution in Ghana are women (Opare & Wrigley-Asante, 2008). Decreased rainfall
also has signicant eect on livestock production. In both regions, more males (26.5%) than females
(18.3%) reported to have experienced decreased livestock production in the last 5years. Again, this
result is not surprising as typically, it is the male who own livestock as compared to females.
Also the FGDs further revealed that decreases in rainfall have consequences on other socioeco-
nomic and psycho-social aspects of the lives of the respondents such as their inability to cater for their
children’s education and even attend other social functions as noted:
…yes because the rains are no longer reliable, it creates a lot of challenges for us, both men and women but espe-
cially the men, because we can’t pay school fees, no money for the home and even attending funeral is a challenge
because you can’t donate. It creates a lot of thinking for us. (Male farmer in Akrobi in the Brong-Ahafo region)
Similar sentiments were raised in all the communities visited even though some women also
insisted that the women, particularly the single mothers and widowed ones, also suered the most.
Nevertheless, the males in particular appeared to be more aected psychologically possibly due to the
traditional role of providing for the home. us in all the communities there were reports of males
going through marriage instabilities and mental stress for their inability to support their families and
most importantly their inability to repay their bank loans. For instance, in Sekyedumase in the Ashanti
region, it was reported that three men died as a result of their inability to repay loans they took from
the bank due to drought and bush res. ese challenges have pushed both males and females to adopt
strategies to manage the situation as discussed in the next section.
Table 3.Food production in the last 5years.
Source: Fieldwork, 2013.
Production in last 5years
Ashanti Brong-Ahafo
Male Female Male Female
Decreased a lot 48.7 56.6 25.7 35.9
Decreased a little 23.1 20.0 51.5 45.8
Remain the same 5.1 3.4 2.4 5.2
Increase a little 8.5 8.0 11.4 9.2
Increase a lot 14.5 12.00 9.0 3.9
Total 100.0 100.0 100.0 100.0
8 C. WRIGLEYASANTE ET AL.
Gender and climate change adaptation practices
In this study both male and female farmers reported using similar adaptation strategies which com-
prised of both o-farm non-agronomic and on-farm agronomic practices (as shown in Table 4). is
is not surprising as the majority of both male and female respondents reported that they have received
some form of information from both governmental and non-governmental organizations (through
their farmer based organizations), to help cope with the climatic shocks such as drought, oods and
other disasters. Nevertheless, there were gender dierences in the adaptation strategies due to gender
inequalities in access to and control over resources, dierent agricultural activities especially the types
of crops grown and dierences in o-farm agricultural activities.
e use of chemical fertilizers was a key agronomic adaptation strategy that both male and female
farmers used. is practice is to improve yields and it appeared to be very common as many of the
traditional crops have given way to maize production, which requires articial fertilization to improve
yield. us even though the farmers during FGDs mentioned that improper handling of agrochemicals
could lead to detrimental eects to both the environment and human health, it was the best option
under the circumstances that they found themselves as it provided a more immediate return on their
investments. However, more males (83.3%) than females (69.1%) used chemical fertilizer as an adap-
tation strategy in both the Ashanti and Brong-Ahafo regions. is was due to the fact that men had
the ability to aord to purchase these chemicals as compared to women. It was revealed during the
FGDs that men tended to have access to agricultural inputs as compared to women since most oen
men had a higher control over household nancial resources, and also could access possibly access
to bank loans to purchase these resources. is conrms other studies that show that many women
small-holder farmers have limited access to credit (see for instance Duncan & Brants, 2004; Opare
& Wrigley-Asante, 2008). is nding about cost and a higher level of non-adoption among women
is also consistent with Ishaya and Abaje (2008) studies in Kaduna State in Nigeria that revealed that
lack of access to resources impede adaptation among women farmers.
Again, slightly more male (58.4%) than female (56.4%) farmers mentioned doing early planting
(a management technique well-known by both male and female farmers) as a favored adaptation
strategy. However, a few more female (56.9%) than male (56.1%) farmers in the Brong-Ahafo region
reported using early planting as a favored adaptation strategy. is perhaps has been associated with
consistent reports of increases in rainfall variability and early cessation of rainfall in the Transition
Zone, particularly in the Brong-Ahafo region (Owusu & Waylen, 2013).
