Emic definitions of empowerment for just development: learnings from Kenya

Abstract

This mixed-methods study examines how men and women of different ages engaged in agriculture in Kilifi and Kiambu counties in Kenya define empowerment and describe empowerment pathways, and how empowered people are perceived by their families and communities. It also examines how emic understandings of empowerment—definitions which originate from within a given cultural context—compare to quantitative empowerment measures based on externally defined indicators from the Project-level Women’s Empowerment in Agriculture Index (pro-WEAI). The study design used mixed methods, with qualitative data collected through focus group discussions, life histories, and community profiles, and quantitative data collected through the pro-WEAI survey tool. Although Pro-WEAI results indicated that women and men in the study communities have achieved gender parity and experience a similar level of empowerment, qualitative findings suggest that experiences of empowerment don’t necessarily fit with Index scores. Rather, conceptualizations of empowerment by people of different ages and genders are highly diverse. Participants of all genders suggested that women’s empowerment is more limited than men’s, as women who are not. Subordinate to men risk being rejected or stigmatized by their communities. Qualitative findings suggest that empowered men are seen as “desirable” by their communities, whereas empowered women are “openly admired” and “treated as a threat/feared” in equal turn. Furthermore, results revealed that the under-involvement of men in empowerment initiatives has contributed to feelings of resentment and neglect. These findings suggest that capturing and accounting for multiple and diverse emic definitions of empowerment for different social groups, which lies beyond the scope of standardized quantitative measures, is essential to measuring and supporting empowerment in ways that are valuable and recognizable to the target group(s) of a given development initiative. Finally, results show that to support the effective empowerment of women and men, there is a need to include men, integrate safeguards for backlash against empowered women, and transform norms that hinder the acceptance of empowered women by their families and communities.

Introduction

Efforts to define, measure and operationalize women’s empowerment have received increasing attention from developmental researchers and practitioners in recent years. This includes an ongoing effort to develop standardized measures of women’s empowerment to evaluate and track the global progress toward Sustainable Development Goal 5—‘Gender equality and the empowerment of all women and girls’. However, developing a single standard to define and measure (women’s) empowerment in the incredible diversity of contexts and cultures around the world presents a major challenge. Thus, the operationalization of such standardized understandings of empowerment has been met with notable criticism. For the purposes of this paper, we focus on two key critiques: (1) empowerment is not conceptualized or realized in the same ways for everyone everywhere, and therefore imposing a one-size-fits-all, top-down definition or metric of empowerment is antithetical to empowerment itself; and (2) a narrow focus on empowerment for just one group (i.e., women) can overlook or overshadow the (dis)empowerment of other marginalized groups.

The work presented here aims to contribute to building more textured, nuanced, and context-specific empowerment approaches by and for diverse age and gender groups. To do so, this paper presents empirical evidence and ‘emic’ perspectives from men and women of two different age groups (18–35 years old and 35 +) from four communities in two counties in Kenya. An ‘emic’ understanding of empowerment is defined according to the internal values, understandings, and worldview of the culture or community being studied. Using Kabeer’s (1999) resources-agency-achievements framework for a “useful definition of empowerment that takes into account the nexus of the different conceptualizations of the concept which may be applicable across contexts” (p. 62), this study aims to answer the following questions:

• How do local communities define empowerment? How do these definitions vary by age and gender within each case study?

• How do the communities conceptualize pathways to this locally defined empowerment?

• How does the lived experience of empowerment processes compare and contrast to local conceptualizations?

• How are women and men who are considered empowered viewed and treated by their families, spouses, and communities?

Conceptualizing empowerment

Resources, agency, and achievement

While a lively discourse about the operational definition of empowerment is ongoing, this paper adopts Kabeer’s foundational (1999, p. 435) definition: “the process by which those who have been denied the ability to make strategic life choices acquire such an ability”. We adopt this definition for its flexible approach which allows for significant contextual variation. Kabeer’s understanding of empowerment is based on three overarching and interrelated dimensions which can be adjusted and triangulated to fit diverse empowerment contexts: resources (pre-conditions), agency (process), and achievements (outcomes):

• Resources: Material, human and social resources which enable one to make choices.

• Agency: The ability to define one’s goals and act upon them.

• Achievements: Well-being outcomes.

These three dimensions are indivisible; if an indicator of empowerment does not speak to an increase in all three, it is an invalid measure (Kabeer 1999). While Kabeer’s text specifically refers to women’s empowerment, her definition is both encompassing and flexible enough in scope to be used for men as well. Therefore, we use her definition not only to frame women’s empowerment, but also empowerment of marginalized men.

Why emic definitions of empowerment?

