gender land and asset survey uganda - icrw

UGANDAGender Land and Asset Survey

FundersAn Anonymous Donor

PartnersAssociates Research Uganda, LimitedUniversity of KwaZulu-Natal

The International Center for Research on Women (ICRW) gratefully acknowledges the contributions of Hema Swaminathan,Margaret Rugadya, Herbert Kamusiime, Christine Kajumba, Solomon Kabanda, Charles Mukasa, Lawrence Lubyayi, Fred Kalema, Urmilla Bob, Vadi Moodley, Nitesh Poona, Zola Dyasi , Dorah Ncobo, Jyoti Jaggernath, Sazile Mtshali, Innocent Shezi, Humayrah Bassa, Timothy Wiggill, Ms ZaakirahBassa , Roshini Bob, Juan Muñoz, and Rodrigo Muñoz.

Special thanks to the community members who participated in the evaluation research.

Several colleagues at ICRW provided tremendous input into the research: Meredith Saggers, Gwennan Hollingworth, Rekha Mehra,Mary Ellsberg, Ellen Weiss, and Sandy Won. We also thank our editor, Margo Young.

© 2011 International Center for Research on Women (ICRW). Portions of this report may be reproduced without express permission from but with acknowledgment to ICRW.

Cover photo by Krista JacobsDesign: Barbieri & Green

Acknowledgements

Gender Differences in Asset Rights in Central Uganda

Aslihan KesKrista JacobsSophie Namy

Though data showing the importance ofasset ownership for women are provingvaluable in raising awareness about theneed to recognize and enforce women’srights, awareness-raising is only a firststep. Deeper analysis is needed to un-derstand individuals’ engagement withassets beyond ownership and into useand decision-making, how women’s en-gagement with assets differs frommen’s, and why asset rights are impor-tant. This entails building an under-standing of the factors that may becorrelated with women’s asset rights. Amore specific understanding of thesefactors in turn will allow movement fromawareness to action by providing thedata needed to guide programs and pol-icy decisions that will ensure theserights are upheld.

To provide some of this more in-depthdetail on determinants of women’s assetrights, ICRW and Associates ResearchUganda Limited. developed and piloted asurvey methodology for collecting andanalyzing individual- and household-level quantitative data on women’srights over assets.1,2 The Gender, Land,and Asset Survey (GLAS) is one of thefirst studies to undertake a quantitativeand gendered assessment of men’s andwomen’s rights over assets – includingownership, documentation, and degreeof control over use, transfer, and trans-actions – and the implications thereof.Specifically, this study attempts to an-swer the following questions in the con-text of central Uganda:

• What are the differences in women’sand men’s ownership, use, and deci-sion-making over land, housing, mate-rial assets, livestock, and financialassets?

• Which socioeconomic/structural fac-tors influence women’s and men’sasset rights and in what ways?

Given the evidence from other contextssuggesting a link between women’srights over assets and domestic violence,the study also examines whetherwomen’s asset rights influence their ex-perience of intimate partner violence incentral Uganda.

Background For individuals and households, assetownership translates to a secure placeto live, means to earn a livelihood, andthe ability to mitigate the economic andsocial risks associated with natural dis-asters, disease, and economic shocks(Doss, Grown, and Deere 2008).

A growing literature demonstrates theparticular benefits of women’s assetownership, not only for themselves, butalso for their families and the economyas a whole. In various studies, women’sasset ownership has been linked to in-creased spending on food, housing, anddurable goods, and children’s schooling(Katz and Chamorro 2003; Quisumbingand Maluccio 2003; Doss 2006). Duflo(2000) and Thomas (1990) also found thatasset ownership by mothers can im-prove children’s health outcomes, in-cluding survival rates, and long-termand short-term nutritional status.Women with more assets are also foundto use prenatal care at higher rates thanwomen with fewer assets (Beegle,Frankenberg, and Thomas 2001).

Besides enhancing the well-being ofwomen and their families, asset owner-ship is also found to empower women intheir relationships and to give them astronger voice in public forums (Katzand Chamorro 2003). A number of stud-ies exploring the pathways amongwomen’s asset ownership, socioeco-nomic status, and HIV risk suggest thatasset inequality increases women’s vul-nerability to the HIV infection (Beegle

and Ozler 2006; Hallman 2005; Swami-nathan, Walker, and Rugadya 2006).Other studies have established that as-sets may protect women from experi-encing domestic violence (Panda andAgarwal 2005; Bhatla, Chakraborty, andDuvvury 2006; Swaminathan, Walker,and Rugadya 2006).

At the macro level, gender equality inasset ownership is shown to improveagricultural productivity, bolster resist-ance to economic shocks, and fostereconomic growth (Deere and Doss 2006).

Current Knowledge on Gender and Asset Rights in Uganda The literature on men’s and women’sasset rights in sub-Saharan Africa isquite limited and for the most part fo-cused on land. Evidence from thesestudies points to a substantial and per-vasive gender gap in asset ownership,with women owning less land than menthat is of lower quality (Doss 2006;Mason and Carlsson 2004; SOFA Team2011). Data on assets beyond land areeven scarcer, though general trendsagain suggest a strong advantageamong men in terms of assets such asfarm animals and transport vehicles(Doss, Grown, and Deere 2008).

Evidence from Uganda is similar to whathas been found elsewhere in the region.Despite their significant role in the agri-cultural sector,3 only about 16 percent ofUgandan women own land in their ownright (Rugadya 2010). Their ownership ofregistered land is even lower at 7 percent(Rugadya 2010; Bikaako and Ssenkumba2003). Housing, often considered a com-bined asset with land, particularly in ruralareas, is also overwhelmingly owned bymen (Rugadya 2010).

About half of the farming households inUganda are engaged in livestock rear-ing (Nayenga 2008), which providesthem an additional source of livelihood.Differences also exist in ownership oflivestock between men and women aswell as in the types of livestock they

2

ecent studies have affirmed the importance of women’s

asset ownership, including land, housing, and other

assets, for economic development and social security. R

own. Oluka et al.’s study of men’s andwomen’s livestock ownership in north-eastern Uganda finds a significant gen-der gap. The disparity is most severe incattle ownership with 62 percent ofmen and only 14 percent of women re-porting ownership, but also extends topoultry and other small stock animals.Only 22 percent of women versus 39percent of men indicate they own poul-try, for instance, which is surprisinggiven that it is often seen as a femaleasset (Oluka et al. 2005).

Land, by virtue of being the main produc-tive asset and means of wealth accumula-tion for a majority of Ugandans, dominatesthe discourse on asset rights in the coun-try. Evidence on gender inequities in assetownership, particularly land, has in recentyears helped move an extensive legal re-form process to define and protectwomen’s right to own, use, and transferland and other assets. This includes theratification of a new Constitution and sev-eral pivotal pieces of legislation such asthe 1998 Land Act and the pending reformof the Succession Act. The draft NationalLand Policy further prescribes specificstrategies to ensure that constitutionalrights and obligations with regard to gen-der equality are upheld, particularlyaround reforming customary laws andpractices (Rugadya 2010).

A fairly large literature is dedicated tounderstanding the underpinnings of per-sistent gender inequity in land rights, inUganda and elsewhere in sub-SaharanAfrica. Commonly examined in this liter-ature are customary laws and practicesthat continue to shape many individuals’relationship with land. Molded by soci-ety’s patriarchal norms, these laws andpractices give ownership of land to menor male heads of extended families,while women enjoy “secondary” rightsin the form of access to and use of landthrough their husbands, fathers, broth-ers, or other male relatives (Bikaakoand Ssenkumba 2003; Benschop 2002;Rugadya 2010).

The implications for Ugandan women ofthe primacy of customary law is thattheir rights over land and housing de-pend significantly on the quality of theirpersonal and social relationships, prima-rily with their intimate partners and to alesser extent with other kinship groups(Walker 2002). A number of factors, in-cluding the weakening of the institutionof marriage and increased scarcity ofland, have contributed to the deteriora-tion of women’s rights under the custom-ary framework (Tripp 2004; Khadiagala2002). Like most other countries heavilyhit by the AIDS epidemic, in Uganda, HIVhas further contributed to increased vul-

nerability of women’s land rights (Drimie2002; Izumi 2006). Recent studies revealthe extent of property grabbing in HIV-af-fected households; with about 30 percentof widows and orphans the victims ofthis practice following the death of aspouse/parent (Ampaire et al. 2008).

The limitations of the statutory legalframework that includes gender equalityprinciples have also been discussed inthe literature (Cooper 2010). AlthoughUganda’s statutory laws grant men andwomen equal rights to land and otherproperty regardless of their marital sta-tus, application of these laws has beenmired by a number of factors. First, legalpluralism that stems from the recognitionof both customary and statutory landtenure in the Constitution and the LandAct results in multiple conflicts betweencustomary and statutory laws, which areoften not mediated by statutory provi-sions as prescribed in the Constitution.The application of statutory laws is fur-ther complicated by a number of factors,including lack of knowledge of the lawsand the high costs of legal action to ad-minister an estate (Ampaire et al. 2008).The weak institutional capacity of thelegal system, particularly in rural areas,also prevents effective enforcement andadministration of the laws (Whiteheadand Tsikata 2003; Young 2010).

3

Krista Jacobs

The confluence of these factors resultsin customary law and practices stillprevailing over formal law amongUganda’s four predominant tenure sys-tems – mailo, customary, freehold, andleasehold.4,5 Practices around maritalproperty and inheritance, in particular,continue to be dominated by customsthat impose serious constraints onwomen’s ability to own and controlland (Rugadya 2010).