Table 4.On-farm agronomic practices.
Source: Fieldwork, 2014.
Adaptation strategies
Ashanti Brong-Ahafo Total
Male Female Male Female Male Female
Soil improver (inorganic fertilizer) 87.1 79.0 79.3 59.2 83.2 69.1
Early planting 60.7 55.9 56.1 56.9 58.4 56.4
Early harvesting 46.1 49.4 56.9 54.8 51.5 52.1
Same crop two or more times 71.1 58.7 33.4 42.5 52.2 50.6
Planting on raised ridges 34.8 47.1 62.3 49.9 48.5 48.5
Soil water conservation/mulching 10.4 10.8 77.3 78.1 43.8 44.4
Mixed cropping 25.1 47.3 40.8 56.7 32.9 52.0
Drought resistant varieties 49.3 48.7 38.4 20.8 43.8 34.7
Late planting 47.0 45.8 39.3 31.7 43.1 38.7
Late harvesting 42.2 32.6 39.1 46.8 40.6 39.7
Early maturing varieties 27.1 32.3 42.3 72.7 34.7 52.5
Late maturing varieties 30.0 78.8 48.6 53.2 39.3 66.0
Planting in valleys/wetlands 13.8 6.9 10.9 9.5 12.3 8.2
Irrigation 9.2 7.8 11.5 9.2 10.3 8.5
Soil improver (compost/manure) 2.5 0.8 8.5 7.8 5.5 8.6
Rain water harvesting 2.7 – 1.1 – 1.9 –
AFRICAN GEOGRAPHICAL REVIEW 9
When we consider early harvesting, slightly more female (52.1%) as compared to male (51.5%)
farmers mentioned early harvesting as a favored adaptation strategy. Indeed, more females in the
Ashanti region as compared to their male counterparts used this as a preferred adaptation strategy.
It was explained during FGDs that this management practice was a well-known one which farmers
use to avoid attack by pests as well as bush res that occurred during the dry season (following the
minor cropping season). Most importantly, early harvesting was done so that cash-strapped farmers
could gain income from early sales. is possibly explains why more females favored this strategy as
they claimed they needed the nancial resources to support or supplement their household budgets.
Similarly, planting same crop two or more times was a preferred adaptation practice, with more
females in Brong-Ahafo region (42.5%) doing that as compared to their male counterparts (33.4%).
is is a practice where farmers planted the same crop at dierent intervals within and across cropping
seasons to hedge against crop failure and at the same time protect them against dry spells at critical
stages of the crop development. It is therefore not surprising that female farmers in the Brong-Ahafo
region preferred that since the region experience increases in rainfall variability and early cessation
of rainfall (Owusu & Waylen, 2013).
Mixed cropping, also a well-known practice was more prevalent among female farmers (52.0%)
than male farmers (32.9%) and has been intensied. For instance, in Nkonsia in the Brong-Ahafo
region, men resorted to dry season farming of vegetables, such as okro and cabbage, but on a small
scale (usually one acre). ey also planted dierent crop varieties such as maize, tomatoes, pepper,
yam and plantain. On the other hand, women used wetlands to grow exotic vegetables. Similarly, in
the Ashanti region, whilst men in Babasso grew maize, cassava, and cowpea, women grew pepper,
garden eggs, groundnut, beans, yam and cassava. Similar practices existed in all the communities
visited. e intensication of this practice was to ensure that food availability is not compromised. It
is therefore not surprising to nd more women adopting this practice in both regions as it is mostly
women’s responsibility to ensure food security at the household level (World Bank, 2009).
e use of new varieties of existing crop types are important adaptation strategies as observed
by Acquah and Onumah (2011). In this study, both male and female farmers reported using new
varieties of existing crops in the sense that when the rains are more reliable and last for over three
months, farmers manage to cultivate late maturing varieties that may yield more. On the other hand,
they use early maturing varieties during the minor rainy season as a result of the short duration and
erratic nature of the rainfall. However, more females (66.0%) than males (39.3%) reported using late
maturing varieties as well as early maturing varieties as reported by 52.5 and 34.7% of females and
males, respectively. e higher percentage of females using these existing crop varieties is due to the
dierences in adaptation preferences, in relation to the types of crops grown by men and women as
well as the o-farm strategies. All of these were oen dened by the gender norms and practices as
well. For instance, in Akrobi in the Brong-Ahafo region, men mostly engaged in the planting of maize,
cowpeas, tomatoes and new crops mostly watermelon and cashew. On the other hand, women were
engaged more in the planting of traditional vegetables such as tomatoes, green pepper and garden
eggs. is was because women preferred the traditional crops as they were not so sure of the benets
of new crops such as watermelon. is reinforces the view that women were not likely to take risks
as household food security is a priority to them (Opare & Wrigley-Asante, 2008; World Bank, 2009).
