LAND MARKETS AND WOMEN’S LAND ACCESS
IN NORTHWESTERN TANZANIA
AYALA WINEMAN AND LENIS SAWEDA LIVERPOOL-TASIE
Department of Agricultural, Food, and Resource Economics, Michigan State University, USA
[email protected]
Paper prepared for presentation at the
“2016 WORLD BANK CONFERENCE ON LAND AND POVERTY”
The World Bank - Washington DC, March 14-18, 2016
Copyright 2016 by author(s). All rights reserved. Readers may make verbatim copies of this
document for non-commercial purposes by any means, provided that this copyright notice
appears on all such copies.
Abstract
Land markets can have potentially divergent effects on the gender equity of land access. These markets
may be important for female-headed households if participation in the market is less subject to restrictive
gender norms around land ownership, as compared with traditional systems of allocation. At the same
time, markets may not operate in a gender-neutral manner, with gender restrictions beyond the land
market continuing to influence women’s options. Using both qualitative and quantitative data collected in
2014 and 2015, this paper explores the extent and manner in which women participate in the land sales
and rental markets in northwestern Tanzania. Results reveal that female heads engage with the land
market, particularly using it to augment small land endowments left from their marriages. Nonetheless,
female-headed households are still less likely than others to participate in the market. A qualitative
analysis confirms that, although women do not feel excluded from the market by traditional norms, they
find it most difficult to mobilize the necessary capital. Women also face unique restrictions on selling
land, though the ‘rules’ of women’s engagement in the market are nuanced. The paper concludes with
policy recommendations to better ensure that women may benefit equally from market participation.
Key Words: gender, land markets, qualitative-quantitative methods, Tanzania
1. Introduction
Land access is strongly related to economic well-being in rural sub-Saharan Africa (Jayne et al. 2003),
and land markets, particularly those operating in customary settings, are an important avenue through
which rural households access land. However, these markets are poorly understood and sometimes even
overlooked in policy discourse. As noted by Chimhowu and Woodhouse (2006), “failure to understand
the nature and extent of land markets under customary tenure regimes risks obscuring the processes
through which the poor have access to land and disabling efforts to maintain or improve that access.”
Land markets may be especially important for female-headed households. They can potentially improve
the gender equity of land access if they provide women with an avenue of access outside traditional
channels (World Bank 2008). Conversely, land markets may marginalize women if their functioning
remains limited by traditional gender norms around land ownership, or if gendered restrictions in other
realms leave women unable to mobilize the necessary capital to participate in the land market. Yet little is
known about the extent to which women engage with the land market, the constraints on their
engagement, or the market’s anticipated effects on women’s land access (Whitehead and Tsikata 2003).
This paper begins to fill this knowledge gap by exploring the manner in which female-headed households
participate in land sales and rental markets in rural Tanzania.
We make several contributions to the literature. First, we provide evidence on the extent of female
participation in informal land markets in present-day Tanzania, where land allocation has long been the
responsibility of democratically elected village authorities (Daley 2005a and 2005b), as opposed to tribal
leaders, as is common in other contexts. Second, to our knowledge, this is the first empirical analysis to
explicitly investigate the extent to which women are able to augment their often-limited landholdings
through land markets. Furthermore, our econometric analysis of female-headed households is
strengthened with a qualitative exploration of the opportunities and constraints that land markets offer
women in Tanzania.
The remainder of this paper is organized as follows: Section 2 provides background on women and land
markets, while section 3 presents our conceptual framework and the hypotheses to be examined. Section 4
introduces the quantitative data and provides descriptive statistics regarding the land market behavior of
female-headed households. Section 5 includes results of our econometric analysis. Section 6 introduces
the qualitative data and provides an assessment of the gendered patterns of land market engagement.
Section 7 concludes.
1
2. Women and Land Markets
Land rights are recognized as vital for women’s economic well-being in rural contexts. In India, Agarwal
(1994) cites the gender gap in ownership and control of property as an important determinant of the
gender gaps in income and levels of empowerment. Control of property affects women’s welfare through
multiple pathways. Women’s possession of immovable property is found to be associated with less
spousal violence in India (Panda and Agarwal 2005). This indicates that access to land and housing
outside of marriage effectively provides wives with an ‘outside option’ to marriage, shifting the balance
of power within their households. A similar dynamic is observed in Tanzania (Dancer 2015), where men
seem to perceive that women become more ‘arrogant’ within marriage if they believe they can leave their
husbands and return to their fathers’ land. Dancer (2015) specifically notes that these women are more
likely to protest physical abuse. In a case study conducted in India, Agarwal (2003) finds that land
ownership also seems to provide women with security from eviction, a serious concern upon widowhood.
In Tanzania, Wineman and Liverpool-Tasie (2015a) find that parents seem to utilize their bequest
decisions to solicit care from their children as they age. This suggests that, for both men and women,
bequest rights over land are also likely to be associated with better care in old age.
In sub-Saharan Africa, the policy discourse around land tenure is typically based on an assumption that
women have weaker land rights than men (Pedersen 2015; Whitehead and Tsikata 2003). Women’s rights
are found to be contingent on their position within the family, with access mediated through their
husbands, fathers, and brothers. Furthermore, women’s rights are assumed to be ‘secondary’, in that they
may be able to use land but rarely possess the right to transfer land to others (e.g. through sale or bequest)
or to exclude others from the land (Lastarria-Cornhiel 1997). Although some authors observe that
customary practices are able to provide for women’s needs specifically because they are flexible,
reflecting evolving needs and social relations (Tripp 2004), others maintain that women’s land access
within such systems is precarious. More specifically, where land scarcity is a growing problem, women’s
rights to land are no longer guaranteed within customary systems. In the Arusha region of northern
Tanzania, Dancer (2015) finds that the welfare of widows is contingent on the goodwill of their in-laws,
while divorced women are often unable to claim any marital land (the land held jointly with their
husbands during marriage). Land inherited by women is more readily subject to challenges, especially
from their male relatives. In the event of family disputes over land, women’s interests often depend on
their kinship relations and on the power dynamics within their families. In response to this pattern, a
World Bank (2008) report on gender and agriculture cites land markets as potentially able to reduce the
influence of family structure for women’s land access.