MethodologyInformed by qualitative research con-ducted in the study areas, GLAS offerstwo main methodological innovations.First, it asks not only about ownershipbut also about use and decision-mak-ing over assets. Second, it allows fordisaggregation of data by gender inasking each woman and man sepa-rately about her/his own ownership,use, decision-making, and docu-mented claims over particular assets.Specifically, for women and men, theGLAS captures:

• Documentation beyond land title: TheGLAS asks about multiple forms ofdocumentation beyond a land title orcertificate of registration, includingpurchase agreement, rental agree-ment, receipts, wills, and written per-mission from traditional authorities.

• Joint ownership: The GLAS data de-scribe the extent of joint ownership ofassets such as land, housing, mate-rial assets,6 livestock, and financialassets; the proportion of women’sjoint property ownership to their over-all asset holdings; and with whomwomen and men jointly own assets.GLAS data also evaluate whetherboth respondents believe that theyshare ownership. For material and fi-nancial assets, respondents wereasked if they own the assets by them-selves or with someone else.TheGLAS also allows more than one per-son to be recorded in cases wheretwo (or more) names are on land andhousing documents.

Specific Measures of Asset RightsThe GLAS measures a spectrum ofrights over several types of assets.This report includes findings on land,housing, livestock, material assets,and financial assets. Depending onthe asset, GLAS measures rights in-cluding (1) ownership (whether a re-spondent states that an asset belongsto him/her), (2) documented rights(whether a person’s name appears ondocumentation for the asset), (3) userights, and (4) role in making decisionsabout the asset. Within ownership, wealso distinguish between sole andjoint property ownership. Table 1gives more detail on the primarymeasures of asset rights used in theanalysis. All measures in Table 1 areat the individual level and are for bothwomen and men.

It is important to note, especially in thecontext of joint ownership, that themeasure of ownership used is per-ceived ownership (i.e., based on whatthe respondent reports versus based onlegal proof of ownership). Perceptionsof who owns what differ across respon-dents and may be influenced by normsfavoring communal ownership or con-solidating ownership with men. Peoplemay be more inclined to report assetsthat many household members use,

such as furniture, as belonging to thehousehold as a whole or belonging to acouple. Additionally, anecdotal evi-dence suggests that women may bemore inclined to report property as be-longing to others in the household (tothe children or their partner) than tothemselves.

Sampling The project team in Uganda pilotedthe GLAS survey in Butenga sub-county, in Masaka District. Across atotal of six parishes, the total popula-tion of Butenga subcounty is 45,148.The site selection was intended torepresent the mailo land tenure sys-tem in rural central Uganda. A multi-stage sampling approach was usedthat drew on a district-level “commu-nity information survey,” which listsall parishes and villages in the district.First, two parishes were randomly se-lected; this was followed by a randomselection of four villages in eachparish. Households were randomly se-lected from each village, with thenumber of households selected pro-portional to the village’s share of theparish population. The total samplesize was 545. Two respondents wereinterviewed per household – thehousehold head (either male or fe-male) and a randomly chosen woman.In many instances, female-headedhouseholds did not have another eligi-ble woman to be the randomly chosenwoman respondent. As a result, thefinal sample included a total of 698 in-dividual interviews. The analysis wasbased on data collected from 539households and 674 individuals, 345men and 329 women. Of these menand women interviewed, 344 men and117 women were heads of theirhouseholds. All but one of the womeninterviewed who were not heads ofhouseholds were spouses of men whowere interviewed.

Analytical ApproachesThis report presents findings from threeanalytical methods: (1) summary statis-tics, (2) profile analysis, and (3) multi-variate analysis.

4

Meredith Saggers

The summary statistics provide snap-shots of what assets individuals have,use, and control. In comparing individ-ual-level means and tabulations of vari-ables by sex and women’s headship(see Table 1) this analysis describes anydifferences in overall rights over assetsbetween women and men.

Profile analysis is used to examine thecharacteristics of women with differentlevels of asset rights. Using indices thatcondense ownership, self-ownership, de-cision-making, and use into scores of low,medium, and high rights over land, hous-ing, material assets, and livestock, womenwere grouped according to their rightsscore.7 This grouping allows for comparingthe main attributes, or profiles, of womenwith different rights scores to traits suchas age, education, marital status, etc.

Finally, multivariate linear regression8 isused to disentangle the influences of mul-tiple socioeconomic factors on women’sand men’s property rights. Economic the-ory and the literature9 on asset rights andgender suggest several factors are re-lated to a person’s asset rights. Individual-level determinants include age, literacyand education, partnership status, rela-tionship to the household head, havingsons and daughters, proximity to natalfamily, main livelihood, whether earningcash income, and the prevailing tenuresystems. Household-level variables cap-ture household composition (number ofadult men and adult women), proxies forhousehold socioeconomic status (numberof persons per room, quality of the physi-cal dwelling), and how the household ac-quired its lands. The models includecommunity-level fixed effects.

A binary variable for whether an individ-ual in the sample is female is included inthe multivariate estimations to capturethe gender differentiated asset rightsoutcomes. This variable is also inter-acted with several other potential deter-minants (partnership status, headshipstatus, prevailing tenure systems, howland was acquired, numbers of broth-ers, sons, and males in the household)to investigate whether these variableshave the same relationship to women’sasset rights as they do to men’s. For ex-ample, the multivariate regressions canspeak to questions like, “Is being in anindividual customary land tenure systemequally beneficial for women’s andmen’s land ownership?”

5

Table 1. Rights Measured by Gender: Land, House, Material Assets, Livestock and Financial Assets

LAND

Own Whether a person owns land either jointly or by him/herself

Document Whether a person reports having his/her name on any written documentation for land, including titles, rental agreements, receipts, permission to occupy orders, etc.

Decision-making A measure of individual’s decision-making over each plot of land associated with the household including land transactions (selling, renting, and collateralizing), who will inherit the land, who may use land, what to grow on land, selling the harvest, and keeping money from harvest sales

HOUSE

Own Whether a person owns a residence either jointly or by him/herself

Own (self) Whether a person owns a residence by him/herself

Decision-making A person’s decision-making power over transactions on the residence (sell, gift, rent, collateralize, and bequeath)

MATERIAL ASSETS

Own The number of different asset types that a person owns either jointly or by him/herself

Own (self) The number of different asset types that a person owns by him/herself

Use (total) The number of different asset types a person uses regardless of who owns them

LIVESTOCK

Own Whether a person owns poultry either jointly or by him/herself

Own (self) Whether a person owns poultry by him/herself

Decision-making (total) A person’s decision-making over use of livestock owned by all household members, including self

FINANCIAL ASSETS

Cash (self) Whether a person reports having cash

6

Characteristics of Study Location and SampleThere are some notable differences inthe sample in terms of the characteris-tics of men, women heads of house-holds, and women who are not heads(see Table 2). While men and womenare on average of similar age, womenheads of households are significantlyolder than women who are not heads ofhouseholds. Although there is a slightdifference in the overall sample be-tween men’s and women’s educationlevel, a more significant disparity ariseswhen comparing women who areheads of households and those whoare not, with women household headsbeing significantly less likely to be edu-cated or literate.

Among women who are not heads ofhouseholds, monogamous customarymarriage is most prevalent (around 33percent), followed by monogamous reli-gious marriages (30 percent) and cohabi-tating (28 percent). Meanwhile, mostwomen heads of households are wid-owed (59 percent). Approximately 28 per-cent are divorced or separated, and afew (6 percent each) are married or sin-gle. Polygamous marriage is not common,constituting about 2-4 percent of all re-ported marriages. This is significantlylower than anticipated based on the De-mographic and Health Survey data fromthe region, which suggests that about 24percent of women are in polygamous re-lationships (Uganda Bureau of Statisticsand Macro International Inc. 2006). The south-central region, whereMasaka is located, is among the moreaffluent parts of Uganda, with more thanone-third of its population in the highesttwo wealth quintiles nationwide

(Uganda Bureau of Statistics and MacroInternational Inc. 2006). The region alsofares better than the rest of the countryin terms of education. Women in particu-lar have higher school attainment andliteracy rates than those in other parts ofthe country. Although lower than the na-tional average, the fertility rate in the re-gion is high at 5.6 percent, adding to thepopulation pressure in this alreadydensely populated region. Masaka isalso among the regions nationwide withthe highest HIV prevalence, and womenare affected more than men (Govern-ment of Uganda 2010).

Agriculture is men’s and women’s mainsource of livelihood in Masaka(Uganda Bureau of Statistics andMacro International Inc. 2006). This isreflected in the study sample, whichshows agricultural self-employment asthe main source of cash income for 84percent of women and the highestranked activity in terms of time spent.Men appear slightly more diversifiedwith 69 percent earning an incomethrough agriculture. Much like the restof the country, land is therefore the keyproductive asset and means of wealthaccumulation for a majority of thosewho live in the region.

Though several land tenure systems co-exist in Uganda, mailo is the predomi-nant tenure in Masaka.10 Transactionsunder this tenure are governed by statu-tory law, which only accords recogni-tion of absolute rights to the individualin whose name the title is issued. Allother powers relating to access, con-trol, and use are governed by the cus-tomary law of Buganda (Rugadya 2007).In our sample, about 70 percent of men

and 54 percent of women reportedbeing associated with mailo tenure, followed by historical occupation re-ported by 15 and 11 percent, respec-tively (see Table 3).11 Other tenureforms, including freehold, leasehold,rent, individual customary, squatting,and informal tenure were reported by10 percent of the respondents. It is im-portant to note that about one-quarterof all respondents, most of themwomen who are not heads of house-holds, indicated that they don’t knowtheir tenure types. Specifically, 50 per-cent of women who live in householdsheaded by men do not know theirtenure types. This may be due to thepatrilocal marriage patterns inUganda, which result in women mov-ing to their partners’ clans and awayfrom the tenure regimes of their natalfamilies, likely the only regime withwhich they are familiar

Table 2. Select Demographic Characteristics of Respondents, by Sex and Headship

Men (n=345) Women (n=329)

All men All women Women, head of Women, not head of household (n=117) household (n= 212)

Average age in years 43.9 41.2 55.1 36.2Read and write in any language (%) 79 69 49 76Ever been to school (%) 87 82 67 87

Meredith Saggers

7

Asset Rights of Men andWomen in Masaka District

Land and HousingAs presented in Table 4, in Masaka,men own land and houses at higherrates than women. When accountingfor both individual and joint ownership,nearly three times as many men aswomen own land (88 percent vs. 32percent). Meanwhile, individual owner-ship for land is 43 percent among menand 17 percent among women. Interms of house ownership, there is asimilar gender divide, with more thanhalf of men owning a house as op-posed to about one-third of women(see Table 5). Joint ownership is signif-icantly more prevalent among women,with 15 percent of women indicatingthat they own a house jointly as op-posed to 3 percent of men.