Similarly more males (43.8%) as compared to females (34.7%) reported to have resorted to the
use of drought resistant crops implying that more men than women may have access to agricultural
resources such as drought resistant varieties. is supports Swai, Mbwambo, and Magayane (2012)
argument that men led in the use of improved seeds of short-maturity and drought-tolerant crops.
For instance in Awisa in the Brong-Ahafo region, male farmers had diversied into lucrative business
of cashew production (a drought resistant crop). Women, however, engaged mostly in short-term
traditional crops such as garden eggs and okro. ey also worked as factory hands in a cashew pro-
cessing factory rather than producers because cashew production requires large acres of land for which
many women did not have access, limiting their entry into this lucrative venture. is situation also
conrms studies that show that traditionally women’s access to and ownership to land is oen limited
10 C. WRIGLEYASANTE ET AL.
aecting their economic opportunities (Benneh, Kasanga, & Amoyaw, 1995; Duncan & Brants, 2004;
Wrigley-Asante, 2008).
When we compare the two regions, the use of drought-resistant crop varieties was higher among
farmers in the Ashanti region than in the Brong-Ahafo region. is could possibly be the establishment
of the Crop Research Station at Ejura (the municipal town) in the Ashanti region that could be con-
tributing to the better sensitization of farmers and hence their adoption. Ejura also has a long history
of mechanized commercial production of crops especially maize. In other words, there is a stronger
institutional support in the Ashanti region as compared to the Brong-Ahafo region. Moreover, both
male and female farmers generally reported that they received institutional support including support
from NGOs to assist them resolve problems caused by changes and uncertainties in climate. But it
appears more males in both Ashanti and Brong-Ahafo regions had received more support than females.
Interestingly, few farmers had adopted irrigation, rain water harvesting, and planting in valleys
and wetlands as well as the use of compost and manure. e limited use of irrigation in particular
is not strange as agriculture in Ghana is almost entirely rain-fed with less than one percent of arable
land irrigated (Ministry of Food & Agriculture, [MOFA], 2003). In cases where irrigation facilities
are available possibly through the construction of small dams, it may be expensive and it is likely that
women may have limited access to irrigation facilities (Opare & Wrigley-Asante, 2008). is came
up clearly in our study where less than 10.0% of farmers with the majority being males (10.3%) than
females (8.5%) relied on irrigation. Most of these farmers were from Akrobi in the Brong-Ahafo
region, who explained that they depended on a government supported small dam irrigation project
but complained that the pipes were not enough and so only a few farmers benetted from that project.
With regards to the o-farm adaptation strategies, men, particularly the younger ones migrated to
more urban areas and beyond. is came up strongly in the Brong-Ahafo region, particularly Akrobi
where it was reported that young men oen went to Libya as a major strategy. On the other hand,
women oen resorted to more diversied activities particularly petty trading in both household dura-
bles and consumables. is is a major adaptation strategy for women that conform with many similar
studies such as Arku’s (2013) and Dzah’s (2011) studies in the Eastern and Central regions of Ghana,
respectively, which also highlighted petty trading as a key coping strategy for women.
It was explained that it is culturally wrong for men to trade (in food stus) and therefore trading
was the preserve of women as noted in other studies (see for instance Arku, 2013). e engagement
of women in this activity provided them with some form of ‘comfort’ to adapt to climatic changes and
its associated low food crop production as this enabled them to maintain and support their homes.
Subsequently, it provided some gains for women in terms of decision-making at the household level
and these have been discussed in the next section.