Leading universal benchmarks for empowerment, such as the gender empowerment measure (GEM) developed by the United Nations Development Programme (UNDP), have been criticized for their Western and especially Eurocentric bias, which “does not do justice to the context specificity of empowerment, especially with regards to the means or process available and the specific local barriers one may face” (Adjei 2015; Völker and Doneys 2021, p. 126). For example, a woman may score high on GEM empowerment indicators such as participation in professional economic activities and decision-making, but may nevertheless remain disempowered if her socio-cultural context does not allow her to exercise strategic life choices (Adjei 2015). The imposition of externally defined indicators of empowerment may lead to empowerment outcomes that look like empowerment to the implementing agency, but may not feel like empowerment to the participants.

Different studies have pointed to a problematic mismatch between ‘increases’ in empowerment and participants’ desired empowerment outcomes (Adjei 2015; O’Hara and Clement 2018). For example, a program focused on economic empowerment may ‘successfully’ increase women’s incomes, but also increase her labor burdens in a culture which values leisure time over money, making the venture a failure from an emic perspective (Völker and Doneys 2021). To achieve meaningful empowerment outcomes, a marginalized community member does not just need to be seen as empowered, they need to feel empowered in order to have the confidence to deviate from disempowering norms and act upon their own strategic life choices (Kabeer 1999; conceptualized as ‘power within’ in Rowlands 1997). This requires understanding and working within local value systems (Kinati et al. 2022). For example, a Western conceptualization of individualistic empowerment may not be desirable to women in a cultural context where family togetherness is highly valued (Kabeer 1999; Galiè and Farnworth 2019).

Understanding and addressing context-specific barriers and entry points to empowerment is therefore key to the success of empowerment-related interventions. Moreover, initiatives which fail to adequately capture, account for, and ameliorate unjust social structures and systems not only risk impotence to move the needle on empowerment, but may actually exacerbate or perpetuate existing systems of inequality (Slegh et al. 2013; O'Hara and Clement 2018; Meinzen-Dick et al. 2019; Völker and Doneys 2021). For example, an increase in women’s income (especially if it means a woman earns more than her male relatives) can lead to severe backlash, including gender-based violence and ostracism. On the other hand, such disruptions to traditional social systems and hierarchies “can be both necessary and positive” (Völker and Doneys 2021, p. 126). As such, empowerment interventions may call for complementary, emic approaches to ensure that the benefits of the intervention outweigh the drawbacks, thus making empowerment a worthwhile and desirable pursuit for the marginalized members of a community.

Despite their critical role in developing appropriate and effective empowerment strategies, emic understandings of empowerment remain underexplored. A 2014 review of empirical evaluations of women’s and girls’ economic empowerment (Pereznieto and Taylor 2014) found that less than 25% of studies included a strong gender analysis, suggesting that the local context was poorly understood. This led to methodology based on misinterpretations and misconceptions which did little to serve those who were supposed to have been empowered. Furthermore, the review found that there are especially few empirical evaluations of women’s and girls’ economic empowerment that gather information on emic definitions from more than one region or community. This indicates a lack of robust data to inform regional, national, or international empowerment benchmarks that are sensitive to diverse empowerment experiences. The findings presented in this paper aims to contribute to filling this gap.

Beyond women’s empowerment

Historically, conversations around empowerment have been focused on women, and for good reason; on a global level, systems of (dis)empowerment are characterized by a prevalence of patriarchal hierarchies in which women are subordinated. However, in recent years, there has been a pushback against this simplistic understanding of gender and empowerment. A binary view of empowerment tends to gloss over the intersectional nuances of power structures in which gender is just one axis of identity that can determine one's social status among many others (e.g., age, race, class, caste, religion, ethnicity, ability, rurality) (Corrêa 2010; Gill and Pires 2019; Leder and Sachs 2019; Trivelli and Morel 2021). In some contexts, gender may not be among the most important factors of (dis)empowerment. In these cases, a top-down imposition of a ‘gender agenda’ by development agencies may mask more nuanced realities while perpetuating colonial attitudes which clumsily prescribe assumed meaning to local power structures (Frewer 2017; Gill and Pires 2019). Furthermore, the binary conceptualization of women as disempowered and men as empowered forms a biased narrative in which women are victims, men are oppressors, and development agencies are saviors (Corrêa 2010). In reality, gendered experiences of (dis)empowerment are often complex, multiple, and even contradictory in nature; in some contexts, for example, men may observe disempowerment while women perceive the opposite (Petesch et al. 2018a).