Worth noting is women heads of house-holds’ significantly higher levels of landand house ownership in comparison towomen who live in households headedby men. In fact, women heads of house-holds’ house ownership as well as theirindividual ownership of land is at compa-

rable levels with those of men. Womenwho are not heading a household reportsignificantly higher levels of joint houseownership than do men overall.

Around 60 percent of all householdshave some kind of land document,most commonly a purchase agreement,on at least one of their plots. Regard-less of their position in the household,a significantly lower number of women,13 percent, have their names on theseland documents compared to 48 per-cent of men.

While no significant difference is de-tected between men and women’s re-porting of having their names on thesedocuments, the subgroup of womenwho are in male-headed households ap-pears significantly less represented inhouse documents. Overall, significantlyfewer households have documentationfor housing (15 percent) than for land(60 percent).

Inheritance, purchase, and transfersfrom living family members emerge asthe most common forms of land acqui-sition in the region (see Table 6). Asked

how the lands they are associated withwere acquired, men and women gaveslightly different responses, with a ma-jority of men indicating they are associ-ated with inherited land, while themajority of women report purchase.12

A significantly higher number of womenheads of households also report theyare associated with land acquiredthrough marriage, as compared towomen who are not household headsand men. This finding is consistent witha high percentage of female householdheads being widowed, divorced, orseparated.

The respondents were asked about theirability to make decisions regarding landand house, including transaction deci-sions (sale, renting, gifting, collateraliz-ing), decisions regarding their use (cropsgrown and sold) and use of income theygenerate, and bequeathing. The ability tomake an independent or joint decisionwas scored as 1, and lack of control overa decision was scored as 0. Based on thisscoring, across all decisions – transac-tions, use, use of income, and bequeath-ing – men emerge as holding a relativelystronger position with respect to making

Table 3. Tenure Types (%), by Gender, by Sex and Headship


All men All women Women, head of Women, not headhousehold (n=117) of household (n= 212)

Mailo 69 54 59 52Historical occupation 15 11 9 12Other 10 8 13 6Unknown to respondent 15 37 23 42

Table 4. Ownership of Land (%), by Sex and Headship



Person owns land 88 33 75 17Person owns land – by self 43 17 38 9Person owns land – jointly 53 20 41 13Person’s name on document for land 48 13 30 7

decisions regarding land and housetransactions (see Figure 1), as well as useof house and land. A more nuanced pic-ture emerges, however, when differentgroups of women and particular deci-sions are taken into account. Specifically,though men have a clear upper hand onland transactions, including selling, gift-ing, and collateralizing land as well as be-queathing decisions, women heads ofhouseholds appear to be better posi-tioned than both men and women whoare not heads in decisions on agriculturaluse of land. Both men and women headsof households report similar levels ofcontrol over income from land. With re-spect to house-related transactions,women who are heads of householdsonce again have a very similar level ofcontrol over these decisions compared tomen, while women who are not headsare significantly disadvantaged.

Material AssetsData were also collected on men’s andwomen’s ownership and use of variousmaterial assets including agriculturaltools and machinery, household goods(both durable and nondurable), trans-portation assets, and communicationassets (see Table 7 for findings pertain-ing to ownership).

Once again, men emerge with a signifi-cant advantage in material asset owner-ship, particularly in individual ownershipof key productive assets such as agri-cultural tools and machinery, transporta-tion assets, and communication assets. The gender divide is somewhat miti-

gated when factoring in headship sta-tus of women. In fact, there is no no-table difference between men overalland women household heads in theownership of agricultural and commu-nication assets, though men have aclear advantage in ownership of trans-port assets. Women in householdsheaded by men, on the other hand, re-port significantly higher levels of jointownership of some of these assets,specifically agricultural machinery andtools and communication assets. Thesewomen also own jewelry at signifi-cantly higher rates than both men andwomen household heads.

Similarly, the level of diversification interms of different types of material as-sets individually owned is indistinguish-able between men and women heads ofhouseholds, while women who are nothousehold heads own a significantly lessvaried basket of material assets. How-ever, ownership seems to be less of afactor in use of material assets as thenumber of material asset types used bymen and women is the same. Both menand women use, on average, 4.6 assettypes across all assets in the householdregardless of the owner.

Livestock Cattle, goats, pigs, and poultry are themost commonly owned livestock amongrespondents. As shown in Table 8, own-ership levels for livestock are higheramong women than men, although thedifference is not statistically significantexcept in the specific case of poultry.

The absence of a gender divide in live-stock ownership is also reflected in com-parable levels of decision-making on useof livestock reported by women and men(not shown in Table 8). On a scale from 0to 1, both men and women have compa-rably strong decision-making controlover use in self-owned livestock with ascore about 0.44. While the absolutescores of decision-making control dropfor jointly owned livestock (0.16 forwomen and 0.05 for men) and livestockowned by others in the household (0.03for women and 0.02 for men), they remaincomparable between men and women.

Financial assetsAll survey respondents have a fairly lim-ited array of financial assets, which con-sist of cash and individual and groupsavings accounts. In terms of cash hold-ings, there is no difference betweenmen and women in general. Womenwho are heads of households are signifi-cantly more cash poor than men and, ifjoint cash holding is considered, alsomore cash poor than women who are inmale-headed households. A significantlylarger number of men hold individualsavings accounts than women, whilethere is no difference in group savingsaccount holdings. There is little or nojoint ownership of financial assets.

Discussion In line with what has been suggested byprevious studies, a significant genderasset gap exists not only in the owner-ship of and decision-making over largerassets such as land and housing, but

8

Table 5. Ownership of House (%), by Sex and Headship


All men All women Women, head of Women, not head household (n=117) of household (n= 212)

Person owns house 57 33 59 24Person owns house – self 54 19 56 5Person owns house – jointly 3 15 3 19Person’s name on document for house* 18 11 21 2

(n=211) (n=118) (n=66) (n=52)

* Data on house documentation comes from the subset of individuals who indicate owning a house. This explains the low n.

9

Table 6. Land Acquisition (%), by Sex and Headship



Land acquired through inheritance 43 40 33 42Land acquired through purchase 36 43 29 49Land acquired through family 19 17 17 18Land acquired through marriage 1 5 18 1

also with regard to key productive mate-rial assets such as agricultural machin-ery and tools and communicationassets. Across all these assets, individ-ual rights predominantly lie with men,while women have some joint ownershipand limited decision-making control.Livestock is the only asset that appearsto be equally owned and controlled.

What appears to be a very stark genderdivide in asset ownership is signifi-cantly mitigated or eliminated whenwomen’s position in the household isconsidered. There appear to be signifi-cant differences in the levels of assetownership between women who areheads of their households and womenwho are members of a household

headed by men. Results from the profileanalysis, which allows for the compari-son of the characteristics of womenwith low, medium, and high rights overland, house, livestock, and material as-sets, support that headship is a signifi-cant factor in determining women’srights over property. Across differentasset profiles described in the Method-ology section, we find that a signifi-cantly higher percentage of womenwith high land, house, and materialasset profiles are heads of their house-holds. What is more, women heads ofhouseholds emerge as having similarlevels of asset ownership and controlas men, with a few notable exceptions.Significant among them is women’slower land documentation, lower ability

to make land transaction decisions, andthat they are more cash poor than men.

Women in male-headed households, onthe other hand, are more asset poor thanmen and women heads of households,particularly in terms of individual owner-ship. Arguably, joint ownership can bethought of as equally valuable to sole own-ership in partnerships, and jointly ownedassets might be expected to form a sizableportion of partnered women’s asset hold-ings. Yet, the pattern of reporting of individ-ual and joint ownership seems to suggestthat in fact a number of assets perceivedas jointly owned by women are consid-ered as individually owned by men. This isparticularly apparent in the reporting ofmaterial asset ownership.

Figure 1. Decision-Making over Land and House, by Gender and Headship, by Sex and Headship

1.0

0.5

0.0Men Women

head of householdWomen

not head of household

Land House

0.53

0.18

0.50

0.14

0.31

0.06

0=no role; 0.5=some input but decision made by other; 1=primary/joint decision-maker

10

Table 7. Ownership of Material Assets (%), by Sex and Headship



Respondent has self-owned transport assets 30 1 1 1Respondent has joint transport assets 1 5 0 7Respondent has self-owned agricultural assets 92 76 94 70Respondent has joint agricultural assets 20 50 2 67Respondent has self-owned communication assets 73 35 59 26Respondent has joint communication assets 12 43 2 58Respondent has self-owned household durables 3 3 0 4Respondent has joint household durables 0 1 0 2Respondent has self-owned household nondurables 70 65 90 57Respondent has joint household nondurables 33 58 5 77Respondent has self-owned jewelry/watches 10 22 12 26Respondent has jointly owned jewelry/watches 0 0 0 0

Table 8. Ownership of Livestock (%), by Sex and Headship



Person owns livestock 54 64 59 66Person owns livestock – self 45 49 57 46Person owns livestock – joint 12 22 2 29

Table 9. Ownership of Financial Assets (%), by Sex and Headship


All men All women Women, head of Women, not head of household (n=117) household (n=212)

Has savings account – self-owned 13 4 5 4Has savings account – joint 0 0 0 0Has cash – self-owned 43 35 25 38Has cash – joint 6 6 0 9Has group savings account – self-owned 6 12 11 12Has group savings account – joint 0 0 0 0

Determinants of AssetRights

Multivariate regression analysis wasused to understand the underlying fac-tors that shape men’s and women’sasset rights and to identify possibledrivers of the gendered nature of theserights. (Figures 2 through 5 present re-sults from the regression analysis, high-lighting the factors that emerged assignificant in shaping asset rights.Please refer to Appendix A for more de-tailed results.)