Other emerging strategies employed by both males and females included the intensication of
livestock rearing (especially goats, sheep and pigs) and the introduction of new livestock such as
rearing of snails, grass-cutters and poultry-raising. A few farmers also mentioned sh farming and
this was done mainly by the male farmers. ere was also the use of ecosystem services such as col-
lecting wildnuts, mushrooms and spices, mainly for home consumption and surplus for sale which
was a strategy mostly used by the females. is strategy was however, at a very minimal level, as the
high levels of environmental degradation have reduced the provision of ecosystem products especially
mushrooms in supporting adaptation. Rain water harvesting emerged as the least adaptation strat-
egy used because it is mainly required for post-harvest activities (such as washing of foodstus) and
therefore not considered a priority by the farmers.
Gender, adaptation and household dynamics
One of the key objectives of this study was to examine how the decision-making role of particularly
females at the household level has been inuenced by the adaptation strategies. It is well noted that
women are oen excluded and underrepresented in decision-making at the household level (Ardayo-
Schandorf, 1994; Wrigley-Asante, 2008) and especially issues pertaining to climate change (World
AFRICAN GEOGRAPHICAL REVIEW 11
Bank, 2009). Our ndings revealed that women’s engagement in o-farm adaptation strategy particu-
larly with regards to petty trading appeared to have positively inuenced their socioeconomic status
and in turn inuenced their decision-making role at the household level. is conrms other studies
conducted in Ghana that shows that women’s ability to support the home nancially improves their
decision-making role (Brown, 1994; Oppong, 2005; Wrigley-Asante, 2008).
In this study wives were increasingly taking decisions at the household level but oen in consultation
with their husbands in both the Ashanti and Brong-Ahafo regions. According to the male respondents,
wives did participate in decisions on household livelihood activities as well as on household needs.
us joint decision by husbands and wives came up as a common phenomenon. Women contributed
in decisions concerning children’s education, healthcare and daily meals. Indeed, the FDGs further
revealed that women were increasingly taking part in decision-making as result of the increasing
contribution to household budget as reported as follows:
… these days the men listen to the women because they’ve realized women give good advice and also because
they trade, they are sometimes nancially better o and can support the man to take care of the home. Also
women have a lot of information on what is going on with agricultural goods because they trade and that even
informs us on which types of crops are in high demand. (Male farmer at Akrobi in the Brong-Ahafo region)
is situation was attributed to the increasing economic contributions of women, and the structural
transformation of the Ghanaian society that has changed the status quo as highlighted in other stud-
ies such as Oppong (2005) and Wrigley-Asante (2012). Moreover, the wider education on economic
empowerment of women also came up strongly as a factor inuencing the increasing participation of
women in household decision-making. As noted by Kabeer’s (1998) earning an independent income
can lead to an increase in women’s ability to exercise ‘voice’ in the household decision-making process.
is situation creates more freedom for women to take decisions and assume more responsibilities in
the aairs of the household as explained:
… these days the women also have a say particularly in our homes because the men can no longer support their
wives as it used to be. e wives support themselves and even some of the husbands and so the husbands listen
to the women. Also women now know their rights, they talk about these things on radio and at community
durbars and so increasingly the men understand that women must take part in issues concerning the welfare of
the home. (Male farmer in Anyinasu in the Ashanti region)
Other members of the family, such as adult children and adult family members also contributed
to decisions on household livelihood activities. A few women respondents in both regions reported
taking decisions on household livelihood activities alone. ese are most oen the divorced, separated,
widowed or single women and they are most likely to be household heads. is conrms other studies
such as Wrigley-Asante (2012) study in Ghana and Kabeer (1998) study in Bangladesh, that show that
such category of women oen take major decisions on their own.
However, the study found out that decisions concerning the use of land at the household level
was within the domain of husbands and other male adults (such as uncles and male children) in the
household. is is consistent with other studies that show that, traditionally, land has been recognized
as the primary source of wealth, social status and power and therefore men tend to control and take
decisions over such assets (Duncan & Brants, 2004; Opare & Wrigley-Asante, 2008).
Concluding discussions
is paper sought to highlight the gender dierences in climate change adaptation among small-holder
farmers’ and its implication for women’s inclusion in decision-making. e major ndings were that
both male and female farmers knew and could describe the nature of changing environmental con-
ditions such as observed increased temperatures, reduction in rainfall, and variations in the rainfall
regime. In general, it appears both male and female farmers are aected more by drought-related
conditions than oods. e emphasis on drier conditions as observed changes was very important to
them as it resulted in decreased food crop production and its resultant adaptation choices.
