Innovative Grassroots NGOS and the Complex Processes of Women’s Empowerment: An Empirical Investigation from Northern Tanzania

World Development Vol. 66, pp. 762–777, 20150305-750X/� 2014 Elsevier Ltd. All rights reserved.

www.elsevier.com/locate/worlddevhttp://dx.doi.org/10.1016/j.worlddev.2014.09.005

Innovative Grassroots NGOS and the Complex Processes

of Women’s Empowerment: An Empirical Investigation

from Northern Tanzania

MARA J. GOLDMAN and JANI S. LITTLE*

University of Colorado-Boulder, USA

Summary. — We argue that women’s empowerment is an ongoing, incremental, and relational process that occurs across scales andpathways. Using a contextualized mixed methods approach, we measure empowerment processes unfolding across Maasai villages innorthern Tanzania as related to the interventions of two innovative grassroots NGOs. Our results indicate increased personal agencyin the household and increased political participation at the societal level as related to NGO involvement. Moreover, we show links be-tween increased personal agency, and changes in beliefs regarding gender norms at the societal level. This highlights the transformativepotential for change and the links across empowerment pathways.� 2014 Elsevier Ltd. All rights reserved.

Key words — empowerment, women, development, agency, Tanzania, Maasai

“Self-congratulatory ‘feel good talk’ about empowering women—thatpretends to put women at the forefront of achieving peace, prosperity,democracy and development—is no longer possible. There are noshort-cuts to gender justice, but it is vital that we strive towards it, withhonesty and conviction and without pretense.”

[Harcourt, 2009: p. 202]

* This paper would not have been possible without the support and coll-

aboration of Ndinini Kimesera Sikar and the rest of MWEDO, Sinandei

Makko and the UCRT team and the men and women of Monduli and

Longido Districts. To them we owe our sincere gratitude. Additionally we

are especially grateful for the support and assistance provided by Sophia

Parkipuny and Neema Laizer, Esupat Ngulupa as well as the Maasai

enumerators who helped with the original survey. We thank COSTECH

for supporting the research efforts in Tanzania and our funders in the US:

the University of Colorado-Boulder Innovative Seed Grant, Center to

Advance Research and Teaching in the Social Sciences (CARTS), and the

Council for Research and Creative Work (CRCW), Grant-in-Aid; and the

National Science Foundation collaborative research grant (NSF grant #

0921507). We thank Alicia Davis and Nicole Smith for comments on

earlier drafts, and Nancy Thorwardson for help designing the figures and

maps. The paper was greatly improved from comments made by three

anonymous reviewers and the journal editor. All opinions and omissions in

the paper are solely responsibility of the authors. Final revision accepted:September 13, 2014.

1. INTRODUCTION

Empowerment has become a key aspect of women’s devel-opment agendas, featured as the third Millennium Develop-ment Goal (MDG). 1 Yet there remains a great deal ofambiguity and contestation over what empowerment means,how it is measured, and how it is best achieved. Initial effortsto conceptualize empowerment stressed that it was a socio-political process, involving shifts in political, social, andeconomic power between and across both individuals andsocial groups (Batliwala, 2007, emphasis in original).Feminist theorists and critical development scholars suggestthat today “empowerment” has become just another develop-ment buzzword, perhaps the most widely used, abused, andhollowed out of critical political meaning (Batliwala, 2010;Parpart, Rai, & Staudt, 2002; Rowlands, 1998). 2 The trans-fer of the word into development lexicon, diluted or erasednotions of power tied to original theorizing and strategizingbehind women’s empowerment (Batliwala, 2010). This iscommon practice in development where ‘magic bullet’ quickfixes are sought for complex social, political, and economicproblems. 3

The use of empowerment discourse within development iso-lates different aspects of empowerment processes for targetedintervention and measurement at different scales (Malhotra,Schuler, & Boender, 2002), often based on external assump-tions of importance (Kabeer, 1999). A focus on personalempowerment targets girls’ enrollment schools (i.e., MDG#3); an economic empowerment focus leads to micro-enterprisedevelopment interventions; and a political empowerment focuspromotes leadership training and quotas. Empowerment is

762

then measured as isolated outcomes, such as an increase inthe number of girls completing school, women’s participationin the workforce, and female representation in government.These outcomes are isolated not only by the sphere of engage-ment (personal, economic, political), but also the scale of anal-ysis (individual, household, community/society). 4

Similarly, scholarly analyses of empowerment processeshave focused almost exclusively on one sphere and scale ofengagement (Malhotra et al., 2002). However, empowermentis a multidimensional process that follows multiple pathways(personal, economic, political) through and across multiplescales: individual (mind and body), household (family), andcommunity (society at large) (Mahmud, Shah, & Becker,2012). Interactions across scales and between different path-ways matter for determining the impacts that projects haveon individual women’s lives and the prospects for transforma-tional change at the societal level. To reinvigorate the term“empowerment” with ‘power’ and ‘process’—we highlightchanges occurring across scales and pathways, through

http://dx.doi.org/10.1016/j.worlddev.2014.09.005

http://crossmark.crossref.org/dialog/?doi=10.1016/j.worlddev.2014.09.005&domain=pdf

INNOVATIVE GRASSROOTS NGOS AND THE COMPLEX PROCESSES OF WOMEN’S EMPOWERMENT 763

rigorous measurement of empowerment-based developmentinterventions using a grounded mixed methods approach withparticular attention to local context.

In this paper we discuss changes occurring in four villagesacross two districts in Northern Tanzania as related to Maasaiwomen’s involvement in ‘empowerment’-based developmentinterventions. We focus on the work of two innovative grass-roots Non-Governmental Organizations (NGOs) workingacross pathways and scales to address issues particular toMaasai women and men: Maasai Women’s DevelopmentOrganization (MWEDO) and Ujamaa Community ResourceTeam (UCRT). While our study provides only a snapshot ofan evolving process, we suggest changes are happening thatcan be understood as part of the empowerment process.Women are gaining more access to resources and knowledgeto challenge power dynamics personally, in the household,and at the community level. While there are other dynamicsat play, we focus on the NGO work in particular as a newintervention bringing new kinds of resources and knowledgeto these areas.

Our study draws from qualitative and quantitative data col-lected over a four-year period (2009–13) as well as over a decadeof ethnographic engagements with the communities by the leadauthor. Our analysis combines insight from feminist develop-ment work on empowerment, with Maasai understandings ofwhat it means to be empowered, and changes Maasai womenare seeking in their communities. 5 We illustrate the impactsacross organizational and institutional scales, highlighting thatdifferent types of interventions work together and separatelyleading to different outcomes such as changes in personal beliefsor increased agency. Echoing Batliwala (2010: p. 115) empower-ment (in a transformational sense) cannot be achieved by tack-ling any one element of social power.

In the following section we outline our conception ofempowerment and address the role that can be played by inno-vative grassroots organizations. We then provide backgroundon the study sites and organizations under review, followed bya discussion of methods and our conceptual framework. Wepresent quantitative and then qualitative results and concludeby reviewing how they reinforce our conceptual model to illus-trate the multi-layered nature of empowerment as a process.

2. CONCEPTUALIZING EMPOWERMENT AND THEROLE OF GRASSROOTS ORGANIZATIONS

Naila Kabeer suggests that innovative grassroots-basedorganizations are best positioned to adhere to the originaltenets of empowerment in their project conceptualizationand implementation. Such organizations can provide spacefor women’s voices to be heard, and are more ‘downwardly’accountable than larger national and international NGOs(1994: 230). This is, of course, not always the case with grass-roots NGOs (Hodgson, 2011; Igoe, 2003), and the adjective‘innovative’ is used to differentiate NGOs that manage to trulymake way for participatory women-led development. Kabeer’sanalysis of “innovative NGOS” in India suggests they aremore prone to recognize that no single magic bullet interven-tion is likely to succeed, and instead work across different cat-egories of need. NGOs need to provide women with access to“new kinds of resources, thereby signaling new potentials andpossibilities” (Kabeer, 2003, p. 261).

Kabeer’s description of innovative NGOs reflects her con-ceptualization of empowerment as a process including agency,resources, and achievements (Kabeer, 2005). Agency refers to“the processes by which choices are made and put into effect;”

resources set the context “through which agency is exercised;”and achievements refer to “the outcomes of agency” (2005, p.14). Resources can be material, relational, legal, and/orknowledge, and are linked to institutional norms and a per-son’s position in society (Kabeer, 1999). Women’s ability toaccess resources is key to implementing change. Empower-ment can thus be seen as a process where, with resources andthrough agency, choices are made by those normally not ina position to do so in a given situation. Achievements occurin the process, often measured as outcomes or proxies forempowerment (i.e., education, employment, political represen-tation). Achievements can also be seen as resources forempowerment if empowerment is viewed as a process ofsequential, and sometimes simultaneous stepping-stones—changes in one sphere leading to subsequent changes in otherspheres.