A number of individual-level factorsemerge as significant determinants ofwomen’s and men’s asset rights. Amongthem, age is positively associated withhigher likelihood of owning land, ahouse, and livestock, with probability ofownership increasing by 3 to 5 percent-age points for each 10 additional years.The likelihood of having cash holdings,on the other hand, appears to fall withage by 6 percentage points for every 10years. Education, measured by whetherindividuals are literate or have ever beento school, has a varied effect on individu-als’ rights over assets, most notably ontheir land rights. The likelihood of havingone’s name in land documents increasesby 10 percentage points if the person hasever been to school. Literacy only slightlyincreases decision-making ability re-garding use and transfer of land, by afactor of 0.06 on a scale of 0 to 1. Mean-while, education is associated withlower cash holdings by individuals. Thereis no effect detected between individu-als’ education status and their houserights or their rights over livestock.

Partnership status and, particularly forwomen, position in the household havesignificant implications for asset rights.First, headship appears to be a signifi-cantly important variable for women.This is captured particularly in land andhouse rights. Female headship is foundto have a large positive association withthe likelihood of land ownership and tohave a negative association with abilityto make transaction decisions overhouse. Second, regardless of their gen-

der, individuals who are divorced, wid-owed, or separated emerge as morelikely to own a house and are signifi-cantly more in control over decisions ofuse and transfer of land. Third, marriageand cohabitation seem to affect men’sand women’s asset rights differently.Women who are married have lowerlikelihood, by 27 percentage points, ofhaving their names on land documentsthan men who are married. Married andcohabitating women also have signifi-cantly less diverse material asset port-folios and use a smaller portion of thetotal household assets than men whoare married or cohabitating. Finally,while marriage increases the likelihoodof livestock ownership for men, it re-duces the likelihood of ownership forwomen by about 12 percentage points.Marriage also weakens women’s abilityto make decisions over use of animals.

Two individual-level variables – numberof brothers and presence of natal familyin the community – help identify the po-tential role of natal family in influencingwomen’s and men’s asset rights. Whatemerges very clearly is a positive asso-ciation between natal family and rights

over livestock. The likelihood of individ-ual livestock ownership is significantlyincreased by community presence ofnatal family members, particularlybrothers. Number of brothers also sig-nificantly increases the use of totalhousehold livestock holdings. For bothmen and women, natal family presencein the community is associated withhigher likelihood of house ownership,although women are less likely thanmen to solely own their own home. Fi-nally, having brothers significantlyweakens women’s ability to make deci-sions over house transactions

At the household level, an individual’slikelihood of land ownership is slightlyreduced with each additional malehousehold member. For house owner-ship, men experience a reduction in thelikelihood of owning with the presenceof an additional male household mem-ber, particularly individual house owner-ship, which is reduced by 16 percentagepoints. For women, the effect of an addi-tional male household member on likeli-hood of house ownership is very small. Tenure systems and forms of land ac-quisition emerge as particularly impor-

11

Land Acquisition & Tenure

Land Acquisition: Family

Land Acquisition: Purchase

Tenure: Historical Occupation

Tenure: Mailo

Individual Characteristics

Married

Divorced, Widowed, Separated

Female Head of House

Female

Weaker (-) Stronger (+) Weaker (-) Stronger (+)

Figure 2. Determinants of Asset Rights: Land

Land Rights

N/A

N/A

Women Mendenotes no effect

Note: Reference categories are for partnership status: single; for tenure: other tenure; and forland acquisition: acquisition through marriage. Regression results should be interpretedin comparison to these categories

12

tant elements in shaping women’s andmen’s land rights, often in significantlydifferent ways. Mailo tenure, the pre-dominant tenure in the region, emergesas favorable to men and women. It isassociated with a 25 percentage pointhigher likelihood of land ownership forboth groups. However, mailo’s positiveassociation with decision-makingpower over use and transactions onland is slightly smaller for women.Somewhat surprisingly, tenure has noeffect on individuals’ ownership and de-cision-making rights over house.

Historical occupation, the second mostprevalent tenure type in the study area,appears to be particularly favorable formen. Being associated with this type oftenure not only increases men’s likeli-hood of land ownership by about one-quarter, it also is associated withslightly stronger control over land deci-sions among men. On the other hand,historical occupation has a negligiblepositive effect on women’s likelihood ofowning land and is associated withslightly diminished decision-making authority among women.

How land was acquired also affects in-dividuals’ rights, chiefly over land butalso over other assets.13 Our results re-veal that individuals who are associ-ated with inherited land have about a16 percentage point higher likelihoodto have their names on land docu-ments. These individuals are also morelikely to own livestock and own individ-ually a slightly more diverse basket ofmaterial assets. For men, purchasedland is associated with an increase ofalmost 50 percentage points in the like-lihood of having their names on landdocuments. For women on the otherhand, the association with purchasedland only slightly increases the likeli-hood, by about 14 percentage points, ofbeing named as a party in the landdocument. Association with purchasedland also increases individuals’ likeli-hood of house ownership and is corre-lated with higher diversification inmaterial assets.


Person’s name on land document




Natal Family in Community

Number of Brothers

Divorced, Widowed, Separated

Female Head of Household


Figure 3. Determinants of Asset Rights: Housing

Housing Rights

N/A





Land Acquisition: Inheritance


Tenure: Historical Occupation

Tenure: Mailo


Number of Brothers

Cohabitating

Married

Female


Figure 4. Determinants of Asset Rights: Material Assets

Material Assets

N/A



13

Surprisingly, association with land ac-quired through family seems to result insomewhat weaker rights over land aswell as house compared to land ac-quired through marriage, particularly formen. Women who are associated withland acquired through family appear tohave significantly stronger rights thanmen in decisions over land use andtransactions as well as house transac-tions. Unfortunately, the GLAS data donot capture whether the land is from theman’s or the woman’s family, but thepredominant practice in Uganda is forland to pass through male lines.

A person having his/her name on docu-mentation for land is associated with in-creases in decision-making over landand housing: an increase of 0.06 on ascale of 0 to 1 for land and 0.24 on ascale of 0 to 2 for housing. The analysiscannot disentangle whether documen-tation led to stronger decision-makingpower or vice versa. It can say thateven accounting for socio demographicand economic factors, land tenure sys-tems, and means of land acquisition,persons with their name on land docu-ments have stronger decision-makingpower over land and housing.

DiscussionAs evident in the findings above, thereare gendered differences in the factorsthat shape individuals’ asset rights,shedding some light on the main driversof the gender asset gap.

Arguably most noteworthy is the emer-gence of women’s relationships and po-sitions in the households as significantfactors in defining their asset rights. Inparticular, headship emerges as a sig-nificant determinant of women’s assetrights. Specifically, female headship isassociated with higher likelihood of landownership but weaker decision-makingauthority concerning house transac-tions. The latter finding, combined withfindings that both women heads andnon-heads have weaker control overland transactions, lends support to theevidence in the literature that womenare particularly disadvantaged with re-spect to rights to sell, bequeath, or giftassets (Gray and Kevane 2001). A poten-tial explanation for female headship’sassociation with weaker decision-makingauthority over housing is that widows,who form 58 percent of heads in thesample, may be permitted to stay in thedwelling they shared with their late hus-

bands, but are often not considered toown it or the land on which it sits.Rather, her children or her in-laws areconsidered the owners, and their per-mission may be required for any trans-actions. Even mailo lands tend to bemanaged within families according tocustom, which dictates that whenwomen do inherit land, it is generally inthe form of user rights. They have theright to use the land, but not to sell ormake long-term decisions about invest-ments without the consent of male rela-tives (Eilor and Giovarelli 2001)

Marriage is found to have a negative ef-fect on women’s asset rights. Marriedwomen are significantly less likely thanmarried men to own livestock by them-selves, and marriage weakens theirability to make decisions over use oflivestock. Married women are also lesslikely to have their names on land docu-ments than married men.

Natal family presence in the communityis a significant support for men’s assetrights, particularly in housing, while ithas a significantly dampened effect onwomen’s asset rights. For women, hav-ing brothers is also associated withlower decision-making ability overhouse transactions, in line with otherfindings that having brothers may havean adverse effect on women’s assetrights (Rugadya 2010).

These findings are consistent with theliterature at large, which articulates thecentrality of intimate partner relation-ships for women’s asset rights inUganda and also supports the existingevidence that natal family relations arenot particularly supportive of women’sasset rights, likely attributable to thefact that in this region of Uganda,women’s engagement with their natalfamily is minimized after marriage.

How land is acquired is highly impor-tant in defining men and women’sasset rights, once again resulting indifferent outcomes. Most significantly,for purchased land, women are foundto be less likely than men to have their


Land Acquisition: Inheritance



Number of Brothers

Natal Family in Community

Married

Female


Figure 5. Determinants of Asset Rights: Livestock

Livestock


N/A

Note: Reference categories are for partnership status: single; for tenure: other tenure; andfor land acquisition: acquisition through marriage. Regression results should be inter-preted in comparison to these categories

names on land documents. This findingis particularly important in light of thefinding that having one’s name on landdocumentation is associated with sig-nificantly stronger decision-makingover land. That married women arealso less likely to have their names onland documents supports the recentCOHRE report finding that women tendnot to be on land documents evenwhen they contribute to the purchase(COHRE 2010).