12 C. WRIGLEYASANTE ET AL.
To address the negative impacts of climate change, farmers have adopted both o-farm non-ag-
ronomic and on-farm yield enhancing agronomic strategies to address the challenges. e on-farm
strategies included the use of chemical fertilizer as a key adaptation strategy with more males than
females resorting to it. More males than females were also found to have more access to and used
drought resistant crops. Mixed cropping was more prevalent among female farmers than male farm-
ers and with females resorting to traditional crops. is was to ensure that food availability at the
household level is not compromised.
Female farmers also resorted to petty trading in agricultural and consumable goods, a key o-
farm strategy as this t more into their traditional gender role. e income and oen crop market
information gained has in turn increased the resilience of women in adapting to climatic hazards and
empowered them in the sense that it has subsequently improved their decision-making role at the
household level.
Our ndings show that there are some dierences between female and male farmers’ adaptation
strategies. Whilst men are more into on-farm agronomic practices, women are interested and utilize
more o-farm adaptation strategies especially petty trading, in addition to the on-farm agronomic
practices. Petty trading which is linked to the traditional roles of women, therefore provide some form
of resilience to climate change as well as empowering women. We therefore argue that institutions
that support climate change adaptation initiatives at the local level must take gender dierences into
consideration in the discussions of climate change. Whilst mainstreaming gender is important in these
discussions, women must be targeted to consolidate their empowerment in particular. Men must also
not be le out as it appears crop failures aect their psychological well-being. e situation of both
male and females small-holder farmers could therefore be improved through access to micro nance
programs as part of the adaptation support. However, this should be supported with wider capacity
building on entrepreneurship skills as well as continuous sensitization and educational programs on
women’s empowerment issues at the community level. is should also be supported with counseling
services targeting possibly men at the community level so as to assist them withstand the challenges
associated with climate variability and change.
Note
1. e United States Agency for International Development (USAID) by Tetra Tech ARD (USAID Contract No.
AID-EPP-I-00-06-00008, Order Number AID-OAA-TO-11-0006).
Acknowledgment
e authors are grateful to the United States Agency for International Development (USAID) by Tetra Tech ARD (under
the African and Latin American Resilience to Climate Change [ARCC] program), for providing funding for the eld
work that resulted in this publication.
Disclosure statement
No potential conict of interest was reported by the authors.
Funding
is work was supported by the United States Agency for International Development (USAID) by Tetra Tech ARD,
through a Task Order under the Prosperity, Livelihoods, and Conserving Ecosystems (PLACE) Indenite Quantity
Contract Core Task Order [USAID Contract No. AID-EPP-I-00-06-00008, Order Number AID-OAA-TO-11-0006].
AFRICAN GEOGRAPHICAL REVIEW 13
Notes on contributors
Charlotte Wrigle y-Asante is a senior lecturer in the Department of Geography and Resource Development at University of
Ghana, Legon. Her research areas include gender, poverty and empowerment issues of rural women and urban migrants.
Kwadwo Owusu is a senior lecturer in the Department of Geography and Resource Development at University of
Ghana. His areas of research interest include climatology, climate variability and change impacts on agriculture and
water resources and smallholder adaptation to climate change.
Irene S. Egyir is a senior lecturer in the Department of Agricultural Economics and Agribusiness at University of Ghana,
Legon. Her research areas include agri-business and micro nance, climate change and gender.
Tom M. Owiyo is a development economist at the African Development Bank. His research interests include agriculture
and rural development, climate change and development and natural resource management.
References
Abrahamson, W. G., & Weaver, M. R. (2008). EcoSampler: A learning object for community sampling, community
sampler and succession. Bulletin of the Ecological Society of America, 89, 435–447.
Abubakari, A., Lawson, E.T., Mensah, A., Gordon, C., & Padgham, J. (2016). Adaptation to climate change or non-climatic
stressors in semi-arid regions? Evidence of gender dierentiation in three agrarian districts of Ghana. Environmental
Development, 20, 45–58. doi:10.1016/j.envdev.2016.08.002
Acquah, H. D., & Onumah, E. E. (2011). Farmers perception and adaptation to climate change: An estimation of
willingness to pay. Agris on-line Papers in Economics and Informatics, 3, 31–39.