In practical terms, failing to include both women and men in empowerment initiatives risks that resulting empowerment mechanisms will be insufficiently inclusive and ineffective. Not only may disempowered men be insufficiently targeted (e.g., Tesfaye et al. 2022), a failure to engage men can also impede advances in women’s empowerment, as men’s involvement is seen as critical to the acceptance and support of empowerment initiatives (Pereznieto and Taylor 2014). However, Cornwall and Rivas (2015) warn that an uneven focus on “empowering women and girls” and “engaging men and boys” is also a poor and overly binary strategy for ending gender-based discrimination. Moreover, it is impossible to understand the full context for points of entry as well as barriers to empowerment without understanding both women’s and men’s roles, priorities, and perceptions in the given social setting. Yet, Pereznieto and Taylor’s (2014) review of empirical evaluations of women’s and girls’ economic empowerment found that many studies do not directly include men at all. This represents a major gap in the literature.

This study therefore seeks to fill this gap by including emic explorations of empowerment for women as well as men, but also for different age groups. Age can play a major role in experiences and conceptualizations of (dis)empowerment which may be critical for effective and inclusive positive change, and which are intersectional with gender considerations. For example, in a study of emic definitions of women’s empowerment in Laos, Vietnam, and Myanmar, Völker and Doneys (2021) found that middle-aged women had much broader understandings of empowerment than older and younger generations, as understandings of gender have changed over time. Considerations of youth empowerment have received increasing attention from development spaces in recent years (Elias et al. 2018; Rietveld et al. 2020; Petesch et al. 2022). However, empirical evidence of empowerment disaggregated by age is uncommon. A review of 254 empirical evaluations of women’s and girls’ economic empowerment found that just 21% explored empowerment for adolescent girls, and most lacked age disaggregated data overall (Pereznieto and Taylor 2014).

Methodology

Study sites

According to the UNDP’s Gender Inequality Index (GII), Kenya has a relatively high level of gender inequality compared to other countries around the world, ranked 128 out of 191 countries (UNDP 2022). This low rating is based on Kenya’s relatively high maternal mortality ratio (342 deaths per 100,000 live births) and adolescent birth rate (64.2 births per 1,000 women ages 15–19), and relatively low representation of women in parliament (23.2% of seats held by women). In addition to the gaps highlighted by the GII, gender gaps also persist in terms of literacy (Akala 2019), income, and access to credit (World Economic Forum 2015; Kivuva and Kinuthia 2021).

While inequality persists, Kenya has made some important political inroads toward gender equality over the last three decades. These include gender-mainstreaming in the Constitution of Kenya and other pieces of legislation, policies and programs, including the establishment of the National Gender and Equality Commission (NGEC). However, customary traditions and cultural beliefs continue to pose barriers to achieving gender parity in key human development indicators (Kivuva and Kinuthia 2021).

Gender is a key factor of marginalization in Kenya, but it is far from the only factor. Other key factors include age (including high stunting and wasting rates for children) and region/rurality (World Bank 2018). For those living in the arid and semi-arid lands (ASAL), development challenges are particularly severe due to poverty and vulnerability factors stemming from intense and prolonged drought associated with climate change (ibid.).

This study took place in Kenya’s Kilifi and Kiambu counties, both of which are ASAL regions, as shown in Fig. 1. In Kiambu, data was collected in Thika and Limuru administrative units, while in Kilifi data was collected in Magarini administrative unit. To uphold participant confidentiality, village names have been omitted.

Fig. 1

Map of study area (by Victor Nyamolo)

Kilifi County, located on Kenya’s east coast, has high rates of poverty (71.7%) and food insecurity (67%). Rain-fed agriculture is the primary livelihood source, contributing 52.7% of household income and employing over half the female population, putting the local population at heightened risk from climate change (GoK 2013). Key commodities are cassava, chili, local poultry, and dairy cattle. Despite the Country’s commitment to eliminate violence against children and gender discrimination through legislation and institutions, experience on the ground indicates that girls in Kilifi face double discrimination due to their gender and age (County Government of Kilifi 2018). In Magarini, the sub-county where data was collected, census data indicates that 25.64% of the total population of Magarini are between 18 and 35 years old (the defined age range for the younger group of participants in this study). The majority of the population belongs to the Giriama ethnic group, the dominant ethnic group in Kilifi County.

Kiambu County is located in the Central Highlands of Kenya, close to the capital city Nairobi. Kiambu’s population density is much higher than that of Kilifi (952 versus 116 people per square feet, respectively) (GoK 2019). Most of the population in Kiambu County is ethnically Kikuyu, a majority ethnic group. Kiambu also has a much larger economy than Kilifi, ranking in the top three Kenyan counties in terms of gross country product (GCP) in 2017 as well as GCP growth from 2013 to 2017 (KNBS 2019). However, key informant interviews (KIIs) suggest that the economy in Kiambu has been negatively impacted by prolonged drought and the Covid-19 pandemic, which have simultaneously affected income generating activities and contributed to the rising cost of agricultural inputs, making it increasingly difficult for locals to pursue agricultural livelihoods.