Women associated with land acquiredthrough families, on the other hand,emerge as having significantly greatercontrol over decisions on land trans-actions and use as well as transac-tions regarding house. Data do notallow for further exploration of thisfinding as they do not reveal whosefamily land is in question. If in factland is acquired from women’s natalfamilies, this finding emphasizes theimportance of improving policies andprograms that promote girls’ inheri-tance rights.

Finally, as expected, tenure systemshave a significant effect on individuals’land rights. Mailo tenure in particular isfound to be favorable for both men andwomen, while women appear to be dis-advantaged under historical occupa-tion. The relative advantage mailotenure is found to offer to women isconsistent with anecdotal evidencefrom the region, including that women inBuganda buy land at higher rates thanthose in other regions in the country. Itwould be important to explore the fea-tures of mailo land that provide womenwith favorable terms and identify waysto further improve women’s rights underthis tenure.

Exploring the Links BetweenWomen’s Asset Rights andIntimate Partner Violence

In a separate module of the GLAS ad-ministered to women only, the surveyasked women about their experienceof intimate partner violence (IPV) dur-

ing their lifetime and in the 12 monthsprior to the survey. In this report, IPVincludes physical, sexual, and emo-tional violence by a relationship part-ner and is for the most part based onthe World Health Organization method-ology (World Health Organization 2005).The prevalence rates established byGLAS data are similar to those re-ported in other studies (Uganda Bu-reau of Statistics and MacroInternational Inc. 2006), with about 74percent of women reporting having ex-perienced intimate partner violence intheir lifetime. Current prevalence of in-timate partner violence was found tobe 48 percent with about 50 percent ofwomen reporting intimate partner vio-lence with their current partner. 14

A substantial literature delineates sev-eral factors that may increase or de-crease a woman’s risk of violence,including socioeconomic status,women’s empowerment, or prevalenceof violence in the community. A smallsubset of the literature explores the ef-fects of asset rights on women’s experi-ence of violence with variedconclusions. In a majority of these stud-ies, asset rights are found to reducewomen’s likelihood of experiencing vio-lence through empowering women intheir relationships and also through pro-viding them with an exit option (Swami-nathan, Walker, and Rugadya 2006;Panda and Agarwal 2005).

Multivariate analysis was used to ex-plore whether asset rights have a similarprotective effect in the study sample15 bymodeling a woman’s probability of hav-ing experienced IPV in the 12 monthsprior to the survey. Other factors ac-counted for in the model were drawnfrom the literature and findings from thequalitative fieldwork. They include thesame demographic and household com-position measures in the models forasset rights; main livelihood or earnedcash income; quality of the physicaldwelling and total land area associatedwith the household; the woman’s and hercurrent or most recent partner’s alcoholconsumption and partner’s frequency ofdrunkenness; frequency of quarreling

14Krista Jacobs

15

with current or most recent partner andan index of communication betweenpartners; exposure to family violence inyouth; and a series of measures associ-ated with women’s empowerment, in-cluding power in making householddecisions, mobility, involvement in com-munity groups, and self-esteem. Differentmodel specifications were used to ac-count for weak instruments used to ad-dress potential endogeneity.16

The analysis reveals robust evidence onthe significant connections thatwomen’s housing rights and their cashholdings have to lower risk of recentlyexperiencing intimate partner violence.Specifically, across different modelspecifications, having stronger controlover house transaction decisions wasassociated with 18 to 21 percentagepoints lower likelihood of intimate part-ner violence. Also fairly consistentacross different models is the link be-tween women’s cash holdings andlower likelihood of intimate partner vio-lence. In one of the models, cash hold-ing is associated with a 17 percentagepoint lower likelihood of experiencingsuch violence. Both these results sup-port the evidence from various contextsthat asset ownership may protectwomen from experiencing violence intheir personal relationships. What ismore, the particular assets that emergeas significant, housing and cash, sug-gest that asset ownership deters inti-mate partner violence by providingwomen with an exit option. These re-sults support findings from an earlierqualitative study by ICRW in Uganda,which suggested that housing in partic-ular offers women a concrete exit optionfrom abusive relationships (Swami-nathan, Walker, and Rugadya 2006).

Among the key drivers of intimate part-ner violence often cited in the literatureis alcohol consumption, particularly bywomen’s male partners. Evidence fromthis study supports this linkage, as bothwomen’s own alcohol consumption aswell as their partners’ frequency ofdrunkenness are found to be associatedwith higher likelihood of intimate part-

ner violence. Specifically, higher fre-quency of drunkenness by the partnersis associated with 14 percentage pointshigher likelihood of intimate partner vio-lence, while the respondents’ owndrinking is found to increase the likeli-hood of intimate partner violence by 10to 17 percentage points in differentmodel specifications.

The analysis also reveals that womenwho have brothers may be slightly moreprotected from intimate partner vio-lence. Specifically, an additional brotherreduces women’s likelihood of experi-encing violence by 4 to 7 percentagepoints. Having sons does not appear toprotect women from intimate partner vi-olence. On the contrary, the likelihoodof experiencing such violence is foundto increase for women with sons byabout 6 to 8 percentage points.

As revealed in this study, asset rights –particularly housing – may have the far-reaching effect of reducing women’slikelihood of experiencing violence intheir homes, possibly by providing thema viable exit option, as has been notedin other studies (Swaminathan, Walker,and Rugadya 2006; Panda and Agarwal2005). The numerous social, economic,and health-related costs of intimatepartner violence are widely docu-mented in the literature and provide astrong argument for more robust assetrights for women so they can protectthemselves from such violence.

Conclusions and Policy Recommendations

Adding to the growing body of evidence,this study points to significant gendergaps with respect to women’s asset own-ership in Uganda. Further, it sheds light onmore detailed aspects of asset owner-ship, looking beyond land to a wider arrayof assets, and not just asset ownershipbut also control and decision-making au-thority over assets. The gender asset gapis most stark if individual rights are takeninto account and only slightly narrowed ifjoint rights are also considered.

The results also point to significant nu-ances in the nature of the gender assetgap and its drivers. Specifically, thegender-based differences in assetrights are shaped to a large extent bywomen’s position in the household andby their relationships with men. Womenheads of households, more than 85 per-cent of whom are widows, divorced, orseparated and not in an intimate partnerrelationship, emerge as having similarasset rights to men and much strongerrights than women who live in male-headed households. Women who live inmale-headed households as intimatepartners of the male head were found tohave significantly weaker rights. The asset rights of women heads ofhouseholds and of widows have longbeen central to the discussions andpolicy around gender, asset rights, andpoverty. Shaped largely by the dis-course around both the vulnerability ofHIV widows and unfavorable customarypractices in land inheritance, land, andother asset rights of widows and fe-male household heads in most develop-ing countries have been described asquite insecure (Ovonji-Odida et al.2000). Results from this study, however,find female household heads and wid-ows to be slightly better off than otherwomen in terms of their asset rights,though all women tend to have weakerasset rights then men. In some re-spects, particularly decision-making,men also face limitations on their assetrights. These results should not be in-terpreted as diminishing the injustice ofproperty grabbing from women or assaying that female household headsand widows are not vulnerable eco-nomically and socially. Rather, theysuggest that, overall in the study re-gion, widows and female householdheads may be less disadvantaged interms of asset rights – only one aspectof welfare – than previously believed.The next steps are to identify and un-derstand the mechanisms, whetherlegal, customary, economic, or other-wise, that support the asset rights ofthese otherwise vulnerable women andleverage them to strengthen the assetrights of all vulnerable groups.

16

Findings also draw attention to the needto better understand and address theasset rights of women who live in male-headed households. Until recently, fewstudies have been able to explore ex-tensively the asset rights of thesewomen, for the most part due to a lackof individual-level data. What emergesin Masaka is that women who live inhouseholds headed by men tend to ownsmaller bundles of assets, often jointlywith their partners, over which theyhave significantly weaker control. Thisfinding brings to the forefront the needto understand the dynamics of jointasset ownership in households and toidentify the key levers for change thatcan strengthen both the individual andjoint asset rights of women in male-headed households. Current law allowsfor joint ownership of property and pro-tects spouses’ rights to property usedfor sustenance. However, mechanismsneed to be put in place to ensure thatthe law granting spouses joint owner-ship is enforced (Rugadya 2010). Simi-larly, joint titling efforts need to besupported to ensure that claims of jointownership, by women as well as men, canbe substantiated and rights associatedwith this ownership properly exercised.

Recent proposed policy and legislationinclude positive steps to strengthen jointasset ownership rights and to addressgender inequalities in asset rights withinpartnerships. The draft National Land Pol-icy also obligates the Ugandan govern-ment to provide for joint or spousalco-ownership of family land and home(Rugadya 2010). A recent version of theproposed Marriage and Divorce Bill takessteps to clarify property rights withinpartnerships by defining matrimonialproperty and considering it as jointlyowned by default. (Jacobs, Asiime-Mwe-sige, and Hollingworth 2010) The Bill fur-ther states that both parties have equalrights to use and benefit from matrimonialproperty. A provision that entitles personswho have invested substantially in theirpartners’ individual property to an inter-est in the property could be especiallybeneficial for women who devote theirresources to improving land or housing.