2
The debate on the value of land markets for women continues. Several studies have provided accounts of
women independently participating in the land market in sub-Saharan Africa. In the Buganda region of
Uganda, women across the wealth spectrum have been observed to purchase land (Troutt 1994; Tripp
2004), with female-headed households even more likely than male-headed households to do so. In areas
with the most active land markets, the holdings of female heads closely resembled those of men,
suggesting that land markets effectively improved women’s land access. According to Tripp (2004),
“purchasing land has, in effect, become a way of circumventing the traditional authorities.” This
observation is also echoed by Gray and Kevane (1999), with reference to both East and West Africa. In
the Iringa region of Tanzania, Daley (2008) finds that the land market itself is not directly eroding
women’s land rights, though women remain marginalized as the market privileges those with money. By
the late 1990s, approximately one-fifth of all market transactions were undertaken by female-headed
households. Daley concludes that there are “no absolute gender obstacles to obtaining land”, and while
land access is not gender-neutral, women with adequate finances or social capital are indeed able to
acquire their own land. These accounts suggest that, in a fluid land market driven by the laws of supply
and demand, the impersonal nature of transactions can potentially ease women’s access to land.
On the other hand, a number of authors conclude that women have little to gain from land markets. In an
early examination of the gender implications of privatized tenure systems, Lastarria-Cornhiel (1997)
concludes that women’s customary claims to land are often diminished when all the rights associated with
land ownership are claimed by a single person. In Kenya, Mackenzie (1990) observes that women’s
claims to land grew insecure as land became increasingly commoditized. As potential buyers, even elite
women were seen to have difficulty purchasing land in their own names.1 Razavi (2007; 2009) similarly
claims that “the market system of land allocation is not free and impersonal, but rather deeply embedded
within local power structures”. Along these lines, Sitko (2010) documents how the development of an
informal land market in southern Zambia has excluded women from participation. This market is illegal
under customary law, and owing to its underground nature, participation requires social capital to hide or
protect a transaction from the authorities. Sitko notes that “the market system of land allocation is not free
and impersonal, but rather deeply embedded within local power structures” that effectively sideline
women.
Yet another obstacle to women’s market participation as renters or buyers is that such participation
requires the mobilization of capital (Daley 2008). As noted by Lastarria-Cornhiel (1997), “often women
1
As noted by Yngstrom (2002), Mackenzie’s (1990) study site in Central Province, Kenya, was characterized by an
increasingly severe land shortage in the 1950s. These unique circumstances may render this case study less
generalizable to other regions.
3
enter the market system with no property, little cash income, minimal political power, and a family to
maintain.” Women find it difficult to accumulate capital, partly because land is the most important asset
used to build wealth in rural Africa, and women begin with a weaker claim to land in most customary
systems. Even in settings where women inherit land, they have been found to inherit (on average)
significantly less than their brothers (Wineman and Liverpool-Tasie 2015a). Furthermore, women’s
productive work within the household is unremunerated, and husbands may curtail their wives’ access to
cash income, leaving them without savings (World Bank 2008). Razavi (2007) concludes that, for the
majority of women, land markets are unlikely to serve as a channel for inclusion.2
This dispute over the merits and drawbacks of land markets for women entered into the debates around
land tenure reform in Tanzania that preceded the Land Laws of 1999 (Tsikata 2003).However, the
diverging observations and arguments of these authors have not been tested through rigorous quantitative
analysis in any setting. Consequently this paper contributes to the ongoing debate by providing a rigorous
analysis for rural Tanzania. Using data from the Kagera region of Tanzania and various econometric
methods, we empirically test whether female-headed households engage with land markets in response to
the amount of land they were left with at the time their marriage ended. We also augment this with a
qualitative analysis that explores the extent to which cultural norms and other factors restrict their ability
to participate in the market.
3. Conceptual Framework and Empirical Strategy
This analysis is based on a conceptual framework (Figure 1) that illustrates the various avenues through
which female heads are able to access land in a rural setting. A woman who becomes household head
(almost always after widowhood or divorce/ separation) can potentially retain land that had been held
during her marriage. This may have been held jointly by husband and wife, or independently by either
individual. As noted in section 2, it is not uncommon in some contexts for women to see their claim to
marital land challenged after a marriage has ended. A female head can also receive land from her own
family in the form of a gift or inheritance once she becomes single. Together, these forms of inheritance
constitute the more ‘traditional’ modes of land access for female heads. Alternatively, a woman can also
access land through the market, either by purchasing land on her own, or by renting/ borrowing land. (As
will be discussed, the distinction between renting and borrowing is imprecise.) Because a woman’s
2
Another argument against the value of land markets for women is that markets do not prevent –and sometimes
open new possibilities for– the mistreatment of women. For example, husbands may sell off the marital land before
exiting a marriage, leaving their newly single ex-wives without a livelihood. However, this topic is beyond the scope
of this paper, which focuses solely on female household heads.
4
landholdings can be comprised of land sourced through any of these avenues, it is possible for women to
compensate for a lack of land from one source by increasing the area accessed through another source. It
is also possible for women to avail themselves of other modes of aquisition, including allocation by local
authorities or clearing unclaimed land, although this is quite uncommon.
This paper explicitly investigates two hypotheses. First, female heads are less likely to participate in the
land market, as compared with households headed by men. The following general equation is used to
evaluate this hypothesis:
𝑌𝑖 = 𝛼 + 𝛽1 [𝐹𝐻𝐻𝑖 ] + 𝑿𝒊 𝜷𝟐 + 𝜀𝑖
(1)
where 𝑌𝑖 is a measure of land market activity for household 𝑖, 𝐹𝐻𝐻𝑖 indexes whether a household is
headed by a woman, 𝑿𝒊 is a vector of household characteristics, and 𝜀𝑖 is a stochastic error term. 𝑌𝑖 takes
the form of a binary indicator for having purchased, sold, or rented in land. 𝑿𝒊 should capture the
household characteristics that might otherwise determine land market behavior. Holding these constant, a
negative value for 𝛽1 indicates that female-headed households are less likely to engage with the market,
as compared with their male counterparts.
Our second hypothesis is that female heads access land through the market more readily if they possessed
less land when first becoming a female head. In other words, they use the market to compensate for a
small endowment left from their marriage. We evaluate this hypothesis with the following general
equation:
𝑌𝑖 = 𝛼 + 𝛽1 [𝐼𝑛𝑖𝑡𝑖𝑎𝑙_𝑎𝑐𝑟𝑒𝑠𝑖 ] + 𝑿𝒊 𝜷𝟐 + 𝜀𝑖
(2)
where 𝑌𝑖 can take the form of a binary indicator for having purchased or rented in land, or a continuous
measure of the total amount of land purchased or rented. 𝐼𝑛𝑖𝑡𝑖𝑎𝑙_𝑎𝑐𝑟𝑒𝑠𝑖 is a measure of a household’s
land endowment at the start of the study period. This equation will be applied to female-headed
households. A positive value for 𝛽1 indicates that women with relatively larger initial land holdings
participate most actively as purchasers and/ or renters, and the opposite is true if 𝛽1 is negative.