Adger, W. N. (2000). Social and ecological resilience: Are they related? Progress in Human Geography, 24, 347–364.
African Development Fund – Human Resource Development Department. (2008). Ghana country gender prole. Accra:
ADF-OSHD.
Ardayo-Schandorf, E. (1994). Household headship and women’s earnings in Ghana. In E. Ardayo-Schandorf (Ed.),
Family and development in Ghana (pp. 30–47). Accra: Ghana Universities Press.
Arku, F. S. (2013). Local creativity for adapting to climate change among rural farmers in the semi-arid region of Ghana.
International Journal of Climate Change Strategies and Management, 5, 418–430.
Beg, N., Morlot, J. C., Davidson, O., Afrane-Okesse, Y., Tyani, L., Denton, F., … Parikh, J. K. (2002). Linkages between
climate change and sustainable development. Climate Policy, 2, 129–144.
Benneh, G., Kasanga, R. K., & Amoyaw, D. (1995). Women’s access to agricultural land in the household: A case study of
three selected districts in Ghana. Accra: Family and Development Programme (FADEP), Assemblies of God Literature
Centre.
Brown, C. K. (1994). Gender roles in household allocation of resources and decision making in Ghana. Family and
Development Programme (FADEP) Technical Series 2. Accra: Assemblies of God Literature Centre, University of
Ghana, Legon.
CARE International. (2013). Climate change vulnerability and adaptive capacity in Northern Ghana. Accra: CARE
International and Adaptation Learning Programme (ALP).
Codjoe, S. N. A., & Owusu, G. (2011). Climate change/variability and food systems: Evidence from the Afram Plains,
Ghana. Regional Environmental Change, 11, 753–765. doi:10.1007/s10113-011-0211-3
Codjoe, S. N. A., Atidoh, L. K., & Burkett, V. (2011). Gender and occupational perspectives on adaptation to climate
extremes in the Afram Plains of Ghana. Climate Change, 110, 431–454.
Denton, F., Sokona, Y., & omas, J. P. (2000). Climate change and sustainable development strategies in the making:
What should West African countries expect?. Dakar: OECD, ENDA-TM.
Djagbletey, G., Bosu, P., Ametsitsi, G., Addo-Danso, S., Foli, E., Cobbinah, J., Nkrumah, E. (2012, June 25–30). Assessment
of coping and adaptation strategies to climate change in Onso North and South Districts, Ashanti Region, Ghana.
1st UFRO-FORNESSA Regional Congress, Nairobi Kenya.
Duncan, B. A., & Brants, C. (2004). Access to and control over land from a gender perspective. A Study Conducted in the
Volta Region of Ghana. FAO, SNV, WILDAF. Accra: e Advent Press.
Dzah, D. E. E. (2011). Gender dynamics of climate change in Ghana: An intersectional perspective. (MA thesis).
International Institute of Social Studies, e Hague, Netherlands. Retrieved March 12, 2017, from Climate%20
Change/Ellen_Dzah_WGD_2011.pdf
Egyir, I. S., Owusu, K., Jatoe, J. B. D. & Wrigley-Asante, C. (2014, March). Climate change and agricultural adaptation
measures in the transition zone of mid-Ghana: Final research report. Ghana: US Agency for International Development
(USAID) and Multi-Features and Capacity Enhancing Services (MFCS).
Gyampoh, B. A., Idinoba, M., & Amisah, S. (2008). Water scarcity under a changing climate in Ghana: Options for
livelihoods adaptation. Development, 51, 415–417.
14 C. WRIGLEYASANTE ET AL.
Hughes, C. (2012). ose le behind: e impact of migration on Guatemalan women. Montreal: McGill Queens University
Press.
Institute of Social Statistical and Economic Research. (2014). e state of the Ghanaian economy report 2014. Retrieved
from http://isser.edu.gh/index.php/sger/50-the-state-of-the-ghanaian-economy-report-2014
Intergovernmental Panel on Climate Change. (2007). Climate change: Impacts, adaptation and vulnerability (IPCC
Working Group II Report, Chapter 19). 779–810.