Drawing from Kabeer (1999, 2005) and Parpart et al. (2002)we call for an iterative, complex, and nuanced conceptualiza-tion of empowerment. Parpart et al. (2002, p. 4, followingRowlands 1995, 1997, 1998) argue that empowerment“involves the exercise rather than possession of power,” andincludes both individual conscientization (power within) andpoliticized power with others, which can lead to the power tobring about change (Parpart et al., 2002, p. 4; see alsoKabeer, 2005). Empowerment is thus, not just about thepower of participation in decision-making, but the processesthat lead individuals to believe they have the capabilities tomake decisions. In other words, empowerment is about per-sonal, relational, and collective processes, and includeschanges in beliefs and attitudes as well as structural and mate-rial change.

This conceptualization of empowerment reflects the viewsvoiced by Maasai women themselves. In interviews with edu-cated ‘empowered’ Maasai women working for change in thecommunities (often through NGOs), a common theme cameup that was echoed in discussions with Maasai across thevillages. People spoke of empowerment as having the aware-ness and confidence to achieve one’s goals, and the ability toshare such knowledge and awareness with the larger commu-nity. One woman who worked for UCRT (and previouslywith MWEDO) explained what she thought constitutes anempowered Maasai woman:

The one who has self-confidence, freedom of her own life and controlof decision-making processes among other women in that particularvillage. She is the one who can work hard for her own activities andothers follow her example. She can look after her children and her hus-band even if the man is a drunkard. She is the one who can participatein government, NGOs and traditional meetings, and can present herideas in front of the people. She is the one who can do her own smallbusiness and not depend too much on her husband. 6

While most Maasai men and women we spoke with dis-cussed the empowered woman as someone who drew fromher internal strength to make due despite her circumstances,there was a sense that such a woman would need to knowwhat she was capable of. Another educated Maasai womenworking for an NGO articulated this clearly by explainingthat to be empowered requires, “knowing that you have thispower. Because in a Maasai community you are regarded as achild, and so on, but you know what you are and you won’tlet anyone take advantage of you, or keep you back becauseyou know your purpose and can go forward.” 7 She suggestedthat this often means workers like herself need to bringawareness to communities, to bring them knowledge to givethem the boost to empower themselves. In other words,NGOS can facilitate the resources needed for agency to beactualized.

764 WORLD DEVELOPMENT

3. STUDY SITE AND ORGANIZATIONS

We focused on the activities of two “innovative grassrootsorganizations” across two districts in northern Tanzania,dominated by people of the Maasai ethnic group. Here we dis-cuss Maasai as a particular focus for the study, the specifics ofthe organizations under review, and the similarities and differ-ences across villages.

(a) The Maasai case study

Maasai are often recognized in both academic and popularliterature as one of the most patriarchal societies in East Africa(Lazaro, 2008; Mitzlaff, 1988; Spencer, 2003) with strongcultural norms legitimizing several forms of institutionalizedgender inequality. 8 Maasai women are often discussed andtreated as ‘children’ and historically were excluded from com-munity meetings. Today women participate in meetings,though in some places for ‘traditional’ reasons they do notstand while speaking. As Hodgson (1999) has noted, the prac-tice of referring to women as ‘children’ is used as a technique tokeep women “in their place” and to reinforce structuralinequalities in access to resources and control over productionand social relations. Hodgson has also shown that women domaintain power in domestic as well as the public sphere in waysthat often go unrecognized by outsiders (Hodgson, 2000, 2001,see also Talle 1987, 1988). Additionally, with increased diversi-fication, Maasai women are taking on new roles and expandingold roles in relation to livestock production (Wangui, 2008),and market exchange (Brockington, 2001; McCabe, 2003;Smith, 2014). Recognizing these factors, we are concerned herewith the structural and ideological contexts which keep womenin a subordinate position relative to men within Maasai com-munities; as well as the work of NGOs targeting Maasaiwomen for ‘empowerment’ development interventions.

(b) Innovative grassroots organizations

We focus on two organizations that we perceive as “Innova-tive Grassroots Organizations,” which are based locally inArusha (the largest city in the area, see Map 1), were startedby Maasai, and are staffed by Tanzanians.

(i) MWEDOMWEDO works in four Maasai dominated districts in Tan-

zania, including Longido and Monduli. At the start ofresearch in 2009, MWEDO had been working in Longidoand Engare Naibor the longest, in Kimokowa for just over ayear, and had just started in Oltukai village. MWEDO is awomen’s organization, with a mission to support effortsthat “empower women within their cultural traditions . . .affirm[ing] their right[s] to benefit both economically andsocially through educational rights.” 9

MWEDO promotes a rights-based agenda (Hodgson, 2002)linked to specific programs (education, economic empower-ment, health education/maternal health, and HIV/AIDS edu-cation). 10 MWEDO educates the community on the importanceof education as a human right that has been denied to Maasaiwomen and girls, and then gets support for girls education 11

and women’s adult education. Women members learn abouthuman rights, women’s rights, and the right to own property,and are encouraged to request legal titles to land for theirfamilies. 12 Land ownership is also discussed at the villagelevel, where support is gained for women to play a vital rolein the process of obtaining titles. 13

MWEDO works on economic empowerment by registeringwomen’s groups (20–25 people), with goals for enterprisedevelopment. Members pay dues, attend annual meetings,have a say in program planning and elect MWEDO boardmembers. MWEDO helps groups find markets for productsand with training on business development and group dynam-ics. Some MWEDO members have come together to form awomen’s leadership ‘forum’, with representatives at the vil-lage, district, (except Monduli), and national levels, to addressissues such as land and human rights; MWEDO facilitates theforum through leadership training.

(ii) UCRT and the community forum (CLF) projectUCRT is a community-based organization that works with

indigenous communities in Tanzania to provide informationand skills for improved resource management. In 2007, in col-laboration with the Tanzanian Natural Resource Forum,UCRT initiated community leadership forums (CLFs), whichwork with customary governance for improved naturalresource management and conflict resolution.

Customary governing institutions, often more respected andeffective than elected leaders, are limited in scope and oftenexclusively male. The CLF addresses this by providing trainingon national laws and facilitating a parallel women’s forum towork with customary male leaders. In 2009, UCRT establisheda women’s forum to work independently and then together withthe men for a joint CLF at village, ward, and district levels. 14 Inaddition to improving local governance through increasedwomen’s participation in decision-making processes, there wasan objective to empower women through leadership training.

At the start of research the women’s forum had just begun.Women had selected village leaders who began training on gov-ernance issues, women’s rights and empowerment, and nationalland laws. Women also learned leadership skills and gained con-fidence to speak in meetings with customary leaders. The CLFproject works in Longido, but not in Monduli district.

(c) Study site: villages

Survey research was conducted in one village in Monduli dis-trict (Oltukai) and three in Longido district (Longido, EngareNaibor, Kimokowa). All villages are part of the Kisongo sectionof the Maasai, sharing similar cultural norms. All villages werepart of the same district until 2007. Despite similarities, villageshave had different exposure to NGO activity. This created anatural setting for a quasi-experimental study to help tease outdifferences in empowerment outcomes across a spectrum ofNGO exposure (see Table 1 for village comparisons).

In Monduli district, Oltukai is an hour drive from the maincity (Arusha), and a one-hour walk to the main road andweekly market. The lead author has worked there since2002. Due to district politics, very few NGOS work in Oltukai,especially those with political motivations. MWEDO startedworking there in 2008. 15 Oltukai thus provides for an idealcomparison where early changes coming in due to the workof MWEDO could be observed.

Longido is a new district (2007), 16 with issues of land grab-bing from Kenya and for new administrative buildings in themain town, Longido. There has been a growth of businessesalong the main road in Longido town. Participants in thestudy vary in distance away from the main road (up to overan hour). Longido has benefited from the work of both MWE-DO and UCRT, and from other NGOs, including anotherwomen’s organization, TEMBO. 17

Kimokowa is located just north of Longido on the main road,has only recently been involved with MWEDO, and is also

Map 1.

Table 1. Village comparisons

Village

Oltukai Kimokouwa Longido Engare Naiborn = 58 n = 60 n = 68 n = 27

NGO involvementAverage # 0.97 1.02 1.53 1.19% in 2 or more 1.70 3.30 30.90 18.50% in TEMBOb 0.00 3.30 27.90 0.00

Wealth% Improved housinga 0.00 11.60 13.50 7.40Average # cattle 8.80 12.40 9.10 7.20% Breadwinner 47.30 27.60 50.70 32.00% Tap water in HH 0.00 35.00 11.80 0.00

Education% Completed primary 5.20 17.60 4.50 15.40

a Housing construction beyond traditional mud and grass.b TEMBO is a women’s based development organization.


involved but to a much lesser extent with TEMBO. 18 Kimok-owa has a local female member on the district council. EngareNaibor is the most geographically isolated of all the villages. Itis a 1–2-h drive on a dirt road to Longido town. Close to theKenyan border, Engare Naibor is subject to cross-border landgrabbing, but also active local politics with well-respected femalerepresentatives and an active forum of community leaders.

4. METHODS

Data for this paper were collected over a period of 4 years.In this section we describe the methods used for data collec-tion and analysis, and our conceptual model.

(a) Data collection

The survey used was designed in 19 to explore women’s wellbeing, drawing from similar models used by WHO and byGrabe in Nicaragua, with modifications made for culturalrelevancy. The survey was translated into Swahili and con-ducted in Maa by trained Maasai women. Participants werewomen whose names were listed as MWEDO members. 170surveys were conducted in 2009. We eliminated widows andun-married women from the quantitative analysis as theirresponses could not be used to judge husband–wife relations,reducing the sample to 152 women.