Further, the Marriage and Divorce Billproposes allowing married persons andcohabitants to make arrangements as tohow property will be owned during andafter the relationship, and reiterating ex-isting rights of both women and men toindividual property within a partnership.Legislating asset rights for unmarried co-habiting couples represents an importantadvance. Twenty-eight percent of womenin our sample were cohabiting, and sotheir claims over assets acquired duringthe relationship had limited legal and so-cial protection. Such legislation could behelpful in clarifying, and if well enforced,protecting the rights of women in male-headed households, especially women incohabiting relationships. Underlying efforts to improve women’sasset rights through legislation and pro-gramming is the importance of educatingwomen and men on the rights they haveover jointly as well as individually ownedassets. Although low among both menand women, knowledge of laws is oftenfound to be particularly minimal amongwomen (Deininger, Ali, and Yamano 2008).In this study sample, for example, morethan half of the women in male-headedhousehold are not aware of the tenure

system they are under, which means theydo not know their associated rights. Pro-visions in the draft National Land Policyto integrate land rights and administrationeducation into the national curriculumand to decentralize land services are twouseful steps to improve awareness ofwomen’s asset rights.Finally, further research is needed in atleast three critical areas: 1) Joint asset ownership needs to bebetter understood. In particular, furtherareas to explore include the importanceof jointly owned assets for women’slivelihoods and vulnerability; whetherperceived joint ownership is truly jointand equitable between the sexes; andhow use of and decision-making abilityover joint assets differs. 2) A deeper understanding of the charac-teristics, livelihoods, and vulnerabilitiesof women living in male-headed house-holds is necessary to increase their own-ership, use, and control of assets. 3) Additional research should seek tounderstand what makes mailo landtenure favorable for both women’s andmen’s ownership and decision-makingover land as opposed to other tenuresystems.

David Snyder

17

Notes1 The University of KwaZulu-Natal in

South Africa, the other country part-ner, helped develop the surveymethodology and led the pilot studyin Uganda. Results from this portionof the study are reported separately.

2 For the study and in this report, theterm “assets” encompasses land,housing, livestock, material assets(farm equipment, modes of transport,etc.), and financial assets.

3 Nearly 85 percent of economicallyactive women in Uganda work in theagricultural sector producing 70 to75 percent of the country’s agricul-tural output (Rugadya 2010).

4 Land tenure “is the relationship,whether legally or customarily de-fined, among people, as individuals orgroups, with respect to land.” It estab-lishes “how access is granted torights to use, control, and transferland, as well as associated responsi-bilities and restraints”(http://www.fao.org/docrep/005/y4307e/y4307e05.htm) “Being associated”with a particular tenure system meansthat a person lives on, owns, uses, ormakes decisions about a piece of landin that particular tenure system.

5 Freehold tenure grants full rights ofregistered ownership. Leaseholdtenure refers to land leased for aspecific period under certain terms.Under customary tenure, land is reg-ulated by customary rules often ad-ministered by clan or family leaders(Rugadya 2007). Under mailo landtenure, the owners of the land, calledlandlords, have legally recognizedownership rights over the land andhave the right to hold registered landin perpetuity. The tenants who live onthe land have legally recognized oc-cupancy rights (Rugadya 2010).

6 Material assets refers to agriculturalassets (hand tools, plough,tractor/thresher/planter, small farmequipment, wheelbarrow, irrigationequipment, processing equipment),communication assets (cell phones,radio, TV), household goods (refrig-

erator/freezer, sewing machine,small household nondurables,household furnishings), and trans-portation assets (bicycle, motorcy-cle, car, or pickup truck).

7 Financial assets are not part of theanalysis due to the low levels ofownership/access reported by re-spondents.

8 The estimations account for cluster-ing and sampling weights in linearprobability and regression estima-tions and in the calculations ofmeans and frequencies.

9 e.g., see Antonopoulos and Floro2004; Katz 2000; Cross and Hornby2002; Feder and Noronha 1987;Brandt et al. 2002; Place and Migot-Adholla 1998.

10 A product of the 1900 BugandaAgreement between Kabaka (King)of Buganda and the British Govern-ment, mailo tenure essentially priva-tized half of the land in the region,dividing it among the Kabaka, othernotables, and chiefs. The smallhold-ers who were settled on this landunder customary tenure were sub-sumed into a feudal arrangement,making them rent-paying tenants.Rent could be paid in cash busuuluor in kind evujjo to the landlordsamong whom the land had been di-vided. While landowners held formalownership rights and Certificates ofTitle, the tenants had permanent userights to the land (Rugadya 2010).

11 In this context historical occupation isequivalent to occupancy rights underBuganda customary law.

12 “Being associated” with land meansthat a person lives on, owns, uses, ormakes decisions on that piece of land.

13 Data on land acquisition was asked atthe plot level and does not include infor-mation on who inherited land fromwhom, whose money was used in pur-chases, or whose family gave it to whom.

14 The high prevalence rates registered inthis as well as in the DHS study are aresult of the inclusion of emotional vio-lence in the prevalence calculations.

15 Rights included in the model were awoman’s reported ownership of land,

decision-making over land, whethershe had her name on any land docu-ments; reported ownership of hous-ing, decision-making over housing,whether she had her name on anydocuments for housing; number ofmaterial asset types owned, numberof material asset types used; owner-ship of livestock; and having cash.

16 To address potential endogeneity ofwomen’s property rights as determi-nant of their experience of recent inti-mate partner violence (IPV), we usedaugmented regression tests and in-strumental variables estimation as ap-propriate. Augmented regression testsrevealed which of the selected prop-erty rights measures were empiricallyendogenous. These property rightsvariables were replaced with predic-tions of themselves based on modelsin earlier sections originally used toexamine the determinants of propertyrights. However, examination of theearlier models indicates that they maynot be well identified, and instrumentsare weak. Because using weak instru-ments can cause inconsistent esti-mates and inaccurate standard errorsand hypothesis tests, we synthesizeresults across the following four esti-mations: (1) and (2) use original prop-erty rights variables with and withoutmeasures of empowerment, alco-holism, relationship quality, and part-ners’ personal problems; (3) and (4)use predicted property rights variablesfor variables that are endogenous withan alpha of 5 percent or lower, withand without measures of empower-ment, alcoholism, relationship quality,and partners’ personal problems. Tolimit simultaneity bias from past vio-lence leading to changes in women’sproperty rights, we measure violenceas intimate partner violence within the12 months prior to the interview. Ourmodel still implicitly assumes that theproperty rights variables are reflectiveof rights that were established beforethe onset of violence.

Appendix A: Multivariate Linear Regression Results for Models of Asset Rights

A.1 Determinants of Land Rights

Own Document Decision-making¨ ¨¨female -0.343* -0.031 -0.096age in years 0.003** 0.004** 0.001literate -0.061 -0.064 0.063**ever schooled 0.072 0.102*** -0.045cohabitating -0.002 -0.051 0.059married -0.01 0.006 0.048divorced, widowed, separated 0.073 0.031 0.106***earns cash income 0.041 0.022 0.053# of sons 0.007 0.007 -0.001# of daughters -0.007 0.009 0.003# of brothers 0.005 -0.001 0.003any natal family lives in community 0.036 -0.001 0.014

# of adult men in the HH -0.024** -0.02 -0.036# of adult females in the HH 0.06 0.039 -0.011# of persons/room 0.041 -0.004 -0.005index of physical quality of dwelling 0.003 0.032 0.001

land tenure: historical occupation 0.263* -0.002 0.101**land tenure: respondent does not know 0.08 -0.033 -0.011land tenure: mailo 0.238*** 0.104 0.076**land acquired through inheritance 0.163** -0.004land acquired through purchase 0.469*** 0.022land acquired through family 0.125 -0.05land acquired through other means 0.024 -0.204***person has name on land documents 0.064**

female * head 0.237* 0.028 0.049female * cohabitating -0.295 -0.109 -0.06female * married -0.277 -0.266* -0.051female * # of boys -0.012 -0.011 0.007female * # of brothers 0.002 0.009 -0.003female * natal family in village -0.057 -0.016 -0.003female * # of adult males in the HH 0.097 0.09 -0.003female * mailo tenure -0.07 -0.058 -0.086*female * tenure unknown to respondent -0.142 -0.009 -0.055female * historical occupation -0.214* -0.044 -0.190***female * land acquired through purchase -0.325*** 0.003female * land acquired through inheritance -0.125 0.05female * land acquired through family -0.111 0.114**female * land acquired through other means -0.037 0.164***female * name on land documents 0.028

Predicted y_hat 0.624 0.318 0.457R2 0.549 0.406 0.424N 670 665 665

¨ includes transaction (sell, rent, collateralize, gift), bequeathing, and use decisions¨ ¨ scale of land decision-making is (0-1)*** statistical significance at the 1 percent level, ** at 5 percent, and * at 10 percent

18

19

A.2 Determinants of Housing Rights

Own Own (self) Decision-making¨ ¨ ¨female -0.226 -0.294 -0.014age in years 0.005** 0.005* 0literate 0.055 0.05 -0.009ever schooled 0.1 0.097 0.114cohabitating -0.033 0.007 0.174married -0.132 -0.073 -0.092divorced, widowed, separated 0.144 0.196* 0.064earns cash income -0.021 -0.084* 0.031# of sons -0.015 -0.015 -0.003# of daughters 0.005 -0.001 0.014# of brothers 0.008 0.011 0.018*any natal family lives in community 0.180* 0.213** 0.148