4. Quantitative Data and Descriptive Statistics
The data used for this analysis come from an impact evaluation of community-based legal aid (CBLA)
undertaken by the International Food Policy Research Institute. This evaluation took place in 2013 and
2014 in two districts of the Kagera region of Tanzania, namely Karagwe and Biharamulo (Figure 2). All
quantitative analyses in this paper draw from the 2014 survey round, as key variables were only collected
in this year. Because the relevant information is retrospective or would not be influenced by this shortterm intervention, it should not affect our results. Kagera is located in the northwestern corner of
5
Tanzania and shares a border with Uganda, Rwanda, and Burundi. Agriculture is the main activity in this
region (de Weerdt 2010), and as will be discussed, Kagera is characterized by a burgeoning land market
in which a majority of households participate.
In the two study districts, 139 of the 142 rural villages were surveyed. A listing was conducted in one
randomly selected hamlet3 in each village to stratify the selection of 12 households equally by gender of
household head. 1,434 households were interviewed in 2014, bringing the rate of attrition from 2013 to
10.0%. Household population weights are used in all analyses, and are adjusted using inverse probability
weights to reflect the likelihood of remaining in the sample in 2014. To ensure comparability, analysis is
limited to households with heads that are either unmarried or monogamously married (668 male-headed
and 629 female-headed households). A community-level survey was administered to key informants in
each village. The household survey included household-level modules regarding asset holdings, land
parcels held, and instances of land disposal for the period 2008-2014. In this paper, the definition of
‘rental’ includes borrowing arrangements because it seems plausible that borrowing entails a cost for the
borrower (e.g. labor to clear the field or protect it from fires), even with no money exchanged. Odgaard
(2006) similarly notes that few borrowing arrangements in Tanzania are genuinely ‘free of charge’.4
Table 1 outlines the basic characteristics of male- and female-headed households (MHHs and FHHs).5 As
expected, FHHs tend to be smaller, but with a higher proportion of non-working age members. Compared
to MHHs, FHHs have significantly smaller owned landholdings (2.6 versus 4.4 acres) and rented holdings
(0.18 versus 0.39 acres). However, they do not differ significantly in terms of land accessed per capita,
where land access includes both owned and rented/ borrowed land. FHHs are significantly less likely to
have purchased land within the past 6 years. However, note that female heads tend to be older (and their
households smaller), placing them at a different stage of the farm-household life cycle. At the same time,
female heads are significantly less likely to report the right to sell any owned plot (54% versus 67%
among male heads), whether as a joint or exclusive decision. Among all households in the study site (both
male- and female-headed), a majority (62%) site possess at least one parcel that was purchased, and this
exceeds the 52% that possess inherited land. Many of these transactions are sealed with a sales contract,
3
Each village is comprised of several hamlets, or sub-village administrative units (mean = 6.7 hamlets, mean hamlet
size = 106.8 households).
4
Although not reported here, results of the econometric analysis remain consistent when borrowed land is excluded
in a test of robustness. In fact, female heads are somewhat more likely to label land as borrowed, and hence even
less likely to rent land when the definition is restricted. Note that the status of sharecropping was not distinguished
in the survey.
5
In actuality, FHHs comprise 14% of households in the study site, though they are over-represented in this sample.
6
even as less than 0.1% of plots have either a land title or ‘customary right of occupancy’ certificate. This
underscores the informal nature of the land market.
Figure 3 provides information on the relationship between households and the individuals or entities from
whom each plot was purchased or rented, disaggregated by gender of the household head. Among
purchased plots, a majority (59%) are acquired from friends or neighbors. However, it seems that FHHs
are somewhat more likely to buy land from an individual of no relation. Among rentals, a larger
proportion of plots are sourced from extended family. For MHHs, this is more likely to be the family of
the male head, while for FHHs, plots are more commonly accessed through the woman’s family. It thus
seems that patterns of women’s land access shift upon becoming a female head. Interestingly, it is
relatively uncommon to rent a plot from an individual of no relation (just 7% of all rented plots). Perhaps
this is because such temporary arrangements, often with delayed payment, require a level of trust not
found amongst strangers.
We next turn our attention to patterns of land purchase within different subsets of FHHs. Among all
female heads (including 457 widows and 160 women who are separated or divorced), the average land
area purchased after a marriage has ended (i.e. after the year of divorce or widowhood) is 0.45 acres.
However, just 21.1% of FHHs have purchased any land during this interval. Table 2 provides summary
statistics for FHHs that have and have not purchased land since the women became household heads. In
the former category, households have purchased an average of 2.13acres since their marriage ended. The
area of land held from their marriage is estimated with reference to the area of currently-retained
household land that had been held at the year of widowhood or divorce. According to our estimates,
female heads that have independently purchased land were left with considerably less land when they
became single (0.55 vs. 2.02 acres). Women who have purchased land have been household heads for (on
average) four years longer than those who have not purchased land. As well, a greater share (42%) of
those who have purchased land are separated or divorced, rather than widowed.
5. Econometric Analysis
This section first evaluates whether FHHs are as likely to participate in the land market as households
headed by men, once we control for other household characteristics. A seemingly-unrelated multivariate
probit regression (SUR) is appropriate to identify the determinants of land market participation, as
decisions to rent, purchase, and sell land are likely to be related, and the SUR allows the error terms to be
correlated across equations. Using the available information on land transactions in 2008-2014,
7
households are categorized by whether they purchased and/or sold land during this time period, and
whether they currently rent land.6 The equation is:
𝑌𝑖𝑣 = 𝛼 + 𝛽1 [𝐹𝐻𝐻𝑖𝑣 ] + 𝑿𝒊𝒗 𝜷𝟐 + 𝑽𝒗 𝜷𝟑 + 𝜀𝑖𝑣
(3)
where 𝑌𝑖𝑣 alternately indicates whether household 𝑖 in village 𝑣 has purchased or sold land since 2008 and
whether it currently rents land, 𝐹𝐻𝐻𝑖𝑣 is an indicator that household 𝑖 in village 𝑣 has been headed by a
woman, 𝑿𝒊𝒗 is a vector of household demographic and wealth characteristics, 𝑽𝒗 is a vector of village
characteristics that may influence household behavior on the land market, and 𝜀𝑖𝑣 is a stochastic error
term. Because a household’s initial land stock is an important determinant of land market behavior, 𝑿𝒊𝒗
includes a measure of the amount of land owned as of 2008 (for the sales market) or one year ago (for the
rental market). To reflect the time interval of this analysis, the status of FHH for the two sales market
equations is given to those who were headed by a woman during the entire 6-year interval (i.e. they were
widowed or divorced prior to 2008). For the rental market equation, this reflects a woman’s headship over
the previous year. In all analyses in this section, standard errors are clustered at the village level to
account for the potential correlation of shocks to the local economy and land market within the same
village.