Qualitative data were collected in 2009, 2010, 2011, and2012 for a total of 47 individual interviews with women


(widows and married women), ten interviews with men, 11group interviews with women, and over 150 hours of ethno-graphic observation including informal conversations withboth men and women in all the study sites. Qualitative datawere used to inform the conceptual and statistical models, toprovide greater depth on context, attitudes, and processes ofchange, and to provide additional information regardingempowerment processes not apparent from the survey.

(b) Conceptual framework and data analysis

Our conceptual model draws from the work of Kabeer (1999)in understanding empowerment as related to three inter-con-nected dimensions: pre-conditions or context (resources andcontrols), personal agency (household level decision making),and outcomes (social norms and practices; see Figure 1).Following Kabeer we focus on empowerment as an ongoingand iterative process, rather than something that can be mea-sured as a final outcome. We also see it as a process that followsmultiple pathways across different scales simultaneously and ina relational fashion. We use the scale designations of individual,household, and community as units of measurement wheresocial relations play out. 20 This is expressed in Figure 1 withthe solid arrows referring to the different pathways of empow-erment measured by our model, and the dotted lines indicatingthe iterative process of empowerment, with “outcomes” eventu-ally leading to further change in time. We therefore discuss‘outcomes’ as forms of transformational change measured asindicators of an on-going process of empowerment.

Empowerment processes occur within particular contexts,which comprise both enabling resources and constraining con-trols. Resource context variables include: the level of formal

Figure 1. Conceptual model

education achieved, the number of cattle a woman has, 21

whether she earns as much or more cash income than herhusband (“breadwinner”), and her level of involvement inMWEDO. MWEDO involvement is a summative scale con-structed from indicators of degree of involvement as a memberof a MWEDO group, i.e., attending adult education classes,going to and talking at MWEDO meetings, and the lengthof time involved with the organization. 22 MWEDO involve-ment involves some choice and therefore it reflects the extentto which a woman is actively seeking to empower herself.

Context is also defined by two variables that measure con-trol or constraints on agency: a woman’s relative age (age-set), 23 and the degree to which her husband controls hermovement and participation in household and village affairs.Older age is often empowering for Maasai men and women,with elder women earning more respect and often having morefreedom and power over younger women (within householdsand society at large). However, age can inhibit changingbeliefs and attitudes. Since older women who are seekingincreased empowerment will likely have higher levels ofMWEDO involvement, the remaining age effect will mostlikely be associated with preservation of traditional norms.

Personal agency measures decision-making at the householdand individual levels regarding topics normally outside ofwomen’s control such as buying and selling livestock or decid-ing on sexual intercourse (see Appendix). Agency must involvenot only making choices, but doing so in ways that challengepower relations (Kabeer, 2005). In our model, these remain inthe realm of the household. Our outcomes, on the other hand,can also be seen as indicators of empowerment or ‘achieve-ments’ in Kabeer’s language in that they challenge socialnorms and practices (Kabeer, 2005). The outcomes in the

of empowerment process.


model include: participation in local meetings (political partic-ipation), belief in non-traditional gender norms, and loweracceptance levels of the use of violence by husbands as a formof disciplining their wives (non-acceptance of violence as disci-pline).

Political participation measures the degree to which awoman participates in societal level meetings. This is specificto Maasai culture because it measures not only if the womanspeaks at meetings (village wide and village governmentbased), but also if she stands while doing so. Historically,according to Maasai custom, women were not allowed tostand while speaking, unless a man also stood at the sametime. 24 Thus standing up while speaking at meetings is a cul-turally relevant indicator of change. The second measurementlooks at culturally relevant indicators that suggest a womanhas begun to uphold non-traditional gender norms, such asbelieving she should have freedom equal to a man’s and choos-ing whom she marries. Our final empowerment scale measuresthe degree to which a woman fails to accept violence from herhusband as a necessary form of wife discipline. This scalereflects a woman’s acceptance of various reasons why her hus-band may beat her. 25 The failure to accept violent disciplineshould be viewed as an indication of both deeply personal(related to personal views of her body, and the rights of herhusband), and societal empowerment. Domestic violence isincredibly prevalent in Maasai communities for reasons tiedto both long-accepted customary beliefs and high levels ofalcoholism. We do not include actual measurements of vio-lence because the data were not reliable and had little varia-tion. 26 This measure of non-acceptance of violence is thus abarometer of potential changes to come at the societal levelas it challenges long-held social beliefs about gender rolesand expectations. 27

In order for these different outcomes to have the impact weare suggesting as transformative forms of agency, they need tofeed into eventual changes in context and agency. The regres-sion models were unable to illustrate this, but our qualitativedata do. 28 For instance, for women to actualize their personalfeelings regarding non-traditional gender norms into societalchange (to move from ‘power within’ to ‘power with’), theyneed to be actively participating in societal meetings, promot-ing such changes. While our data can show this is occurring(i.e. through increased political participation), qualitative dataillustrate how increased political participation put beliefsregarding gender norms into action, exemplifying the transfor-mative types of agency (Kabeer, 2005).

Qualitative data were transcribed and analyzed both inde-pendently and in relation to the quantitative findings. Analysisinvolved looking for commonalities across interviews andobservations, noting changes as discussed by women and asethnographically observed over time (such as women’sincreased participation in meetings), and looking for illustra-tions (through quotations and actions) of processes outlinedin the quantitative analysis. 29 We now turn to our results,moving between qualitative and quantitative findings.

5. RESULTS: EMPOWERMENT PROCESSESUNFOLDING

Here we first present qualitative data to illustrate the contextand relevance of the variables used for measuring empower-ment. We then present results from the regression models,which focus on MWEDO involvement as the primary facilita-tor and resource for empowerment. These results are inter-mingled with qualitative description to provide deeper

explanation of the findings. Qualitative data appear again atthe end to explain links and flows not articulated in the modelsand related to the work of UCRT.

(a) Context examined: underlying conditions for change

Qualitative interviews began with the open ended question,“what are the biggest problems facing women in this commu-nity?” The predominant response across villages was women’slack of ownership/control of anything and therefore depen-dence on their husbands for everything. Some described thisas needing to ask their husband for everything, even salt fromthe market, and not being able to help themselves because,“everything belongs to the husband.” 30 Some suggested that thissituation often leads to arguments and violence. Othersdescribed this lack of control over anything as related to thevery institution of marriage.

As one woman explained, “when you get married you goonly with a calabash (for storing milk) on your back. The cat-tle and children are all under the control of husband. If youwant to be in charge at all, if you want to sell [cattle] your hus-band tells you to be quiet because you came with nothing but acalabash.” 31 This situation is often exacerbated by the mar-riage exchange, cemented through a bride price paid by thehusband to the wife’s family. Women complained that this,combined with not owning anything, leads to a situationwhere their husband, for all intents and purposes, owns them.They explained that because they are caring for his cattle, hisland, and even his children, he could and would beat them ifthey did so poorly.

One woman from Longido explained the situation in generalterms:

We have problems because there is nothing to say that is mine. Even achild that you give birth to is not yours, but his. You do the work, butall of it is his. Cattle he says are yours but only in name because youcannot sell them or give them away, they are his (Elder widow Long-ido, May 26, 2009).

For this reason, we include in our measurement of personalagency, decisions that may be otherwise seen as within therealm of women’s work already, such as making decisionsabout purchasing food or children’s medical care. Womencannot make such decisions without asking their husbandsfor money, and the husband has the final say (i.e. if the childgoes to the hospital or determining how much maize tobuy). 32 In this context, women making decisions alone ortogether with their husbands regarding their children’s healthand household economics does indicate increased agency, asdoes making decisions regarding cattle, land, and sexual inter-course.

Women rarely named violence as a major problem impact-ing their communities, or themselves personally. This is notbecause violence against women is not a problem in these com-munities, but because it is so prevalent and accepted as a socialnorm. We did question women directly about violence theyhad themselves experienced and their thoughts on a man’sright to beat his wife. There was a standard list of legitimatereasons for a man to beat his wife that was repeated by nearlyall women. This list was used to design the questions in thesurvey as well as to construct the scale used to measure thelevel with which a woman rejects violence as a form of disci-pline by her husband. Many women stated that it was neces-sary for a wife to be disciplined. When pushed, somesuggested violence was not needed, though most did think thata beating was permissible in certain circumstances, such as err-ing with the cattle, or being caught having an extramaritalaffair. The most common complaint regarding violence was


what women referred to as ‘beating for no reason,’ and thatwas nearly always associated with excessive alcohol consump-tion.

(b) Change unfolding: models of quantitative change

The processes of empowerment set out in the conceptualmodel (Figure 1) were operationalized and tested with pathmodels that address the processes of gaining personal agency(Empowerment Processes I, Figures 2 and 3) and the processesthat lead to empowerment outcomes (Empowerment ProcessesII, Figure 4). We hypothesize that resources are needed toenact personal agency which can then provide the ‘power to’lead to more transformative forms of agency (our outcomesor empowerment indicators). All of the path models wereestimated with linear regression and are summarized inFigures 2–4.