# of adult men in the HH -0.064** -0.156** -0.039# of adult females in the HH 0.086 0.061* 0.029# of persons/room 0.022 0.01 -0.045

land tenure: historical occupation -0.041 -0.017 0.247land tenure: respondent does not know -0.058 -0.04 0.033land tenure: mailo 0.017 0.07 0.092land acquired through inheritance 0.2 0.129 -0.139land acquired through purchase 0.184* 0.128 0.046land acquired through family 0.145 0.037 -0.178*land acquired through other means -0.162* -0.192* -0.343***person has name on land documents 0.236*

female * head 0.08 0.135 -0.151*female * cohabitating 0.056 -0.049 -0.247female * married 0.04 -0.042 -0.096female * # of boys 0.028 0.005 -0.002female * # of brothers -0.008 0.003 -0.036***female * natal family in village -0.107 -0.148* -0.077female * # of adult males in the HH 0.07 0.143* 0.039female * mailo tenure 0.045 -0.07 0.025female * tenure unknown to respondent -0.056 -0.018 0.046female * historical occupation -0.066 -0.022 0.069female * land acquired through purchase -0.087 -0.146 0female * land acquired through inheritance -0.236 -0.186 0.015female * land acquired through family -0.13 -0.1 0.108female * land acquired through other means -0.005 0.08 0.238**female * name on land documents 0.057

Predicted y_hat 0.464 0.381 0.280r2 0.265 0.372 0.232N 666 666 666

¨ includes transaction (sell, rent, collateralize, gift) and bequeathing decisions¨ ¨ scale of house decision-making is (0-2)*** denotes statistical significance at the 1 percent level, ** at 5 percent, and * at 10 percent

A.3 Determinants of Material Asset Rights

Own Own (self) Use (total)

female 0.886*** -0.044 1.061age in years -0.012 -0.010** -0.016*literate 0.366 0.361** 0.252ever schooled -0.032 0.244 0.051cohabitating 0.679 0.369 0.267married 0.452 -0.036 0.103divorced, widowed, separated 0.025 0.394 -0.074earns cash income 0.822*** 0.697* 0.837**# of sons 0.096* 0.042 0.086*# of daughters 0.085* 0.078** 0.084**# of brothers 0.009 -0.015 -0.001any natal family lives in community 0.122 0.09 0.18

# of adult men in the HH 0.484 0.315 0.527*# of adult females in the HH 0.355* 0.235 0.450**# of persons/room -0.263** -0.175 -0.192**index of physical quality of dwelling 0.168** 0.05 0.236***

land tenure: historical occupation -0.461 -0.761* -0.386land tenure: respondent does not know -0.599 -1.092*** -0.586land tenure: mailo -0.585 -1.168* -0.569land acquired through inheritance 0.778 0.690* 0.541land acquired through purchase 0.987** 1.188* 0.897**land acquired through family 1.521** 1.463** 1.247**land acquired through other means 0.365 0.394 0.394

female * head -0.269 0.172 -0.596female * cohabitating -1.720** -1.748** -1.454**female * married -1.431* -1.641** -1.129**female * # of boys -0.009 0.06 -0.024female * # of brothers 0.059 0.101* 0.074female * natal family in village -0.487 0.003 -0.366female * # of adult males in the HH -0.094 -0.431 0.044female * mailo tenure 0.918 1.243* 0.903female * tenure unknown to respondent 0.705 0.899 0.728female * historical occupation 0.478 0.252 0.459female * land acquired through purchase -0.492 -0.985 -0.461female * land acquired through inheritance -0.753 -0.582 -0.825female * land acquired through family -0.952* -1.279* -0.894*female * land acquired through other means -0.308 -0.496 -0.227

Predicted y_hat 4.377 3.223 4.619r2 0.349 0.368 0.379N 665 665 665

*** denotes statistical significance at the 1 percent level, ** at 5 percent, and * at 10 percent

20

A.4 Determinants of Livestock Rights

Own Own (self) Decision-making¨female 0.462** 0.024 0.297age in years 0.005** 0.006* 0.005**literate 0.151 0.112 0.17ever schooled 0.108 0.113 0.042cohabitating 0.16 0.143 0.157married 0.219** 0.205*** 0.231**divorced, widowed, separated 0.083 0.109 0.056earns cash income 0.076 0.197* 0.184*# of sons 0.006 0 0.003# of daughters 0.003 -0.013 0.006# of brothers 0.021** 0.016*** 0.019***any natal family lives in community 0.062 0.097** 0.058

# of adult men in the HH 0.007 -0.02 -0.064# of adult females in the HH -0.01 -0.074* -0.115*# of persons/room 0.039 0.05 0.055*index of physical quality of dwelling -0.021 -0.005 -0.024

land tenure: historical occupation -0.109 -0.086 -0.094land tenure: respondent does not know -0.047 -0.158 -0.233land tenure: mailo 0.022 -0.057 -0.023land acquired through inheritance 0.190* 0.099 0.151land acquired through purchase 0.105 0.058 0.105land acquired through family 0.173* 0.035 0.133land acquired through other means -0.062 -0.039 -0.044

female * head -0.115 0.027 -0.096female * cohabitating -0.174 -0.127 -0.209female * married -0.296 -0.322* -0.334*female * # of boys 0.014 0.027 0.026*female * # of brothers -0.006 -0.004 -0.009female * natal family in village 0.037 -0.045 0.018female * # of adult males in the HH 0.023 0.055 0.105female * mailo tenure -0.109 0.062 -0.074female * tenure unknown to respondent -0.04 0.093 0.109female * historical occupation -0.013 0.083 0.126female * land acquired through purchase -0.088 0.054 -0.016female * land acquired through inheritance -0.075 0.054 -0.059female * land acquired through family -0.026 0.091 0.01female * land acquired through other means 0.058 0.212 0.129

Predicted y_hat 0.589 0.468 0.528R2 0.198 0.138 0.167N 665 665 665

¨ includes use decisions*** denotes statistical significance at the 1 percent level, ** at 5 percent, and * at 10 percent

21

22

A.5 Determinants of Financial Asset Rights

Cash (self)

female 0.237age in years -0.006**literate -0.027ever schooled -0.157**cohabitating 0.254married 0.259divorced, widowed, separated 0.117earns cash income -0.103# of sons 0.012# of daughters 0.013# of brothers 0.004any natal family lives in community -0.083

# of adult men in the HH -0.058# of adult females in the HH -0.033# of persons/room -0.033index of physical quality of dwelling 0.056

land tenure: historical occupation 0.343**land tenure: respondent does not know 0.1land tenure: mailo 0.327**land acquired through inheritance -0.053land acquired through purchase -0.095land acquired through family -0.1land acquired through other means 0.011

female * head -0.02female * cohabitating -0.254female * married -0.154female * # of boys 0.013female * # of brothers -0.031female * natal family in village -0.085female * # of adult males in the HH 0.031female * mailo tenure -0.208female * tenure unknown to respondent -0.06female * historical occupation -0.316*female * land acquired through purchase 0.104female * land acquired through inheritance 0.048female * land acquired through family 0.082female * land acquired through other means -0.031

Predicted y_hat 0.395r2 0.193N 665


23

Appendix B: Determinants of Current Intimate Partner Violence

Results are marginal effects from multivariate probabilistic estimation

MODEL 1 MODEL 2 MODEL 3 MODEL 4

variable dy/dx dy/dx dy/dx dy/dx

age in years -0.002 -0.004 -0.002 -0.003literate -0.055 -0.169 -0.240** -0.311*ever schooled -0.058 -0.027 -0.148** -0.109cohabitating 0.392 0.097 0.130 0.157married 0.209 -0.081 0.009 0.055divorced, widowed, separated -0.179 -0.393*** -0.446*** -0.434**# of sons 0.065*** 0.077** 0.074*** 0.080***# of daughters 0.020 0.020 0.081*** 0.065**# of brothers -0.047*** -0.048*** -0.079*** -0.074***any natal family lives in community -0.042 0.018 -0.153 -0.112earns cash income -0.356** -0.195 -0.429*** -0.304*media exposure: # of differentsources of information accessed 0.013 0.004 0.032*** 0.023***

# of adult men in the HH -0.011 0.008 -0.114* -0.102# of adult females in the HH -0.064 -0.082 -0.160** -0.144index of physical quality of dwelling 0.075* 0.053 0.118*** 0.093***# of persons/room 0.108* 0.085** 0.043 0.041hh plot area 0.023 -0.009 0.063* 0.032**

growing up witnessed violence directed to others -0.124* -0.046 -0.024 0.049father was present when growing up 0.070 0.003 0.088* 0.023witnessed/heard father violent toward mother 0.115 0.084 0.251*** 0.249**respondent’s alcohol frequencywith most recent partner 0.109*** 0.173***alcohol frequency of most recent partner -0.130 -0.030 -0.192** -0.045seen most recent partner drunk in the past 12 months 0.146** 0.132** 0.143*** 0.114**frequency of quarreling with most recent partner 0.160*** 0.139*** index of quality of relationship with partner -0.439*** -0.436***Rosenberg self-esteem score 0.006 0.006mobily - number of places need permission 0.024* 0.008household decision-making: average score across decisions 0.075 0.042number of community groups belong to -0.102 -0.052

owns land: self or joint 0.023 0.004 -0.073 -0.058has name on land document 0.132 0.124 0.242 0.189land decision-making average scale -0.078 -0.084 -0.122 -0.111owns house: self or joint -0.041 -0.079* -0.011 -0.069house decision-making transactions -0.181*** -0.199** -0.218*** -0.222**number of diverse material assets owned: self or joint 0.050 0.026 0.025 0.016number of total hh material assets used -0.078 -0.060 -0.055 -0.034has livestock: self or joint -0.090 -0.040 -0.130 -0.062decision-making over all livestock owned by hh -0.024 -0.001 0.004 0.014has cash self -0.176** -0.140(instrumented has cash self) -1.476*** -1.219***

Predicted probability of recent IPV 0.386 0.397 0.387 0.409


24

Reference ListAmpaire, C. et al. Baseline Study Report on Suc-cession-Related Property Grabbing in MukonoCounty, Uganda. 2008. Kampala, InternationalJustice Mission.

Antonopoulos, R. and M.S. Floro “Asset owner-ship along gender lines: evidence from Thailand.”Journal Of Income Distribution 13(2004): 57.