Results show that the coefficient on FHH is consistently negative, even when we control for household
demographics and wealth indicators (Table 3).7 These include the household labor endowment (number of
working-age adults), initial land endowment (acres owned at the start of the relevant interval for each
equation), and wealth (value of non-land assets, including farm equipment). Consistent with our first
hypothesis, it seems that women are significantly less likely to participate in the market as buyers, seller,
or renters. This suggests that women are somewhat marginalized in these markets, though it is important
to note that these results may be affected by unobserved factors (omitted variables) that influence the land
market behavior of FHHs.
To test our second hypothesis, we narrow our focus to FHHs and use a seemingly unrelated bivariate
probit (SUR) model to explore what determines whether a FHH accesses land through the market.
Specifically, we focus on whether FHHs respond to the amount of land they were left with at the time
their marriage ended. The equation is:
6
Unfortunately, the data set includes few observations of land leased out, perhaps due to absentee landlords or to
inadvertent or intentional under-reporting. A similar discrepancy is seen in a nationwide agricultural survey in
Tanzania (Deininger et al. 2015).
7
Although not reported here, results remain robust when indicators of wealth are removed from the right hand side
of equations (3) and (4). These are likely to be endogenous with prior land market behavior, as land is used to
generate wealth.
8
𝑌𝑖𝑣 = 𝛼 + 𝛽1 [𝐴𝑐𝑟𝑒𝑠_𝐹𝐻𝐻𝑖𝑣 ] + 𝑿𝒊𝒗 𝜷𝟐 + 𝑽𝒗 𝜷𝟑 + 𝜀𝑖𝑣
(4)
where 𝑌𝑖𝑣 alternately indicates whether the female head has purchased land since she became single, and
whether she currently accesses land through rental. 𝐴𝑐𝑟𝑒𝑠_𝐹𝐻𝐻𝑖𝑣 refers to the area of currently-retained
household land that had been held at the year of widowhood or divorce. (Note that this estimate
necessarily does not account for any land that was sold off, seized, or abandoned since that time. As
𝐴𝑐𝑟𝑒𝑠_𝐹𝐻𝐻𝑖𝑣 is a lower bound estimate of a woman’s true post-marriage endowment, 𝛽1 may be biased
upward.) All of the household and village controls of Table 3 are also included in this model, although
these coefficients are not reported. In Table 4 (columns 1 and 2), results indicate that women are more
likely to purchase or rent if they began with a smaller land endowment, suggesting that FHHs use the
market to compensate for a smaller initial land stock. Recognizing that women face two non-market
channels to access land, including both marriage and personal inheritance (see Figure 1), we include a
new variable for the amount of land the head has inherited or received as a gift since her marriage ended
(columns 3 and 4). Results indicate that women seem to compensate for their personal inheritance through
land purchases.
In Table 5, tobit models are used to estimate the determinants of land area purchased or rented by FHHs,
as a function of land acquired through marriage and the head’s post-marriage inheritance. A left-censored
tobit model is appropriate for the purchase and rent models because a sizable proportion of households
possess no purchased or rented land. Equation (4) is used, with 𝑌𝑖𝑣 now a continuous measure of land area
purchased or rented. The results again indicate that women use the market to compensate for the land area
they were left with, and this is true for both the sales and rental markets (columns 1 and 4). In fact, the
coefficients on initial land endowment are not significantly different from a value of -1, with one less acre
of inherited land associated with one additional acre obtained through purchase or rental. Recalling that
widows are significantly less prevalent among female heads that purchase land (Table 2), we also
disaggregate these households by whether they are widowed or separated. The relationship between land
retained from marriage and subsequent land purchase or rental appears to be strongest for women who are
divorced or separated (columns 3 and 6). In other words, the land market is most important for women
who are separated/ divorced and likely to have retained less from their ex-husbands. Women, and
particularly widows, also use the sales market to compensate for their limited inheritance (columns 1 and
2), and we cannot reject the hypothesis that the coefficients of land retained and land inherited are equal.
Although just 21% of female heads have purchased land since becoming widowed or separated, and just
9
11% rent land at any given time, it seems that these women effectively use the market to compensate for a
small endowment obtained through more traditional channels.8
6. Qualitative Analysis
The qualitative data come from a set of semi-structured in-depth interviews and focus group discussions
held in the study site in 2015. Two villages were selected in each district with the aim of capturing a
diversity of community characteristics (Table 6). This was prioritized in order to identify themes related
to women’s land access that cut across the heterogeneity of different communities (Patton 2015). In each
of the four villages, a comprehensive household census was conducted in one randomly selected hamlet in
order to identify all female-headed households that have ever participated in the land market. Within this
group, five household heads were randomly selected to be interviewed, including (where possible) three
women who had bought or sold land, while the rest were renters. The sample of 20 women spans a wide
spectrum of ages and marital situations (Table 7). In one village, gender-disaggregated focus group
discussions were also held regarding women’s participation in the land market. All conversations were
structured by interview guides, touching on the respondents’ experiences of land market engagement.
Among 64 female-headed households listed in this exercise, 52% have (at some time) independently
rented in land, 5% report having rented out land, 25% have purchased land, and 5% have sold land. A
majority of these women (66%) have turned to the land market at some point since becoming a household
head. Before we attempt to trace out the ‘rules’ associated with women’s participation in the land market,
an example will help illuminate the context in which women navigate this market. Miriam9 (age 72) had
been in a polygamous marriage and successfully campaigned for ownership of a portion of her husband’s
land. He died in 1987 when her children were still young, and she supplemented her income with a
business of selling home-brewed beer. When a troubled neighbor approached with an offer to purchase
their plot of land, she was happy to agree. She said, “I had saved my money for a reason, as I knew my
children were still young and they had to go to school. So I bought that land as a way of keeping my
money, so that I could sell it in the future to take care of my children.” In other words, she viewed land as
a secure way to store savings. Miriam did not feel that being a woman influenced this transaction, saying
“so long as you had your money, you could purchase land.” The year was 1990. Seven years later, one of
her children was imprisoned and required bail money. Seeing no alternative, and with no rental market in
existence, Miriam now sold that same plot to a brother and successfully secured her son’s release. She
8
Note, however, that this analysis necessarily overlooks the women who became widowed or divorced, but then
changed their status. For example, some women remarry or join the household of a sibling, and we cannot observe
their hypothetical land market behavior had they not self-selected out of being a female head.