The first model (Figure 2) summarizes the importantresource and control measures that lead to personal agency.All the measures were included in the regression model, butonly MWEDO Involvement, husband control, formal educa-tion, and breadwinner were statistically significant. Adjustingfor all other contextual covariates, MWEDO Involvementshows the strongest effect on personal agency (beta = .25).Husband control and formal education are similar in the mag-nitude of effect, but, as expected, husband control has arestrictive effect (beta = �.15); the effect of formal educationis positive (beta = .15) and encourages personal agency. Thebreadwinner effect is positive (beta = .13) and demonstratesthat a wife who generates as much or more income as her hus-band tends to have higher personal agency, on average.

Research from various contexts around the world hasshown the link between education and women’s empower-ment. Improved education is the single pathway targeted inMDG#3 for women’s empowerment. MWEDO is clearlyaware of the importance of education for women. The organi-zation has always worked to support girls’ education throughscholarships to needy families and recently established theirown boarding school for Maasai girls.

Figure 2 shows how specific resources and controls affect thetotal personal agency score, but it does not reveal how contextimpacts different components of personal agency. Figure 3summarizes results from five different path models, each onetreating a different personal agency subscale as the dependentvariable. The pathways included are those that have statisti-cally significant regression coefficients, with the sign of the

Figure 2. Empowerment processes I: resources and controls affecting

personal agency. Note: Standardized coefficients, *p < .10, **p < .05.

coefficient reported on the path. (See Table 2 for specificresults for all models.) Starting on the left side, the significantpredictors of decision-making regarding land are husbandcontrol (negative) and MWEDO involvement (positive). Awife’s decision-making regarding cattle is positively affectedby her wealth, i.e., the number of cattle she has, and by herinvolvement in MWEDO.

A wife more often makes household economic decisions anddecisions regarding children (healthcare, school, and mar-riage) if she is the breadwinner. Higher levels of husband’scontrol inhibit the wife’s decision making regarding the chil-dren. Decisions about sexual intercourse between husbandand wife are more often made by the wife if she is moreinvolved in MWEDO, and less often made by the wife if sheis older. 33

MWEDO involvement is important for decision-makingregarding land, cattle, and sexual intercourse, but not forissues involving household economics or children. The effectsof formal education are not significant for any of the subscalesalthough it does have a significant effect on overall personalagency. Husband control has a negative effect on decision-making regarding land and children.

The processes of empowerment that lead to changes insocial norms and practices are reported in Figure 4, whichreflect the results of three different structural regression mod-els. Each of the outcome measures (political participation,belief in non-traditional gender norms, and non-acceptanceof violence as discipline) is the dependent variable in a sepa-rate path model that controls for the resources and controlsvariables, and the total personal agency scale is included inthe model as a mediator variable. The significant pathwaysare reported in Figure 4 along with the signs of the regressioncoefficients. Coefficient values are reported in Table 3.

The model results confirm that personal agency positivelyaffects political participation, non-traditional gender norms,and non-acceptance of violence as discipline. Political partici-pation is also directly and positively affected by wealth (i.e.,number of cattle) and MWEDO involvement. Husband con-trol positively affects changes in beliefs regarding the socialnorms of gender as well as violent discipline. While the modelcannot prove directionality, qualitative data highlight that as awoman learns about her rights, she is more likely to react towhat she sees as unfair treatment by her husband. This sug-gests that the more a husband tries to control his wife, themore likely she is to believe he does not have the right to beather.

How can we explain the positive relationship betweenMWEDO involvement and increased personal agency in thestudy area? The model cannot confirm causality because allmeasurements are occurring at the same time, nor are we ableto show the impacts of the different MWEDO programs indi-vidually. However, we can see that differences in empower-ment outcomes do exist across the villages and thesecorrespond to the length of time that MWEDO has beenworking in each village (in ascending order of time—Oltukai,Kimokouwa, then Longido and Engare Naibor) as presentedin Table 4. These differences across villages are likely leadingto the corresponding differences in personal agency andempowerment outcomes that we see. Again, the tests for differ-ences in means reported in Table 4 show that, on average,women in Oltukai have significantly lower levels of personalagency (total, cattle, land, and children); lower levels of polit-ical participation; and more traditional gender norm beliefsthan the women in the other villages. For many of the samemeasures, the two villages with less exposure to MWEDO(Oltukai and Kimokouwa) have lower levels of personal

Figure 3. Empowerment processes I: pathways to personal agency subscales.

Figure 4. Empowerment processes II: pathways to outcomes as mediated by personal agency.

Table 2. Standardized regression coefficients from models of empowerment processes I, pathways to personal agency (total and subscales)

Total Land Cattle HH Economics Children Sexn = 147 n = 147 n = 147 n = 147 n = 147 n = 147

Resources and controlsMWEDO involvement .25** .28** .18** .05 .08 .17**

Husband social control �.15* �.14* �.07 �.01 �.14* �.11Formal education .15* .13 .09 .10 .08 .05Number of cattle .05 �.05 .21** �.12 .07 .08Breadwinner .13* �.12 .06 .28** .25** .09Age category �.05 .12 �.07 .00 �.10 �.20**

R-squared .15 .17 .10 .11 .11 .08

Probability coefficient equals zero: *p < .10, **p < .05.


Table 3. Standardized regression coefficients from models of empowerment processes II, pathways to outcomes

Political participation Belief in non-traditional gender norms Non-acceptance of violence as disciplinen = 147 n = 147 n = 147

Resources and controlsMWEDO involvement .24* .02 �.09Husband social control �.01 .19** .21**

Formal education .01 .16* �.09Number of cattle .17** �.06 �.04Breadwinner �.12 .01 .32***

Age category .12 �.04 -.21**

Personal agencyTotal .14* .16* .23**

R-squared .16 .09 .25

Probability coefficient equals zero: *p < .10, **p < .05, ***p < .001.

Table 4. Married women characteristics by village exposure to MWEDO

Oltukai (1) Kimokouwa (2) Engare Naibor and Longido (3) Significant contrasts p 6 .10n = 47 n = 47 n = 66

Resources and controlsMWEDO involvement (0,6) 2.06 2.30 3.24 1 vs. 2, 3 and 1, 2 vs. 3Husband social control (0,6) 1.85 1.70 1.65Formal education (0,7) 0.26 0.67 0.45 1 vs. 2, 3Number of cattle 9.47 13.83 8.28Breadwinner (0,1) 0.46 0.22 0.42Age category (0,1) 0.36 0.37 0.58 1, 2 vs. 3

Personal agency (household decision making)Total (0,17) 4.68 6.64 8.04 1 vs. 2, 3 and 1, 2 vs. 3Cattle (0,3) 0.28 0.55 0.76 1 vs. 2, 3 and 1, 2 vs. 3Land (0,5) 0.04 0.57 1.81 1 vs. 2, 3 and 1, 2 vs. 3Household economics (0,3) 1.53 1.64 1.86 1, 2 vs. 3Children (0,3) 0.94 1.57 1.41 1 vs. 2, 3Sex (0,3) 2.02 2.40 2.23

Agents of social transformationPolitical participation (0,4) 0.40 0.81 0.82 1 vs. 2, 3Belief in nontraditional gender norms (0,3) 0.89 1.36 1.55 1 vs. 2, 3 and 1, 2 vs. 3Non-acceptance of violent discipline (0,1) 0.37 0.37 0.35


agency and empowerment outcomes than the villages whereMWEDO is more established (Engare Naibor and Longido).Other differences across the villages that could also be contrib-uting to the different empowerment outcomes include formaleducation (with Oltukai having much lower levels than allthe other villages) and age, with the Longido and EngareNaibor sample being older, on average, than the other twovillages. However, given that age is often a deterrent to chang-ing beliefs (see for instance Figure 4), we suggest that the maincontribution to the changes we are seeing is involvement inMWEDO, as outlined in the next section.

(c) Changes articulated: qualitative findings linked to MWEDO

In Oltukai village MWEDO has been working for the short-est period of time and is the only organization working towardwomen’s empowerment. When asked which MWEDO projecthas been the most beneficial for them personally, womenspoke unanimously about the adult education program. Whilethe program did not necessarily succeed at making all partic-ipants literate, it did teach them the national language (Swa-hili, which most Maasai women do not speak), their politicalrights, and national laws regarding land, and women’s rights.Women explained that now that they know some Swahili they

are able to travel more easily and do business without beingafraid of being cheated. MWEDOs adult education programis turning Maasai female participants into ‘citizens’.

The education program is acting as a form of ‘conscientiza-tion’ (Freire, 2000), creating spaces where women can chal-lenge the belief systems that uphold their oppression. 34

They are learning to read and write, and at the same timelearning their rights as members of society. As one womanfrom Longido village remarked “we have opened our eyes;now we know that [men & women] should be equal.” Whenasked if this was impacting her behavior, the woman replied,that yes, now “we educate our daughters,” and “we know thatwe feed our children, so everything that comes inside thehouse, we go to look for it so we have it.” 35 This highlightshow the education process is leading to changes in how womenthink about themselves personally and about what it means tobe a Maasai woman more broadly. 36 It also illustrates thecontinuous process of empowerment outlined in Figure 1, withinvolvement in MWEDO leading to changes in attitudes andbeliefs related to social norms, which is leading to increasedagency in terms of decisions about children and householdeconomics, as well as linking back into transformative changesin beliefs in gender norms. One woman articulated these con-nections:


If you are given to a man [in marriage] it will keep you from going toschool and getting education. If a girl gets education it will bring herown decision [for marriage] and she will be happy in her heart and willcome to get everything she wants.