Beegle, K. and B. Ozler “Young women, rich(er)men, and the spread of HIV.” Development (2006).

Beegle, K., E. Frankenberg, and D. Thomas “Bar-gaining power within couples and use of prena-tal and delivery care in Indonesia.” Studies inFamily Planning 32(2001): 130-46.

Benschop, M. Rights and Reality: Are Women’sEqual Rights to Land, Housing and Property Im-plemented in East Africa? 2002. Nairobi, Kenya,United Nations Human Settlements Program(UN-HABITAT).

Bhatla, N., S. Chakraborty, and N. Duvvury.Property Ownership and Inheritance Rights ofWomen as Social Protection from Domestic Vio-lence: Cross-Site Analysis. 2006. Washington,DC, International Center for Research onWomen (ICRW).

Bikaako, W. and J. Ssenkumba “Gender, landand rights: Contemporary contestations in law,policy and practice in Uganda.” Women andland in Africa. L. M. Wanyeki, ed., pp. 232-77.New York, NY: Zed Books, 2003.

Brandt, L. et al. “Land Rights in Rural China: Facts,Fictions, and Issues.” China Journal 47(2002): 67-97.

COHRE. The Impact of National Land Policy andLand Reform on Women in Uganda. 2010.

Cooper, E. “Inheritance and the Intergenera-tional Transmission of Poverty in Sub-SaharanAfrica: Policy Considerations.” Chronic PovertyResearch Centre Working Paper(2010).

Cross, C. and D. Hornby. Opportunities and Ob-stacles to Women’s Land Access in SouthAfrica. 1-180. 2002. South Africa, National LandCommittee and the Department of Land Affairs;Centre for Rural Legal Studies; Centre for Ap-plied Legal Studies (Wits). Promoting Women’sAccess to Land Programme.

Deere, C. D. and C. Doss “The Gender AssetGap: What Do We Know and Why Does It Mat-ter?” Feminist Economics 12(2006): 1-50.

Deininger, K., D.A. Ali, and T. Yamano “Legalknowledge and economic development: Thecase of land rights in Uganda.” Land Economics84(2008): 593.

Doss, C., C. Grown, and C.D. Deere. Gender andAsset Ownership: A Guide to Collecting Individ-ual-Level Data. 4704. 2008. Washington, DC, The World Bank Poverty Reduction and Eco-nomic Management Network. Policy ResearchWorking Paper.

Doss, C.R. “The Effects of Intrahousehold Prop-erty Ownership on Expenditure Patterns inGhana.” J Afr Econ 15(March 2006): 149-80.

Drimie, S. The Impact of HIV/AIDS on RuralHouseholds and Land Issues in Southern and

Eastern Africa. 2002. Sub-Regional Office forSouthern and Eastern Africa, Food and Agricul-tural Organization of the United Nations (FAO).

Duflo, E. “Grandmothers and granddaughters:old age pension and intra-household allocationin South Africa.” Unpublished, 2000.

Eilor, E. and R. Giovarelli. Land Sector Analysis:Gender/Family Issues and Land Rights Study.Draft. 2001.

Feder, G. and R. Noronha “Land Rights Systemsand Agricultural Development in Sub-SaharanAfrica.” Research Observer 2(1987): 143-69.

Government of Uganda. UNGASS CountryProgress Report Uganda: January 2008-Decem-ber 2009. 2010. Kampala, Government of Uganda.

Gray,L.C. and M.Kevane “Evolving tenure rightsand agricultural intensification in southwesternBurkina Faso.” World Development 29(2001):573-87.

Hallman, K. “Gendered Socioeconomic Condi-tions and HIV Risk Behaviours Among YoungPeople in South Africa.” Afr. J. AIDS Res.4(2005): 37-50.

Izumi, K. Women’s Property and InheritanceRights and Livelihoods in the Context of HIV andAIDS. Lusaka, Zambia: United Nations Food andAgricultural Office (FAO) Sub-Regional Office forSouthern and Eastern Africa, 2006.

Jacobs, K., J. Asiime-Mwesige, and G. Holling-worth “Women’s Human Rights in Uganda: AGender-Based Review and Analysis of the Ugan-dan Legal Framework, based on the Conventionto Eliminate all Forms of Discrimination againstWomen (CEDAW) and the Protocol to the AfricanCharter on Human and Peoples’ Rights on theRights of Women in Africa (Women’s Rights Pro-tocol).” Unpublished, 2010.

Katz, E. and J.S. Chamorro. Gender, Land Rights,and the Household Economy in Rural Nicaraguaand Honduras. Annual conference of the LatinAmerican and Caribbean Economics Associa-tion, Puebla, Mexico, October 9-11, 2003.

Katz, E. “Does Gender Matter for the NutritionalConsequences of Agricultural Commercializa-tion? Intrahousehold Transfers, Food Acquisition,and Export Cropping in Guatemala.” WomenFarmers and Commercial Ventures. IncreasingFood Security in Developing Countries. AnitaSpring, ed., pp. 209-31. Boulder, Colo.; London,UK: Lynne Reiner, 2000.

Khadiagala, L.S. “Negotiating Law and Custom:Judicial Doctrine and Women’s Property Rightsin Uganda.” J Afr Law 46(2002): 1-13.

Mason, K.O. and H.M. Carlsson. The Impact ofGender Equality in Land Rights on Development.2004. Washington, DC, World Bank.

Nayenga, R. “Gender Dynamics in Agriculture inUganda: What Are the Key Policy Considera-tions.[should end punctuation here be a ques-tion mark rather than a period?]” PolicySummary Paper(2008): 5.

Oluka, J. et al. “Small stock and women in live-stock production in the Teso Farming System re-gion of Uganda.” Small stock indevelopment(2005): 151.

Ovonji- Odida, I. et al. Land, Gender and PovertyEradication, Is There a Case for Spousal Co-Own-ership of Primary Household Property? 2000. Kam-pala, Uganda, Land Act Implementation Project.

Panda P. and B. Agarwal “Marital Violence,Human Development and Women’s Property Sta-tus in India.” World Development 33(2005): 823-50.

Place, F. and S.E. Migot-Adholla “The EconomicEffects of Land Registration on Small HolderFarms in Kenya: Evidence from Nyeri andKakamega Districts.” Land Economics 74(Au-gust 1998): 36-73.

Quisumbing, A.R. and J.A. Maluccio “Resourcesat Marriage and Intrahousehold Allocation: Evi-dence from Bangladesh, Ethiopia, Indonesia,and South Africa*.”Oxford Bulletin of Econom-ics and Statistics 65(2003): 283-327.

Rugadya, M. Gender in Uganda’s National LandPolicy: Issues, Theories and Policy Responses: Im-plications for Poverty Social Impact Assessment inUganda. Presentation for the Poverty Social ImpactAssessment (PSIA) Training, Ministry of Finance,Planning and Economic Development. 2007.

——-. Women’s Land Rights in Uganda: Statusof Implementation of Policy and Law onWomen’s Land Rights for ECA, ACGS AddisAbaba. 2010. Maastricht University.

SOFA Team. Gender Differences in Assets. ESAWorking Paper. 11-12. 2011. Food and Agricul-ture Organization (FAO).

Swaminathan, H., C. Walker, and M. Rugadya."Women's Property Rights, HIV and AIDS & Domes-tic Violence Research Findings from Two Districts inSouth Africa and Uganda.” HSRC Press, PretoriaSouth Africa (2007).

Thomas, D. “Intrahousehold Resource Alloca-tion: An Inferential Approach.” Journal ofHuman Resources 25(1990): 635-64.

Tripp, A.M. “Women’s Movements, CustomaryLaw, and Land Rights in Africa: The Case ofUganda.” African Studies Quarterly (2004).

Uganda Bureau of Statistics and Macro Interna-tional Inc. Uganda: Demographic and HealthSurvey 2006. 2006. Calverton, Maryland, UBOSand Macro International, Inc.

Walker, C. Land Reform in Southern and EasternAfrica: Key issues for strengthening women’saccess to and rights in land. 2002. Harare, Zim-babwe, Sub-Regional Office for Southern andEastern Africa, Food and Agriculture Organiza-tion of the United Nations (FAO).

Whitehead, A. and D. Tsikata “Policy Discourseson Women’s Land Rights in Sub-Saharan Africa:The Implications of the Re-Turn [Is hyphen cor-rect here?] to the Customary.” Journal of Agrar-ian Change 3 (January 2003): 67-112.

World Health Organization. WHO Multi-countryStudy on Women’s Health and Domestic Violenceagainst Women. 2005. Country Findings. 8-26-2010.

Young, D.E. “The Jurisprudence of Vulnerability:Property Rights, Domestic Violence andHIV/AIDS Among Women in Uganda.”

nternational Center for Research on Women (ICRW)

Headquarters1120 20th Street, NWSuite 500 NorthWashington, D.C. 20036 USAtel: 202.797.0007fax: 202.797.0020email: [email protected]

Asia Regional OfficeC – 139, Defence ColonyNew Delhi, India – 110024tel: 91.11.4664.3333fax: 91.11.2463.5142email: [email protected]

East Africa Regional OfficeABC PlaceWaiyaki Way, WestlandsP.O. Box 20792, 00100 GPONairobi, Kenyatel: 254.20.2632012email: [email protected]

Associates Research Uganda, LimitedPlot 114, Bukoto Street − KamwokyaP. O. Box 595 Kampala, Ugandatel.: 256.414.541988email: [email protected] web: aresearchug.org

University of KwaZulu-Natal - WestvilleSchool of Environmental SciencesDiscipline of Geography4th Floor, H Block South Africatel: 2731.2607322 fax: 2731.2607317web: www.ukzn.ac.za

gender land and asset survey uganda - icrw

Documents