9
Names were changed to protect respondents’ identities.
10
said, “for women to sell land is not easy unless someone has an emergency like I had.” Since then, the
land inherited from her late husband has been allotted to each child.
Several themes that emerge from the qualitative analysis are evident within Miriam’s story (and
corroborated by others). First, land is more than a factor of production. For our respondents, land
represents security and independence, and further serves as a savings mechanism and future bequest. This
diversity of land’s value for women is consistent with the accounts of Agarwal (1994 and 2003). One
respondent noted that land is more secure than other potential investments. “For goats or cattle, you can
lose them even the next day. But land is there to stay.” Another widow explained her reason for retaining
uncultivated land as one of providing security for her married daughters. “In case they get divorced and
decide to return home, where will they stay when I have sold their land?” For women who may lose
access to marital property upon the end of a marriage, the security offered by an alternative land source is
evidently quite important, a conclusion similarly reached by Dancer (2015). As with Miriam, several
women who had purchased land noted that it provides a relatively safe means to store their money – an
otherwise challenging task. “If I had not bought land, I would have already used up all the money.” At the
same time, the ability to liquidate land is itself a source of security, as plots may be sold in the event of an
emergency.
The second theme of our work was somewhat surprising, as the female heads we interviewed indicated
that gender generally did not function as a restriction in purchasing or renting in land. Similar to
Miriam’s account, respondents frequently remarked, “so long as you have money, you can buy land.” One
woman commented that, since the land purchased by both men and women goes toward their children’s
inheritance, the joint effort is reasonable. Most respondents simply felt that wealth overshadows gender as
a determinant of land access, and this sentiment was shared by the men in our focus group. One man
noted, “as long as you have money that I need, I will sell my land to you. We don’t need to give priorities
based on gender because if a woman says I want to buy land, it means she has money and she intends to
buy it.” It seems these same conditions extend to the rental market, in a manner similar to that observed
by Daley (2008). This finding is also consistent with the argument put forth by Whitehead and Kabeer
(2001) who claim that limited land access, in itself, is not a major cause of the poverty of women in subSaharan Africa.
Third, we observe that land markets are used by separated/ divorced women, and, for at least one
respondent, land accessed through the market served to facilitate an exit from marriage. In that case,
a divorced woman who had suffered with her husband’s (unspecified) behavior developed the idea of
11
purchasing land while still married. She said, “I knew one day we might separate, so I stayed alert…”
According to her account, she initiated the divorce and was no longer able to use the land she had been
cultivating. She therefore moved back with her mother and managed to save money and purchase her own
land. Though it was not precisely a pre-calculated strategy, it seems the possibility of purchasing land at
least played a role in the decision to leave her husband. This is similar to the account of another woman in
the Kagera region (described in Wineman and Liverpool-Tasie 2015b) who, upon leaving her husband,
chose not to rely on her brother but instead struck out on her own, renting and later purchasing land in a
new village. For at least some women, the land market seems to represent an exit option from a bad
marriage.
Fourth, the interviews unambiguously reveal that female heads (and women, generally) tend to have
less access to money with which to enter the market. First, women with children have many strains on
their budget, with a responsibility to care for the immediate needs of their family. It thus becomes
difficult to save the amount necessary to rent or purchase land. “The challenges are in saving money
because sometimes you use the little amount you have collected to cover some other expenditure.” Within
marriage, it is also difficult (though not always impossible) for women to accumulate savings. One focus
group participant noted that, even if a husband and wife cultivate the farm together, “it is her husband
who claims the money.” For this reason, women do not possess a hefty bank account when they become
household heads – a situation often more critical for women who are divorced or abandoned. One woman
noted that she had never possessed 100,000 TSh (roughly USD $50 in 2015) as her own money. This is
consistent with the observations and concerns of Lastarria-Cornheil (1997) and Razavi (2007). Whitehead
and Kabeer (2001) similarly claim that limited access to non-land factors of production is the underlying
cause of women’s poverty in rural sub-Saharan Africa. In our sample, most respondents had saved up
their rental or purchase fees through agricultural production (selling crops from land they already access),
wages (working on others’ farms), and small businesses (e.g. selling charcoal, making mats, or working
as a seamstress).
Our fifth observation is that, while women who are widowed or separated hold considerable freedom to
engage independently with the land market (at least as buyers or renters), women with husbands possess
far fewer rights over property. Within marriage, decisions over land are commonly made by the
husband (though perhaps after consultation with his wife). Even if a woman has saved money from her
own small business, the intent to purchase land must be vetted by her husband. As one respondent
observed, “A married woman will always be under the feet of her husband… It will be hard for her to buy
land or any other important asset…” Both men and women insisted that a woman who purchases land
12
without the direct involvement of her husband will be regarded with suspicion. “Within no time, you find
yourself back at your parents’ home.” Some wives do attempt to creatively assert ownership of property.
We heard of women who purchased land far from home and beyond their husbands’ gaze. One
respondent had a male acquaintance purchase a cow for her and keep it within his own herd. This
transaction was kept secret from her husband.
Sixth, in contrast to the expansive rights of female heads to purchase and rent land, we heard of many
more restrictions on women’s rights to sell land. It seems women are allowed to sell in response to an
emergency, as in the case of Miriam (above), but not for other reasons, such as the desire to invest in a
business. One woman noted, “For a woman to sell land, it is not easy unless you have a big problem.
Otherwise you cannot do so as a woman, but a man can do so at any time without seeking permission
from anyone.” Thus, a woman who sells land becomes an object of her clan’s scrutiny. “The one who
knows your problem cannot bother you. But those who don’t know what you are going through will look
intently upon you.” Often, a woman’s children must be consulted for approval, and sons, in particular,
may hold veto rights over a mother’s decision to sell land. The women we spoke with prefaced their
narratives by earnestly explaining that they had no choice but to sell land. However, the same
clarifications were never offered before telling of a land purchase. This seems to signify a ‘split morality’
around the land market, with norms that differ on either side of the market.