Through the adult education programs women are alsolearning about the importance of land ownership, which seemsto have opened up many potential pathways for change.Whereas cattle ownership is still seen by many women as cul-turally male, land ownership is a fairly new concept to Maasaiwomen and men and thus easier for women to conceptualize assomething they can participate in. Additionally, since MWE-DO is promoting family land titles, which women seek outand have their names on together with their husbands, menwere often quite receptive and supportive of the process. 37

When asked about the potential for land ownership to bringabout change, women spoke at length about the possibilities,including the ability to grow their own food (or keep theirown cattle), to feed themselves and their children, and asdirectly linked to knowing their rights.

A widow dependent on her eldest son and living in Oltukaiexplained:

[Land ownership] will bring big benefits because if I get my own landand house I can wake up and do my own business you can get 2–3 days to do something. Long ago if you left for 2–3 days you wouldbe kicked out of the boma [homestead]. . . . [With land] you can buyyour own things and you won’t need to beg your husband [or son] any-more for everything—soap, clothes, even a spoon. 38

She also suggested that land ownership could reduce domes-tic violence against women, because women will know theirrights and men will be afraid to violate that,

Up until I get a land title, I will have received knowledge to get thisthing that on my own I own. When men are used to being the only onesto own things, they will know that it is my right to own things so it isnot just them [men] and they will be afraid of other things.

A young married woman with an abusive alcoholic husbandargued that if she had her own land, she would make her owndecisions to buy what she wanted and “not depend on or haveto beg for everything.” 39 Furthermore, she suggested thatland ownership would make it possible for her to speak atpublic meetings:

If I had a piece of land that I knew was mine and was put aside for mypurposes, I would be able to stand and speak at meetings. [Why? Whatis the difference?] It is just bitterness, that I don’t know how I will getmy needs met [that prevents me from speaking].

In Longido and Engare Naibor villages, the process ofobtaining land titles started much earlier, and while manywomen were able to articulate the benefits of land ownership,learned through MWEDO trainings, most were still withoutactual titles, and were frustrated with the process. As thequotes above suggest, it was the education about the processof obtaining land that brought knowledge of rights, even ifthe title to land was not yet actualized. In many cases the addi-tional training in these two villages through the work of theUCRT Women’s Forums highlighted the larger political issuesrelated to land, creating a political consciousness amongwomen as elaborated below.

Some women who did receive land titles articulated the ben-efits in a more political way, as linked to larger societal oppres-sion and land grabbing by the government and outsiders, aswell as freedom from their husbands. One woman with a titleexplained that land brought,

Change. Because we won’t come again to be oppressed by the leaders.They can’t sell again a place where we live. Also we get benefits becauseif I sell anything, the benefit is mine, it is no longer my husband’s. So

we can see we have a benefit. No one can claim it [the land] or tell youwhat to do [with it]. [Q: even if the land is in the name of the womantogether with her husband? He can’t take the profit?] No he can’t. Hedoesn’t know where it (the title) is because I went to get it myself! 40

In villages where MWEDO had worked longer there wasalso more discussion about the benefits of the economicempowerment projects. Women explained they were able tomake money through small business activities and this seemedto decrease beatings by their husbands. Several womenexplained that if they came home late, normally their husbandwould beat them, but if they brought food for the family thenhe could not beat them. Such shifts in economic power in thehousehold are likely leading to an increase in a woman’sagency as she contributes toward the household budget (asillustrated in Figures 3 and 4, and discussed above). This sug-gests that certain aspects of increased agency, such as eco-nomic decision-making, facilitated through involvement inMWEDO (indirectly by becoming a breadwinner), can havespecific impacts on transformative processes of empowerment.

Therefore, while the models may not prove direct linksbetween MWEDO involvement and changing beliefs (in socialnorms and violence), change seems to be happening throughincreased personal agency in the household. Qualitative datahighlight changing attitudes and beliefs among women partic-ipating in MWEDO, and these attitudes are much more visiblein the villages where MWEDO has been working longer andwhere the CLF is also active. The rhetoric around women’srights and land rights in these villages was stronger than inOltukai village, where most women did not know the term‘women’s rights’ and could not articulate what it meant. InOltukai, women also spoke a lot more about the customaryrules governing women’s participation, the only transforma-tive outcome that our model shows MWEDO involvement isaffecting directly (see Figure 4). Women in Oltukai in particu-lar explained that they did not stand at meetings because itwas not their place to do so. In contrast, in the villages whereMWEDO had been working much longer, women were wellversed on the discourse of women’s rights, and commentedstrongly and with pride, that of course they stand while speak-ing, why should they not! Though for some, this was only atMWEDO meetings. Speaking at village and government meet-ings was still a challenge. Change was seen as coming gradu-ally, linked to knowledge of rights, access to resources,education, and the example of local female leaders. Onewoman from Longido village explained her view in responseto a question about women speaking at village meetings:

Yes, women speak. But at meetings, many are ashamed to speak.Today, there are some women who are not afraid to speak, but othersare still afraid. . . . It is not [because of fear] that if I say this I will bebeaten, we just got used to being beaten, so today we are afraid tospeak in front of the elders because of respect [enkanyit]. 41 So youcan not stand, but some, these elected leaders, can stand. . . . But itis changing now. Those who have studied, they know how to standup for our rights. . . .. Those who can speak at meetings can say thata woman gets her rights, gets her land and to stand in meetings. 42

We explain the impact of MWEDO involvement on partic-ipation as linked to education about rights, but also as relatedto “empowerment by example.” Recall that our measurementfor MWEDO involvement includes not only participating inMWEDO projects but also attending MWEDO meetingsand speaking at them. At these meetings women are encour-aged not only to speak, but also to stand while speaking.When women come to Annual MWEDO Meetings in Arushafrom villages where women do not stand at meetings, they seeother Maasai women standing to speak. Women spoke atlength in interviews about the power of such experiences.


“These are women just like us,” they would say. “If they stand,why can’t we?” Many women begin by standing while speak-ing at MWEDO meetings, then take this behavior home withthem exemplifying empowerment by example.

(d) Qualitative findings beyond MWEDO: UCRT and the CLF

In this section we show how the work of the UCRT CLF iscontributing to women’s increased participation and changingbeliefs in social norms, and through both there is increasedaction for change at the societal level.

After the women’s forum was established in a village,women began to hold their own meetings to deliberate prob-lems and set agendas. They also received knowledge abouttheir legal rights through education and training from UCRTand partner organizations. One woman expressed the value ofthe CLF from a rights perspective:

Long ago, women had no worth/value. We were just like children!Long ago, women were very badly oppressed and there was nothingyou could say. If you spoke [at a meeting], the men would say, ‘ohthese are only children’s issues you are speaking of!’ Today we knowour rights; we have had our eyes opened from these seminars. So forme, I see that the oppression and exploitation of women has decreased.

The CLF also encourages recognition of women as leaders,able to participate in decision-making processes at the commu-nity level. This is changing perceptions of women’s role insocial leadership and their capacity to contribute to meetings,as expressed by a woman from Longido:

Now we have come out of the darkness, we have seen the light [kitiatwa ewang’an], because we can share a meeting with men. During dis-cussions amongst men now they say to us, ‘women we want to hearyour ideas.’ They [women] now stand and state their ideas.

Both men and women spoke at length about how the words ofwomen are now more officially recognized as important becauseof the CLF. Men are not only asking women to speak, but theyare listening to them and valuing what they say. It was the workof combining the male and female forums that facilitated thistransformation. Male leaders in Longido suggested that havingwomen on the CLF would help with leadership and unity withinthe community as a whole. One male leader stated,

Together we will be able to work better to help the community. First,women work with society more than men. A man can decide to sell thefarm, and the woman won’t agree because she knows she has childrento raise. Also, if we meet together (as a joint forum), they will contrib-ute more ideas regarding development. 43

When asked what the value was of bringing together the twoforums, one elder woman explained that there was “a big ben-efit because women were given a chance to put forth theirideas.” A younger married woman explained that, “it is help-ful because long ago we couldn’t stand in the middle of men,but now it helps me to become accustomed to standing in frontof men, speaking and being heard!” 44 Here we see more thanempowerment by example, but also empowerment throughexperience and changing expectations about the role of womenin society. It does not presume that empowerment is alwaysabout resistance (Mahmood, 2005) but rather highlights thesubtle ways that power plays out and can be transformedthrough social relations and individual behaviors.