The two administrative units chosen as study sites in Kiambu County are characterized by dissimilar economic conditions, as Thika Town is a peri-urban area economically tied to nearby Nairobi (Muiruri and Odera 2018), while Limuru is a comparatively rural and agricultural area. Thika’s ecology is not well-suited for agriculture, as the soils are low-fertility sand and clay, which are furthermore dissected and easily eroded (County Government of Kiambu in Collaboration with Ministry of Land Housing and Urban Development 2015). In contrast, Limuru is an agricultural unit with major tea estates and a vibrant horticultural sector (County Government of Kiambu 2018).

Sampling frame

Focus group discussion participants were selected from existing farmer groups in Kilifi and Kiambu. Most of these groups were mixed gender, but had more women members than men, as most of the groups were originally for women only. In each focus group discussion, participants were introduced to a leveled empowerment typology (Sect. “Pathways to empowerment”) and asked to express their own level of empowerment using a secret ballot. Based on these anonymous self-assessments (as well as willingness to participate), 16 out of 55 total focus group discussants were selected for life history interviews. The participants that assessed themselves the highest and the lowest in terms of empowerment were selected for the life histories, and thus represented the empowered and disempowered from that specific focus group.

Data collection

The study was designed as a mixed methods study and both quantitative and qualitative data were collected. The results presented here are primarily based on the qualitative data, while the quantitative survey results are used to examine how the emic definitions of empowerment explored in the qualitative data map onto the externally defined quantitative indicators. Data were collected from March 17 to 28, 2022.

The qualitative tools used for the study were adapted and amended based on tools from GENNOVATE (Petesch et al. 2018b) and the Gender, Agriculture and Assets Project, phase 2 (GAAP2)¹ and consist of three parts:

• 1. Community profile: through key informant interviews with men and women community leaders, this module provides an overview of the communities. Three community profiles were conducted (Table 1). This data was used to provide background information and context to situate the findings from the other quantitative and qualitative tools described below.

• 2. Focus group discussions (FGDs): FGDs were used to capture local definitions of empowerment, conceptualizations of empowerment pathways, and attitudes toward empowered men and women. A total of eight FGDs were conducted, disaggregated by age and sex (2 younger men, 2 younger women, 2 older men, 2 older women). On average, seven people participated in each FGD. Broken down by gender and age group, there was an average of 6 participants for younger men’s FGDs, 7 for younger women, 8 for older men, and 8 for older women. These participants were asked to anonymously self-identify their own level of empowerment based on the steps in the Ladder of Power and Freedom Tool.

  • a. Ladder of Power and Freedom: To explore emic pathways to empowerment, FGDs included a participatory tool called the Ladder of Power and Freedom, which invites participants to define how many steps a disempowered person must climb to reach total empowerment, describe what empowerment phases look like in practice in the local context, and name the local mobility factors and barriers that may help or hinder movement up the ladder.

• 3. Life histories (LHs): This semi-structured interview format provides insight into the circumstances in empowered and disempowered individuals’ lives that have impacted their empowerment pathways over time and which influence their views on empowerment. A total of 16 LH interviews (8 men and 8 women) were conducted, with participants selected to represent varied levels of empowerment. This sampling was based on participants’ self-identified level of empowerment as privately expressed in focus group discussions.

Table 1 Community profile sample

The quantitative data were collected through a structured questionnaire² composed of two parts: a household survey which gathered basic information about the households and their members (e.g. age, education, employment, assets), and an individual questionnaire based on the Project-level Women’s Empowerment in Agriculture Index (pro-WEAI) (Malapit et al. 2019) with some indicators from the market inclusion extension of the index (IFPRI 2020). This paper uses information from the pro-WEAI module. The pro-WEAI was administered separately to the main female and male decision-makers of each household; in households without a male adult, only the female adult was interviewed. The quantitative survey was collected in March 2022.

The pro-WEAI focuses on agency (as per Kabeer’s (1999) framework) and uses 12 indicators to measure agency in three domains: intrinsic, instrumental and collective. Intrinsic agency is measured through the following indicators: autonomy in income, self-efficacy, attitudes about intimate partner violence against women, and respect among household members (the last indicator is calculated only for households with two respondents). The instrumental agency indicators include mainly decision-making questions: input in productive decisions, ownership of land and other assets, access to and decisions on financial services; control over use of income, work balance, and visiting important locations. Finally, collective agency is approximated with an indicator for group membership and an indicator for membership in influential groups.