Interestingly, this confusing divergence between a woman’s right to purchase and sell land seems to
disappear when a woman has independently purchased land. For these cases, decisions regarding land
disposal are largely outside the purview of the clan. One respondent noted that, although a woman must
inform her husband of a plan to sell land she had purchased, “he can’t refuse because he knows you
bought it using your own money”. This freedom to sell land was evident across many circumstances (i.e.
different intentions or family structures), as long as it was initially acquired through purchase. It thus
seems that land, once purchased, has somehow ‘exited’ the clan system for as long as a woman (or man)
holds it. This pattern is also reflected in the survey data: While household heads have some sales rights
for 72% of purchased plots, this value is just 55% for inherited plots. Land acquired through inheritance is
evidently subject to greater restrictions.
This qualitative exercise reveals a reality more complex than can be extracted from econometric analysis.
The land market clearly plays a large role in how female-headed households structure their livelihoods. At
the same time, women are subject to gender-based restrictions with regard to selling land that was not
independently purchased, and owing to gender roles, they are less able to raise money to purchase or rent
13
land. However, the manner in which women can or cannot participate in the market is nuanced. Female
heads are far from sidelined, and as the land market continues to develop, it does seem that women’s land
access may simultaneously expand. At the same time, if land prices continue to rise (a common
observation during interviews), then women may be the first to be priced out of the market.
7. Conclusions
This paper has examined the land market behavior of female-headed households using both quantitative
and qualitative methods. Contrary to near-universal claims that women are dependent on men for access
to land, female heads in Kagera are observed to participate in the market. Roughly one in five female
heads have purchased land after they become widowed or divorced/separated, and within this group, the
average size of purchased land is 2.1 acres. This indicates that women in Kagera are not excluded from
the market, as has been documented in other contexts (Sitko 2010). We further find that female heads
effectively compensate for the amount of land held when they became single or widowed, as well as the
land they have individually inherited. Yet female-headed households are significantly less likely than
other households to participate in the market as buyers, sellers, or renters. Although these results do not
imply causation, as we cannot claim exogenous variation in access to the land market, our econometric
analysis controls for many confounding factors, with results that consistently indicate that women are
somewhat marginalized in the land market.
These quantitative results are supported with evidence from our qualitative analysis, which reveals that
women often feel they have the right to buy or rent in land, and in some cases, this option seems to
facilitate an exit from an unhappy marriage. In other words, the right to buy or rent in land appears to
make women somewhat less dependent on their status as a wife in order to access land. Our analysis
complements that of Pedersen (2015), who uses a case study to document how Tanzania’s 1999 Land
Laws have improved land access for women (especially female heads). While noting that access is
becoming less gendered, Pedersen does not consider the role of land markets in this trend. However, we
also find that women face difficulties accessing or raising money, and in fact, many respondents feel that
wealth outweighs gender as a determinant of land market participation. At the same time, our analysis
reveals a fascinating complexity around gender and land markets, whereby women face asymmetric
freedoms on either side of the market. Even if female heads are active renters and purchasers, they are
burdened with particular restrictions on the sale of land and, thus, do not benefit equally from market
engagement.
14
These findings indicate that policies and programs should be designed, as much as possible, to ensure
women’s equal access to land through the market. This avenue of land access seems to be particularly
important for rural women who are separated or divorced, providing them with some security and
independence. Once women are able to purchase land on their own, it seems they also gain transfer rights,
or the freedom to dispose of it at their discretion. Thereafter, women can bequeath or liquidate their land
stock as they need or desire, potentially investing in other productive activities within or outside
agriculture.
As access to land through the market seems to bring benefits, how can women’s market participation be
advanced? First, women’s options to generate income should be expanded. As women increasingly
assume the role of income earners (particularly in higher-return activities), they will progressively gain
access to land through the market – with all the attendant benefits. Income generation can be facilitated
with targeted training to provide job skills, as well as improved and equal access to credit in order to start
or expand a business – or, naturally, to purchase or rent land. Second, women’s incomes would be
augmented if more fathers are encouraged to provide child support upon separation from their wives.
Though this is unlikely to be addressed through statutory law, norms may be shifted through sensitization
around gender equality – a concept likely to be enshrined in the new Tanzanian Constitution. Training
regarding child support can also be offered to local dispute mediators and legal advisers, as well as
members of Ward Tribunals (the lowest rung of the formal legal system). Third, women’s contributions to
household production have been recognized by law for several decades (Law of Marriage Act, 1971).
However, it is evident that women do not actually have equal access to household income and other
resources. This, too, is a topic where norms can potentially be shifted through sensitization, and legal
inroads can be made through the courts.
Further research is needed to explore whether and how women’s post-marriage welfare is correlated with
land market activity, and whether non-market systems effectively provide for women’s needs in the event
of widowhood or divorce. Intriguingly, Dancer (2015) observes that brothers are now less likely to care
for their divorced sisters, as compared with past generations. This change in social norms may reflect
women’s growing market-based options to be self-sufficient, suggesting a more complicated story around
the influence of land markets on women’s welfare. At the same time, women’s exclusion from family
land may be a response to other factors, such as heightened land scarcity. The precise relationship
between land markets and women’s well-being merits greater attention.
15
Acknowledgements
The authors would first like to acknowledge financial support from the Gender, Justice, and
Environmental Change Dissertation Research Fellowship at Michigan State University. They also thank
Dr. Valerie Mueller of the International Food Policy Research Center for providing access to the
household data set. The authors are further grateful to Dr. Leah Lakdawala, who offered helpful feedback
on an earlier draft of this paper, and to Gilbert Ntimba and Victor Rusetta, our excellent fieldwork
assistants in Tanzania.
16
References
Agarwal, B. (1994). A field of one’s own: Gender and land rights in South Asia. Cambridge: Cambridge
University Press.
Agarwal, B. (2003). Gender and land rights revisited: Exploring new prospects via the state, family and
market. Journal of Agrarian Change, 3(1-2), 184-224.
Chimhowu, A. & Woodhouse, P. (2006). Customary vs. private property rights? Dynamics and
trajectories of vernacular land markets in Sub-Saharan Africa. Journal of Agrarian Change, 6(3),
346-371.
Dancer, H. (2015). Women, Land and Justice in Tanzania. Suffolk: James Currey.
Daley, E. (2008). Gender, uenyeji, wealth, confidence, and land in Kinyanambo: The impact of
commoditization, rural-urban change and land registration in Mufindi District, Tanzania. In B.
Englert & E. Daley (Eds.) Women’s land rights and privatization in Eastern Africa (pp. 61-82).
Suffolk: James Currey.
De Weerdt, J. (2010). Moving out of poverty in Tanzania: Evidence from Kagera. Journal of
Development Studies, 46(2), 331–349.