The work of the CLF is therefore doing more than teachingwomen their rights and capabilities; it is also creating an insti-tutional setting where such knowledge can be transformed intoaction and the role of women in society can be re-thought.That women are officially given a seat at the leadership tablemeans they are given the right to discuss certain issues with

the larger community in an official capacity. Women havealways played an important role in mediating the decisionsof men in Maasai society. Some explained the behind thescenes role that women play in decision making processes withthe term, ‘olaigwanani lol-choni’ or the spokesman/leader ofthe bed. This refers to the words a man hears from his wife(or mother), when he goes home to think about an importantissue, or to ‘sleep on it’ before presenting a decision to a largermeeting. The man will then come to the meeting the followingday and state that he was advised in a particular way, suggest-ing that he dreamt or thought while sleeping, of what to say.All present will know that it was the wise words of his wifeor mother that lay behind the man’s decision. 45 The CLF isbringing this important role of women into the open. It is alsoenabling other issues particularly important for women, to beaddressed in official settings. In a group interview in Longidowomen discussed the work of the forum as:

To join women with leadership and to persuade other women regard-ing education, and to teach them their rights. To teach women thatthey have political rights, to be a [ward level] counselor, etc. in govern-ment.To change bad traditions, regarding women not inheriting, and to keepthose [traditions] that are good. And to know that women can own cat-tle, not only men. So if her children need something she can sell cattle,not only the men.

When asked if and how the forum was helping them, onewoman spoke about the forum as being an important spaceto address issues that affect women, such as violence:

I am really happy to have the women’s forum because women have, formany years now, been left behind, and the forum is a tool that will helpus address the problems that affect women. For instance, long ago ifyou were beaten or exploited [by your husband] you had no place togo. Today we can solve our problems by meeting and discussing ourconcerns together.

Even though the CLF is still new, in Engare Naibor, where ithas existed the longest, women have already begun working onspecific issues at the community level. One member (who wasalso a counselor for the area) explained that the women’s forumwas created to do the following things: deal with land rights,alcoholism, and to find the money to send children to secondaryschool. In fact, this particular forum succeeded at passing a locallaw banning the production and sale of homemade liquor, whichis one of the main causes of increased alcoholism in the area lead-ing to more incidences of domestic violence. They successfullyfought for land that was acquired illegally to be returned to fam-ilies/women and they were working on mechanisms to makesuch sales difficult to occur in the future. As one member stated,“we need to protect our land. If we sell all the land then wherewill we graze cattle? The main objective of the CLF is to protectland from being stolen.” They were also working to preventunregulated use of pastures by Kenyans crossing into Tanzania.

In places where the CLF is working, women spoke of it asthe most powerful ‘empowerment’ intervention. Economicswas important, but this was more important. Not only werethey learning their rights, they were building an official spacewithin which to activate those rights at the community level.Yet at the same time, economics remained salient. Some ofthe women participating in the forum requested assistance indeveloping economic groups. Others used the forum as a vehi-cle to create their own economic development projects.

6. CONCLUSION

Empowerment as a goal of individual and societal transfor-mation is difficult if not impossible to measure. Yet it is


important to understand the impacts of empowerment-baseddevelopment programs on both individual participants andsocieties at large (Nagar & Raju, 2003). Empowerment mustoccur at both levels to truly result in transformative change.In this paper we have outlined a way to measure empower-ment outcomes, while holding on to an understanding ofempowerment as an on-going and iterative process occurringalong multiple pathways across different socio-spatial scales.We have shown that women exposed to different NGOempowerment interventions, seem to be experiencing changesthat can be seen as part of a process of empowerment at indi-vidual and societal levels—in actions, beliefs, and attitudes.

Our models do not explain all the changes we are seeing,though the qualitative data suggest that many of the changescan be attributed to activities related to the NGO activity beinganalyzed. There are processes that we did not measure, orwhere our measurements did not work very well, such as: therole of church involvement; social psychological measures suchas friends and family who may serve as role models for per-sonal agency; and indicators of individual character such asresilience, willfulness, tenacity, personal strength, etc. Such fac-tors are difficult to get at in a survey, particularly in a differentcultural setting. The words from Maasai themselves regardingwhat it means to be empowered suggest that many of these per-sonal traits (e.g. willfulness, tenacity, strength) are importantcontributors, but also that education and awareness are neededfor women with such traits to realize their full potential. Ourqualitative data show this by highlighting the value of the adulteducation program, empowerment by example, and the workof the CLF in providing the resources for women to enhancepersonal and collective agency. 46

Our conceptual model draws from Kabeer’s framework thatempowerment implies increased agency and achievementswithin a particular context of resources and constraints. Wecontextualize this to reflect local power dynamics and socialnorms, including an appreciation of the different ways inwhich power plays out within households, across different gen-erations of women, between men and women, within the com-munity at large; and how the introduction of new resourcesfrom NGOS may contribute to changes across these dynamics.For instance, though older women often have more powerwithin the household, our study shows that through resourcesintroduced by NGOs, younger women are gaining the ability

to exercise new forms of power: power to make decisions inthe household; power within, to think differently about a hus-band’s right to beat them and to question other gendered soci-etal norms; and power-with others by standing up andspeaking at meetings and through the CLF to gain power tobe recognized officially for the power they have always had,and to push for societal changes. Some of these changes workto challenge customary norms, some of them work with cus-tomary norms to improve the lives of men and women (i.e.,keeping marriages together, preventing land grabs). MWEDOmembers have also put together their own ‘political forums’ soas to create the necessary conditions for power-with and to cre-ate more female political leaders.

Innovative grassroots organizations know that women’sissues cannot be isolated into separate compartments neatlytreated by individual projects. Women participating in educa-tion programs are learning about political rights and rights toland. As a result, they are gaining skills and simultaneouslytransforming their personal beliefs and attitudes. As womenattend MWEDO meetings they are learning about women’srights and the power of participation. As a result, they aregaining the strength to stand and speak at meetings, whichchallenges personal, household, and societal norms simulta-neously. The CLF is providing a space for women to engagedirectly with men as leaders, giving women a stronger voicein societal affairs and challenging societal gender norms andbehaviors.

We argue that it is through these connections across scales(from personal bodies, to household relations, to larger com-munities) and pathways (economic, political, and personal)that empowerment is happening in these villages. Women aremaking the connections themselves—taking their educationto their economic groups, using economic earnings to demandnew rights in the household, and making sure that they partic-ipate in decision making about their children, especially theirdaughters, to enable empowerment of the next generation ofwomen. Here we can see the way that empowerment is neces-sarily a relational and iterative process. One that involveschanging (and sometimes reinforced) subject positions in termsof what it means to be a Maasai woman (strong, capable,respectful, but also with a new set of rights); new concepts ofrights for men and woman, and new spaces to voice and actout these rights in ways that can lead to changes societally.

NOTES

1. The goal is stated as to “promote gender equality and empowerwomen,” and is to be achieved through eliminating gender disparity inprimary and secondary education. http://www.un.org/millenniumgoals/gender.shtml (accessed March 5, 2014).

2. This is common practice, Batliwala (2007) reminds us, fordevelopment practice relies on ‘buzzwords, catchphrases, and simplerhetoric, that can be used across disciplinary boundaries, cultural contexts,and the policy-academic divide.

3. A key aspect of magic bullet responses is the simplification of the fieldof intervention, and the promotion of deliverable, measurable outcomes.

4. UN gender empowerment measurements do look at multiple mea-surements simultaneously, but have been critiqued for providing onlyaggregate measurements that can be misleading and for being biasedtoward the most educated and economically advantaged members ofsociety (Cueva Beteta, 2006; Moghadam & Senftova, 2005). For instance,education is measured only as enrollment ratios at primary, secondary,

and tertiary schools, political participation is measured by parliamentseats and economic empowerment by looking at participation in formaleconomic spheres.

5. We cannot claim to be doing the sort of collaborative researchpromoted by transnational feminists (Nagar, 2008; Sangtin Writers &Nagar, 2006), though we do place a great deal of value on the women’sown views of change and struggle.

6. Words of Sofia Parkipuny (interview, 2011). She typed this up herselfin response to a series of open questions about what empowerment means.She herself came from a rural village background and worked withdifferent Maasai-based NGOs since completing her education. Shetragically passed away as a young mother in 2012.

7. Interview, Arusha 2011.

8. These include culturally sanctioned wife beating, forced child-mar-riages (often to much older men), men’s control of all property (cattle andcustody of children), and female genital cutting.

http://www.un.org/millenniumgoals/gender.shtml

http://www.un.org/millenniumgoals/gender.shtml


9. http://maasaiwomentanzania.tripod.com/mission.htm. AccessedNovember 17, 2013.

10. This comes from an interview with the executive director, 2009. Sheexplicitly commented on these being “programs” and not “projects”, asprojects are more finite and can change.

11. Through a scholarship fund and with MWEDOs own boardingschool.

12. Women are not dissuaded from pursuing their own individual titlesor titles for land to be held in common by a MWEDO women’s group.

13. The lead author observed one of these meetings, when MWEDO wasjust introducing the issue into Oltukai village, and the active involvementof men in the discussion was a vital part of gaining community support forthe issue.

14. This structure mimics the Tanzanian governance system to facilitateacceptance and collaboration.

15. This is related to the power and influence of the long standing MP forthe area and subtle but long-term tensions between the two dominantethnic groups in the district—Maasai and Arusha. The political tensionsmake it difficult for some NGOs to work in the area and have lead to asituation where the women are less aware and more isolated then womenin other villages, which may be more geographically isolated but have a lotof NGO activity. While this isolation itself could be a factor impactingwomen’s empowerment levels, the fact that changes have been happeningsince the start of MWEDO working in the village suggests they are notinsurmountable and rather provide an ideal context for looking at changeas related to NGO activity. In this paper we look only at the comparisonacross villages. Data were collected to analyze change in year 4 of thestudy and is still being analyzed, but preliminary analysis highlightschanges occurring in this village which echo our findings here.