Ishaya, S., & Abaje, I. B. (2008). Indigenous people’s perception on climate change and adaptation strategies in Jema’a
local government area of Kaduna State, Nigeria. Journal of Geography and Regional Planning, 1, 138–143.
Jost, C., Kyazze, F., Naab, J., Neelormi, S., Kinyangi, J., Zougmore, R., … Kristjanson, P. (2016). Understanding gender
dimensions of agriculture and climate change in smallholder farming communities. Climate and Development, 8,
133–144. doi:10.1080/17565529.2015.1050978
Kabeer, N. (1998). Money can’t buy me love. Re-evaluating gender, credit and empowerment in rural Bangladesh (IDS
Discussion Paper 363). Brighton: Institute of Development Studies, University of Sussex.
Kalame, F. B., Aidoo, R., Nkem, J., Ajayie, O. C., Kanninen, M., Luukkanen, O., & Idinoba, M. (2011). Modied taungya
system in Ghana: A win–win practice for forestry and adaptation to climate change? Environmental Science & Policy,
14, 519–530.
Kemausuor, F., Dwamena, E., Bart-Plange, A., & Kyei-Baour, N. (2011). Farmers’ perception of Climate change in the
Ejura-Sekyedumase district of Ghana. ARPN Journal of Agricultural and Biological Science, 6, 26–37.
Ministry of Food and Agriculture. (2003). Agriculture in Ghana: Facts and gures. Accra: Statistics, Research and
Information Directorate (SRID).
Nhemachena, C., & Hassan, R. (2007). Micro-level analysis of farmers’ adaptation to climate in Southern Africa (IFPRI
Discussion Paper No. 00714). Washington, DC: International Food Policy Research Institute.
Opare, J., & Wrigley-Asante, C. (2008). Assessment of Gender and Agricultural Development Strategy (GADS) report.
Accra,: Ministry of Food and Agriculture.
Oppong, C. (2005). Conjugal resources, power, decision-making and domestic labour: Some historical and recent evidence
of modernity from Ghanaian families. Accra: University of Ghana.
Owusu, K., & Waylen, P. R. (2012). e changing rainy season climatology of mid-Ghana. eoretical and Applied
Climatology, 112, 419–430. doi:10.1007/s00704-012-0736-5
Owusu, K., & Waylen, P. R. (2013). Identication of historic shis in daily rainfall regime, Wenchi, Ghana. Climatic
Change, 117, 133–147. doi:10.1007/s10584-013-0692-9
Swai, O. W., Mbwambo, J. S., & Magayane, F. T. (2012). Gender and adaptation practices to the eects of climate change
in Bahi and Kondoa Districts, Dodoma Region, Tanzania. Journal of Sustainable Development, 5, doi:10.5539/jsd.
v5n12p65
World Bank. (2009). Gender and agriculture: Sourcebook. Washington, DC: e International Bank for Reconstruction
and Development/World Bank, FAO, IFAD.
Wright, H., & Chandani, A. (2014). Gender in scaling up community based adaptation to climate change. In L. Schipper,
J. Ayers, H. Reid, S. Huq, & A. Rahman (Eds.), Community based adaptation to climate change: Scaling it up (pp.
226–238). New York, NY: Routledge.
Wrigley-Asante, C. (2008). Men are poor but women are poorer: Gendered Poverty and Survival strategies in the Dangme
West District of Ghana. Norwegian Journal of Geography, 62, 161–170.
Wrigley-Asante, C. (2012). Out of the dark but not out of the cage: Women’s empowerment and gender relations in the
Dangme West district of Ghana. Gender, Place & Culture, 19, 344–363.
Wrigley-Asante, C., & Agandin, J. B. (2015). From Sunrise to Sunset: Male Outmigration and its eect on le-behind
women in the Builsa District of Ghana. Gender Issues, 32, 184–200. doi:10.1007/s12147-015-9139-0
Yaro, J. A., Dogbe, T. D., Bizikova, L., Bailey, P., Ahiable, G., Yahaya, T., & Salam, K. A. (2010, December). e Social
Dimensions of Adaptation to Climate Change in Ghana. (Discussion Paper No. 15). Washington, DC: World Bank.