Gray, L. & Kevane, M. (1999). Diminished access, diverted exclusion: Women and land tenure in subSaharan Africa. African Studies Review, 42(2), 15-39.
Lastarria-Cornhiel, S. (1997). Impact of privatization on gender and property rights in Africa. World
Development, 25(8), 1317-33.
Jayne, T. S., Yamano, T., Weber, M. T., Tschirley, D., Benfica, R., Chapoto, A., & Zulu, B. (2003).
Smallholder income and land distribution in Africa: Implications for poverty reduction strategies.
Food Policy, 28(3), 253-275.
Mackenzie, F. (1990). Gender and land rights in Murang’a district, Kenya. Journal of Peasant Studies,
17(4), 609-643.
Panda, P. & Agarwal, B. (2005). Marital violence, human development and women’s property status in
India. World Development, 33(5), 823-850.
Patton, M. Q. (2015). Qualitative research and evaluation methods, 4th edition. Newbury Park: Sage
Publications.
Pedersen, R. H. (2015). A less gendered access to land? The impact of Tanzania’s new wave of land
reform. Development Policy Review, 33(4), 415-432.
Razavi, S. (2007). Liberalization and the debates on women's access to land. Third World Quarterly,
28(8), 1479-1500.
Razavi, S. (2009). Engendering the political economy of agrarian change. Journal of Peasant Studies,
36(1), 197–226.
17
Sitko, N. (2010). Fractured governance and local frictions: The exclusionary nature of a clandestine land
market in southern Zambia. Africa, 80(1), 36-55.
Tripp, A. M. (2004). Women’s movements, customary law, and land rights in Africa: The case of
Uganda. African Studies Quarterly, 7(4), 1-19.
Troutt, E. (1994). Rural African land markets and access to agricultural land: the central region of
Uganda. Land Tenure Center, College of Agricultural and Life Sciences, University of Wisconsin Madison.
Tsikata, D. (2003). Securing women’s interests within land tenure reforms: Recent debates in Tanzania.
Journal of Agrarian Change, 3(1-2), 149-183.
World Bank, FAO, and IFAD. (2008). Gender issues in land policy and administration. In Gender in
Agriculture Sourcebook (pp. 125-171). Washington, D.C.: World Bank Publications.
Wineman, A. & Liverpool-Tasie, L. S. (2015a). All in the family: Bequest motives in rural Tanzania.
Mimeo, Michigan State University: East Lansing.
Wineman, A. & Liverpool-Tasie, L. S. (2015b). Land markets and migration trends in Tanzania: A
qualitative-quantitative analysis. Mimeo, Michigan State University: East Lansing.
Whitehead, A. & Tsikata, D. (2003). Policy discourses on women’s land rights in sub-Sharan Africa: The
implications of the re-turn to the customary. In Razavi, S. (Ed.) Agrarian change, gender and land
rights. Oxford: Blackwell Publishing.
Yngstrom, I. (2002). Women, wives and land rights in Africa: Situating gender beyond the household in
the debate over land policy and changing tenure systems. Oxford Development Studies, 30(1), 21-40.
18
Tables
Table 1. Characteristics of male- and female-headed households
(1)
Male-headed
households
Mean
SD
Number of working-age adults (age 15-59)
2.32
(1.12)
Proportion dependents (<15 or >59 years)
0.52
(0.22)
Head's age (years)
41.79 (14.78)
1=HH member completed primary school
0.74
(0.44)
1=Has non-agricultural income
0.13
(0.34)
1=Iron roof
0.72
(0.45)
Value of assets (100,000s TSh)
37.80 (107.81)
Land owned (acres)
4.39
(5.61)
Land rented/ borrowed (acres)
0.39
(1.47)
Land accessed per capita (acres)
1.07
(1.49)
Number of agricultural parcels
2.34
(1.25)
1=HH rents/ borrows land
0.18
(0.39)
1=HH has sold land in past 6 years
0.12
(0.33)
1=HH has bought land in past 6 years
0.30
(0.46)
1=HH has sales contract
0.40
(0.49)
1=HH head has sales rights to any plot a
0.67
(0.47)
Observations
668
(2)
Female-headed
households
Mean
SD
1.55
(1.42)
0.59
(0.31)
56.90
(15.26)
0.52
(0.50)
0.11
(0.32)
0.78
(0.41)
24.45
(58.37)
2.58
(2.54)
0.18
(0.73)
1.01
(1.02)
1.73
(0.80)
0.11
(0.32)
0.06
(0.23)
0.16
(0.37)
0.21
(0.41)
0.54
(0.50)
629
Test
(1) = (2)
***
***
***
***
**
*
***
***
***
***
**
***
***
***
Note: Asterisks denote significance levels of t-test for the difference in means. *** p<0.01, ** p<0.05, * p<0.1
a
This information is only available for land-owning households in which the head was interviewed (n=1,159).
Source: CBLA Evaluation Survey (2014)
Table 2. Characteristics of female-headed households that have independently purchased land
(1)
(2)
Have
Have not
purchased land
purchased land
Test
Mean
SD
Mean
SD
(1) = (2)
1.70
(1.43)
1.52
(1.43)
Number of working-age adults (age 15-59)
0.54
(0.31)
0.62
(0.30)
Proportion dependents (<15 or >59 years)
*
1=Head is widowed
0.59
(0.49)
0.77
(0.42)
***
15.63
(9.87)
11.60
(9.29)
Years since marriage ended a
***
54.46
(12.30)
58.39
(15.43)
Head's age (years)
***
0.53
(0.50)
0.50
(0.50)
1=HH member completed primary school
1=Has non-agricultural income
0.12
(0.33)
0.09
(0.28)
1=Iron roof
0.74
(0.44)
0.79
(0.41)
15.45
(31.84)
23.70
(57.33)
Value of assets (100,000s TSh)
2.99
(3.22)
2.52
(2.34)
Land owned (acres)
Land accessed per capita (acres)
1.06
(0.50)
1.20
(0.78)
a
2.13
(2.46)
0.00
-Land purchased since marriage ended (acres)
N/A
0.14
(0.49)
0.16
(0.69)
Land rented/ borrowed (acres)
0.58
(1.57)
2.01
(2.34)
Land retained from time of marriage (acres)a
***
Observations
145
473
Note: Asterisks denote significance levels of t-test for the difference in means. *** p<0.01, ** p<0.05, * p<0.1
a
16 women in column 2 were unable to report the year their marriage had ended. For these variables, n = 457.