16. Longido split from Monduli District in 2006, but did not receive adistrict council until 2007.

17. Other NGOs include: TEMBO (run by Canadians, working on girlseducation, HIV/AIDs education, preventing FGM, and micro-enterprisedevelopment and 27.9% of women interviewed in Longido are involved);LOSSIP (started by a local Maasai man, working on education andtraining), World Vision, and other smaller organizations.

18. Only 3.3% of women interviewed in Kimokowo were involved inTEMBO.

19. With Shelly Grabe, University of California-Santa Cruz, Departmentof Psychology.

20. Drawing from the literature in human geography on scale, we do notsee these as fixed spatial units, but rather as relational categories or asnested vertical spatial organization of social relations (Brenner, 2001), thatare socially constructed. As Swyngadow states, (1997: 141, cited in Leitner& Miller, 2007: p. 117): “scales ‘are perpetually redefined, contested, andrestructured in terms of their extent, content, relative importance, andinterrelations’ (Swyngedouw, 1997, p. 141).”

21. While Maasai women, by custom, do not officially ‘own’ cattle, theyare given cattle at marriage by their husbands, and some come with cattle(from their father) to the marriage. The women often refer to these cattleas ‘theirs’. Additionally some women today are acquiring their own cattle.We did not specify the difference in this survey so can only speak to cattleownership as a sign of wealth.

22. See Appendix for complete table of variables included in theconstruction of individual scales.

23. As most Maasai women are illiterate, they have difficulty calculatingtheir exact age. Maasai men everywhere, and the women in our study area,are recognized as belonging to age-sets, determined by when they werecircumcised. Using age-set distinctions proved a more reliable method toget the general age of a woman.

24. This knowledge comes from the lead author’s long-standing ethno-graphic research with Maasai communities across Tanzania and Kenya.

25. See description in qualitative findings, and in Appendix.

26. The question that asked if a woman had experienced violence fromher husband in the last year had 32 missing responses. The question thatasked about ever experiencing such violence had little variation (83%reporting yes).

27. One of the variables in the scale include being beaten “for no reason”,which thus covers occurrences of violence beyond disciplining, and forwhich most women did not believe was acceptable. We frame the results as‘violence for discipline’, since beating as a form of discipline is such acommon and long-held belief for Maasai women that changes in it suggestimportant and empowering changes for women.

28. All scales and survey questions used to represent measures ofresources and controls, personal agency, and outcomes, are described indetail in Appendix.

29. For example, quotations appear in qualitative interviews of womenexplaining increased agency; the lead author also noted observed changesin participation or in certain women’s behavior in the household andbroader community.

30. Interview with a woman in Oltukai, March 2, 2010.

31. Interview with elder woman, Kimokowa, May 2009.

32. For instance, there was one household where women were told by themale head of the household that they could not prepare maize in aparticular way because it used too much maize at one time and thereforecost more money than the more efficient technique of grinding and thencooking the maize. In another situation a woman’s husband (abusive andalcoholic) refused to let her take her extremely sick child to the hospitaleven when she secured her own funding.

33. This highlights the ways that age can act as a control, in this casewith younger women gaining power through MWEDO to have morecontrol in decisions regarding marital sex.

34. We prefer this term as it draws from the powerful work of PauloFreire which is based in adult education and critical consciousness as aform of liberation in a developing world context, to ‘consciousness raising’which draws from the work of Western Feminists and is more directlyactivist oriented.

35. Interview, Longido June 2011.

36. The shifting subject positions of women as related NGO involvementis complex and beyond the scope of this paper. Women have begun to seethemselves as citizens with rights they were previously unaware of. Yet thishas not diminished their feelings about what it means to be a properMaasai women, which is often related to respecting and receiving respectfrom men, and not challenging certain cultural norms. For instance, one

http://maasaiwomentanzania.tripod.com/mission.htm


elder woman used her economic independence and success to give herhusband the money to pay for her bride price, which he had never paid. Toher, this was not at all an ironic or contradictory act. She did not feel like acomplete woman, because this important cultural act was incomplete andshe was empowered economically and socially within the relationship tomake it happen, even though it was an act that many women saw as partof their subordination, as described above.

37. Some women, widows in particular, are requesting individual titles.There are also MWEDO groups that are requesting group titles.

38. Interview with widow, Oltukai July 2009.

39. Interview with young married woman, Oltukai July 2009.

40. Interview with widow, Longido May 26 2009.

41. The Maasai notion of respect (enkanyit) can inhibit some fromspeaking in front of others because of “fear” associated with respect andsocial norms governing behavior see Goldman and Milliary (2014) andHodgson (2001).

42. Interview with married woman, Longido May 26, 2009.

43. Interview with male Customary Leader, Kimokowa, Longido June 2010.

44. Interview with five women and five men (all but one, members of theforum), June 24, 2010.

45. I thank Elifuraha Laltika and Alais Morindat for clarification on this term.

46. Here, “agency” refers to shifts if official presence and actions and notnecessarily resistance to subordination or complete political change (seealso Saba Mahmood’s Politics of Piety for an in depth discussion of needto complexify understandings of agency”).

REFERENCES

Batliwala, S. (2007). Putting power back into empowerment: OpenDemocracy. http://www.opendemocracy.net/article/putt-ing_power_back_into_empowerment_0.

Batliwala, S. (2010). Taking the power out of empowerment: Anexperiential account. In A. Cornwall, & D. Eade (Eds.), Deconstructingdevelopment discourse: Buzzwords and fuzzwords (pp. 111–121). Ox-ford: Oxfam.

Brenner, N. (2001). The limits to scale? Methodological reflections onscalar structuration. Progress in Human Geography, 25(4), 591–614.

Brockington, D. (2001). Women’s income and the livelihood strategies ofdispossessed pastoralists near the Mkomazi Game Reserve, Tanzania.Human Ecology, 29(3), 307–338. http://dx.doi.org/10.1023/a:1010906715682.

Cueva Beteta, H. (2006). What is missing in measures of Women’sEmpowerment?. Journal of Human Development, 7(2), 221–241.

Freire, P. (2000). Pedagogy of the oppressed (30th anniversary ed.). NewYork: Continuum.

Goldman, M., & Milliary, S. (2014). From critique to engagement: Re-evaluating the Participatory model with Maasai in Northern Tanzania.Journal of Political Ecology, 21, 408–423.

Harcourt, W. (2009). Body politics in development: Critical debates ingender and development. London: Zed Books.

Hodgson, D. L. (1999). Women as children: Culture, political economy,and gender inequality among Kisongo Maasai. Nomadic Peoples, 3(2),115–130.

Hodgson, D. L. (2000). Rethinking pastoralism in Africa: Gender, culture &the myth of the patriarchal pastoralist. Oxford: James Curry.

Hodgson, D. L. (2001). Once intrepid warriors: Gender, ethnicity, and thecultural politics of Maasai development. Bloomington: Indiana Univer-sity Press.

Hodgson, D. L. (2002). Women’s rights as human rights: Women in Lawand Development in Africa (WiLDAF). Africa Today, 49(2), 3–26.http://dx.doi.org/10.1353/at.2003.0010.

Hodgson, D. L. (2011). Being Maasai, becoming indigenous: Postcolo-nial politics in a neoliberal world. Bloomington: Indiana UniversityPress.

Igoe, J. (2003). Scaling up civil society: Donor money, NGOs and thepastoralist land rights movement in Tanzania. Development andChange, 34(5), 863–885. http://dx.doi.org/10.1111/j.1467-7660.2003.00332.x.

Kabeer, N. (1999). Resources, agency, achievements: Reflections on themeasurement of women’s empowerment. Development and Change,30(3), 435–464.

Kabeer, N. (2003). Reversed realities: Gender hierarchies in developmentthought (1994). New York: Verso.

Kabeer, N. (2005). Gender equality and women’s empowerment: A criticalanalysis for the third millennium development goal. Gender andDevelopment, 13(1), 13–24.

Lazaro, H. (2008). Violence against women on the rise, Arusha Times.http://business.highbeam.com/3548/article-1G1-189913161/violence-against-women-rise. Retrieved from http://business.highbeam.com/3548/article-1G1-189913161/violence-against-women-rise.

Leitner, H., & Miller, B. (2007). Scale and the limitations of ontologicaldebate: A commentary on Marston, Jones and Woodward. Transac-tions of the Institute of British Geographers, 32(1), 116–125.

Mahmood, S. (2005). Politics of piety: The Islamic revival and the feministsubject. Princeton: Princeton University Press.

Mahmud, S., Shah, N. M., & Becker, S. (2012). Measurement of women’sempowerment in rural Bangladesh. World Development, 40(3),610–619. http://dx.doi.org/10.1016/j.worlddev.2011.08.003.

Malhotra, A., Schuler, S. R., & Boender, C. (2002). Measuring women’sempowerment as a variable in international development: BackgroundPaper Prepared for the World Bank Workshop on Poverty and Gender:New Perspectives. http://siteresources.worldbank.org/INTGENDER/Resources/MalhotraSchulerBoender.pdf.

McCabe, J. T. (2003). Sustainability and livelihood diversification amongthe Maasai of northern Tanzania. Human Organization, 62(2), 100–111.

Mitzlaff, U. v. (1988). Maasai women: Life in a patriarchal society: Fieldresearch among the Parakuyo, Tanzania. Munchen: Trickster Verlag.

Moghadam, V. M., & Senftova, L. (2005). Measuring women’s empower-ment: participation and rights in civil, political, social, economic, andcultural domains. International Social Science Journal, 57(184), 389–412.

Nagar, R. (2008). Languages of collaboration. In P. Moss, & K. FalconerAl-Hindi (Eds.), Feminisms in geography: Rethinking space, place, andknowledges (pp. 120–129). Lanham, MD: Rowman & Littlefield.

Nagar, R., & Raju, S. (2003). Women, NGOs and the contradictions ofempowerment and disempowerment: A conversation. Antipode, 35(1),1–13. http://dx.doi.org/10.1111/1467-8330.00298.

Parpart, J. L., Rai, S. M., & Staudt, K. (2002). Rethinking em(power)-ment, gender and development: An introduction. In J. L. Parpart, S.M. Rai, & K. A. Staudt (Eds.), Rethinking empowerment: Gender anddevelopment in a global/local world (pp. 3–21). New York: Routledge.

Rowlands, J. (1995). Empowerment examined. Development in Practice,5(2), 101–107. http://dx.doi.org/10.1080/0961452951000157074.

Rowlands, J. (1997). Questioning empowerment: Working with women inHonduras. Oxford: Oxfam Publishing.

Rowlands, J. (1998). A word of the times, but what does it mean?Empowerment in the discourse and practice of development. In H.Afshar (Ed.), Women and empowerment: Illustrations from the thirdworld (pp. 11–34). Houndmills, Basingstoke, Hampshire: Macmillan.

Sangtin Writers, A., Nagar, R., et al. (2006). Playing with fire: Feministthought and activism through seven lives in India. Minneapolis, MN:University of Minnesota Press.

Smith, N. M. (2014). Gender and livelihood diversification: Maasaiwomen’s market activities in Northern Tanzania. Journal of Develop-ment Studies, 52, 1–14, Epub ahead of print.

Spencer, P. (2003). Time, space, and the unknown: Maasai configurations ofpower and providence. New York: Routledge.

Swyngedouw, E. (1997). Neither global nor local: ‘Glocalization’ and thepolitics of scale. In K. R. Cox (Ed.), Spaces of globalization: Reassertingthe power of the local (pp. 137–166). New York: Guilford Press.

Talle, A. (1987). Women as heads of houses: The organization ofproduction and the role of women among pastoral Maasai in Kenya.Ethnos, 52(1–2), 50–80.

http://www.opendemocracy.net/article/putting_power_back_into_empowerment_0

http://www.opendemocracy.net/article/putting_power_back_into_empowerment_0

http://refhub.elsevier.com/S0305-750X(14)00270-8/h0010








http://dx.doi.org/10.1023/a:1010906715682

http://dx.doi.org/10.1023/a:1010906715682

























http://dx.doi.org/10.1353/at.2003.0010





http://dx.doi.org/10.1111/j.1467-7660.2003.00332.x

http://dx.doi.org/10.1111/j.1467-7660.2003.00332.x












http://business.highbeam.com/3548/article-1G1-189913161/violence-against-women-rise











http://dx.doi.org/10.1016/j.worlddev.2011.08.003

http://siteresources.worldbank.org/INTGENDER/Resources/MalhotraSchulerBoender.pdf

http://siteresources.worldbank.org/INTGENDER/Resources/MalhotraSchulerBoender.pdf















http://dx.doi.org/10.1111/1467-8330.00298






http://dx.doi.org/10.1080/0961452951000157074





























Talle, A. (1988). Women at a loss: Changes in Maasai pastoralism and theireffects on gender relations, Thesis/dissertation (deg). Dept. of SocialAnthropology, University of Stockholm, Stockholm.

Wangui, E. (2008). Development interventions, changing livelihoods, andthe making of female Maasai pastoralists. Agriculture and HumanValues, 25(3), 365–378.

Table 5. Scale construction, survey questions, and distributions

Resources and controls Alpha Survey questions Coding for scale % Scoring 1

MWEDO involvement 0.573 Number of MWEDO meetings attended: village/general assembly?

0 = none, 1 = 1 or more 23.1

Number of MWEDO meetings attended: meeting ofMWEDO members in the village?

0 = none, 1 = 1 or more 51.2

Number of MWEDO annual meetings attended inArusha?

0 = none, 1 = 1 or more 30.0

Speaks at MWEDO meetings? 0 = no, 1 = yes 30.0Participated in MWEDO economic empowerment/business?

0 = no, 1 = yes 45.0

Participated in MWEDO adult education? 0 = no, 1 = yes 82.5

Husband Social control 0.592 Husband always wants to know where she is? 0 = no, 1 = yes 66.0Husband tries to keep you from seeing your friends? 0 = no, 1 = yes 11.9Husband tries to restrict contact with your family ofbirth?

0 = no, 1 = yes 5.7

Husband insists on knowing where you are at alltimes?

0 = no, 1 = yes 65.4

Husband tries to keep you from going to the market? 0 = no, 1 = yes 7.6Husband tries to keep her from attendingcelebrations?

0 = no, 1 = yes 15.7

% Scoring

Formal education Attended school? 0 = did not attend school 75.9Highest level of education achieved? 1 = primary 10.8

2 = completed primary 10.83 = some secondary 0.66 = technical school 1.9

% Scoring 0

Number of cattle How many cattle do you own (have)? 0-300 5.2

% Scoring 1

Breadwinner Amount of money she brings in relative to husband? 0 = less than husband, 1 = about thesame or more than husband

36.3

% Scoring

Age category Woman’s age-set? 0 = younger (Njunjulai, Ngali/Ndyamaragi, Mbuluku)

54.4

1 = older (Isusan, Mosogiro, Mugambo,Ngaimug, Isikanyi)

45.6

Personal agency(household decisionmaking)

Alpha Survey questions Coding for scale % Scoring 1

(continued on next page)

APPENDIX A

See Table 5.





Table 5 (continued)

Resources and controls Alpha Survey questions Coding for scale % Scoring 1

Total 0.583Cattle Subscale 0.749 Do you make cattle decisions? 0 = no, 1 = yes 32.7

You make the final decision in buying livestock? 0 = no, 1 = yes 13.1You make the final decision in selling livestock? 0 = no, 1 = yes 12.5

Land subscale 0.877 You have control of plot? 0 = no, 1 = yes 23.8Your name is on the title? 0 = no, 1 = yes 23.1

% Scoring

Number of acres officially owned? 0 = none 71.9(.05 thru 4 = 1) 16.9(5 thru 9 = 2) 6.3(10 thru highest = 3) 5.0

% Scoring 1

HH Economic Subscale 0.560 You make the final decision in buying food items? 0 = no, 1 = yes 55.0You make the final decision in buying small HHitems?

0 = no, 1 = yes 79.4

You make the final decision in buying clothes forchildren?

0 = no, 1 = yes 35.6

Child Subscale 0.765 You make the final decision in child health care? 0 = no, 1 = yes 60.0You make the final decision in child education? 0 = no, 1 = yes 41.3You make the final decision in child marriage? 0 = no, 1 = yes 30.6

Sexual Subscale 0.834 You make the final decision on use of contraception? 0 = no, 1 = yes 74.7You make the final decision on the number ofchildren?

0 = no, 1 = yes 69.3

You make the final decision to have or not to havesex?

0 = no, 1 = yes 81.8

Outcomes (social normsand practices)

Alpha Survey Questions Coding for scale % Scoring 1

Political participation 0.832 Speaks at governance meetings? 0 = no, 1 = yes 7.5Stands to speak at governance meetings? 0 = no, 1 = yes 6.9Speaks at village meetings? 0 = no, 1 = yes 29.4Stands to speak at village meetings? 0 = no, 1 = yes 25.6

Non-traditional GenderBeliefs

0.470 Men should share in household tasks? 0 = no, 1 = yes 16.3

Woman should choose man to marry? 0 = no, 1 = yes 75.6Woman should expect equal freedom? 0 = no, 1 = yes 38.1

Non-acceptance ofdiscipline

0.615 If a woman does not complete her housework tohusband’s satisfaction it is (not) ok for her husband tohit her?

0 = no, 1 = yes 64.4

If a woman made an error regarding livestock it is(not) ok for her husband to hit her?

0 = no, 1 = yes 10.6

If a woman disobeys or disrespects her husband it is(not) ok for him to hit her?

0 = no, 1 = yes 10.6

If a woman refuses to have sex with her husband it is(not) ok for him to hit her?

0 = no, 1 = yes 16.5

If a woman asks her husband if he has other lovers it is(not) ok for her husband to hit her?

0 = no, 1 = yes 87.3

If a husband suspects his wife has a lover it is (not) okfor him to hit her?

0 = no, 1 = yes 61.2

If a husband finds out that his wife has a lover it is(not) ok for him to hit her?

0 = no, 1 = yes 3.1


ScienceDirectAvailable online at www.sciencedirect.com

Innovative Grassroots NGOS and the Complex Processes of Women’s Empowerment: An Empirical Investigation from Northern Tanzania

Documents