Source: CBLA Evaluation Survey (2014)
19
Table 3. Determinants of land market behavior (2008-2014) (seemingly unrelated multivariate probit)
(1)
(2)
(3)
Currently
Has purchased
Has sold land
rents/ borrows
land in past 6
in past 6
land
years
years
-0.35**
(0.14)
1=Female-headed household
-0.25**
(0.12)
-0.02
(0.02)
-0.00
(0.00)
-0.20
(0.15)
0.03
(0.06)
0.81***
(0.13)
0.24
(0.21)
-0.62***
(0.15)
0.09**
(0.04)
-0.00**
(0.00)
-0.13
(0.17)
0.01
(0.06)
0.27*
(0.15)
-0.16
(0.23)
-0.23
(0.21)
0.02
(0.07)
-0.19
(0.16)
-0.00
(0.02)
-0.31*
(0.18)
-0.24
(0.40)
0.03
(0.15)
-0.02
(0.12)
-0.13
(1.59)
-0.04*
(0.02)
0.21
(0.17)
0.11**
(0.05)
0.03
(0.14)
-0.00
(0.03)
0.17*
(0.09)
-0.17
(0.35)
-0.17
(0.11)
-0.15
(0.12)
0.64
(1.55)
0.03*
(0.01)
-0.04
(0.19)
-0.02
(0.06)
-0.22
(0.17)
-0.02
(0.03)
0.03
(0.12)
-0.32
(0.48)
0.08
(0.14)
0.22*
(0.11)
-5.87***
(1.75)
1,297
1,297
1,297
1=HH has been female-headed for past 6 years
Age of head
0.01
(0.02)
-0.00
(0.00)
0.06
(0.21)
-0.02
(0.07)
0.38***
(0.14)
-0.10
(0.28)
-0.24***
(0.04)
Age2 of head
HH member has completed primary school
No. working-age adults
1=Migrant
1=Has non-agricultural income
Land owned by household 1 year ago (acres)
Land owned by household 6 years ago (acres)
1=HH dwelling has iron roof
Value non-land assets (log)
1=HH is in Karagwe
Village population density (100's people/ km2)
Time to road (hours)
Time to phone (hours)
1=Land is available in village to be allocated
Village median land value per acre (log)
Constant
Observations
Standard errors in parentheses, clustered by village, *** p<0.01, ** p<0.05, * p<0.1
Rho (1 & 2): -0.18** (0.09); rho (1 & 3): -0.26 (0.14); rho (2 & 3): 0.14 (0.09)
Likelihood ratio test that all rhos = 0: 𝜒 2 = 197,348, P > = 0.0000
Source: CBLA Evaluation Survey (2014)
20
Table 4. Determinants of land market behavior among female-headed households (seemingly unrelated
bivariate probit)
(1)
(2)
(3)
(4)
Has purchased
Rents/
Has purchased
Rents/
land a
Borrows land
land a
Borrows land
Land retained from before
marriage ended (acres)
-0.27***
(0.07)
-0.26**
(0.10)
Land inherited by head, since
marriage ended (acres)
No. years head has been
widowed, separated, or divorced
HH demographic/ wealth controls
and migrant status
Village controls
Rho
Observations
-0.30***
(0.07)
-0.27***
(0.11)
-0.32**
(0.14)
-0.04
(0.09)
0.04***
(0.01)
0.004
(0.01)
0.04***
(0.01)
0.004
(0.01)
Y
Y
Y
Y
Y
Y
Y
Y
602
-0.25*
(0.15)
602
602
-0.23
(0.14)
602
Standard errors in parentheses, clustered by village, *** p<0.01, ** p<0.05, * p<0.1
a
Has purchased land since marriage ended
Source: CBLA Evaluation Survey (2014)
21
Table 5. Determinants of land acquisition by female-headed households (tobit)
Land area purchased
since marriage ended (acres)
(1)
(2)
(3)
Widowed
Separated
All FHHs
FHHs
FHHs
Land retained from before marriage ended (acres)
Land inherited by head, since marriage ended (acres)
No. years HH has been widowed, separated, or divorced
HH demographic/ wealth controls and migrant status
Village controls
P > F (land retained from marriage = land since inherited)
P > F (land retained = -1)
P > F (land inherited = -1)
Land area rented/ borrowed (acres)
(4)
(5)
(6)
Widowed Separated
All FHHs
FHHs
FHHs
-0.98***
(0.30)
-1.01**
(0.44)
0.10***
(0.02)
-0.78***
(0.21)
-0.87**
(0.39)
0.12***
(0.03)
-1.78**
(0.74)
-0.83
(0.65)
0.04
(0.06)
-0.70**
(0.28)
-0.14
(0.24)
0.01
(0.03)
-0.29
(0.24)
0.22
(0.25)
0.04
(0.04)
-2.85***
(0.87)
-0.75
(0.56)
0.02
(0.04)
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
0.95
0.95
0.988
0.79
0.29
0.74
0.22
0.30
0.80
0.06
0.28
0.000
0.09
0.003
0.000
0.01
0.04
0.65
156
57
602
79
446
33
156
46
Observations
602
446
Uncensored observations
145
88
Standard errors in parentheses, clustered by village, *** p<0.01, ** p<0.05, * p<0.1
Source: CBLA Evaluation Survey (2014)
22
Table 6. Villages included in qualitative data collection
Study sites
Village description
Karagwe District
Katembe
Public transport readily available
Chabalisa
Difficult to find public transport
Biharamulo District
Town characteristics
Nyakanazi
(daily market, crowded)
Nyabugombe
Remote, no phone reception
Population
(no. households)
Travel time from district
headquarters (hours)
881
660
0.5
2
2,272
1.5
633
2.5
Sources: CBLA Evaluation Survey (2014) and authors’ observations
Table 7. Characteristics of individual respondents (qualitative study)
Age
No.
Marital status
< 35
3
Widowed
35 – 70
13
Divorced/ separated
≥ 70
4
Married, but functioning as head
Never married
Total:
20
Source: Authors’ calculations
23
No.
8
8
3
1
20
Figures
Figure 1. Conceptual framework of female heads’ land access
Source: Authors’ observations
Figure 2. Study site
Source: Wikimedia commons
Figure 3. Relationship with individual/ entity from whom plot was acquired
Proportion of plots
1.0
n=856
462
128
104
MHHs
FHHs
MHHs
FHHs
0.8
0.6
0.4
0.2
0.0
Purchased
Rented/ Borrowed
Man's family
Woman's family
Friends/ Neighbors
Government
Clan/ Traditional leader
No relation
Source: CBLA Evaluation Survey (2014)
24