The survey was administered to 712 households, but some were excluded because of incomplete information; the final sample consisted of 700 households. Of these 700 households, 343 were in the communities where qualitative data were collected. Given the focus on emic and context-based approaches, the analysis for this study only uses information from these 343 households so that the quantitative data could be more accurately mapped onto the qualitative results from the same communities.

Data analysis and limitations

During qualitative data collection, typed notes and recordings were produced, which were then translated into English and systematically reviewed by the field team leaders. The findings presented here are the result of content analysis of all data for each case. A hand-coding exercise was conducted to capture recurring themes and, following the framework laid out by Kabeer (1999), to understand linkages between agency, resources, and achievements that local people expressed to be empowering in their context.

The pro-WEAI is calculated as the weighted mean of two sub-indices: a sub-index that captures adequacy in the three agency domains (3DE), and a gender parity index which compares the empowerment scores of the male and female respondent in the same household. A comprehensive description of the index is available in Malapit et al. (2019).

Challenges were faced during the preparation, implementation and analysis stages of the study. Delays in the preparatory phase related in part to complications from the Covid-19 pandemic led to a reduced window of time to prepare and mobilize respondents and to transcribe the data collected. Issues inherent to translation in addition to imprecise and informal language resulting from the conversational format of KIIs, FGDs, and LHs leave many data points open to varying interpretations (e.g., the difference between being “valued” by the community and “respected” by the community, as explored in FGDs).

Results

Resources, agency, and achievement

The responses that arose from the FGDs and LHs provided rich depictions of local values and emic understandings of empowerment, their dynamic natures, and their nuanced differences and similarities between different social groups. In this section, we structure these findings around Kabeer’s (1999) definition of empowerment indicators based on resources, agency, and achievement.

Resources

Land, livestock, and money surfaced as key resources for women and men alike in FGDs and LHs. However, participants suggested access to these resources is gendered—though this is changing for younger generations. According to community profiles in both Kilifi and Kiambu, although women and men have equal rights to buy and inherit land under the law, most land is still customarily inherited and owned by men. As a result, women largely access land through their husbands or male relatives. In terms of livestock, FGD participants reported that women are increasingly recognized as co-owners of animals. However, this ‘ownership’ often does not translate to decision-making power over the asset or the income it generates, due to deep-seated gender norms that give men greater influence in intra-household decision-making. Additionally, women and men alike reported in FGDs that while most women and men have their own streams of income, women’s incomes tend to be lower and more allocated to necessities for the family, while men were able to spend on ‘pleasures.’

LHs showed a more nuanced view of men’s and women’s financial assets. Women respondents commonly reported that they receive financial support from their adult children or other male family members. In contrast, men largely reported receiving some contributions to their household finances from their wives but said that the bulk of their income is self-generated. There was an exception to this trend: one older Kilifi man said he is too old to work very much, so his wife is looking for work abroad.

Education also surfaced as one of the most highly valued resources for empowerment, but there were notable demographic differences in its valuation. Education was most valued by younger men, and least valued by older men. Younger women valued education nearly as much as men in their age group, but argued that it needs to be paired with other resources to be effective. As a young, empowered Kiambu woman said in her LH,

“You can equip me with information, but if you don’t equip me with inputs like, say, a hoe to use on my land, you have not given me [the necessary] resources. You see, equipping me on the practical side of things, that is where I am stuck.”

Agency

Women and men in FGDs and LHs reported that women have relatively limited decision-making power at the household level. An older Kiambu woman explained in FGDs that even in situations of ‘joint’ decision-making, the women’s input is subordinate to the man’s:

“She gives the husband proposals, saying, ‘if we do this or we do that we would benefit,’ and then her husband weighs her suggestion and then decides. You see, she is not making a decision, but she is proposing."

Women in this FGD also suggested that if a woman does not accept her husband’s decision, she may be hit.

Participants across qualitative groups suggested that being a landowner and/or a homeowner grants you decision-making power regarding those assets, but that such control is not guaranteed for women. Kiambu FGDs suggested that men have final decision-making power over income generated from selling agricultural produce, for example, even when it comes to vegetables that are perceived as women’s crops. According to some LHs, married women may be able to make decisions regarding co-owned livestock on their own, but they must consult or defer to their husbands when it comes to decisions on income generated by these assets. However, a man is not held to the same standard:

“The man will take her assets and sell them without question, but if the woman dares to sell his assets then she can even be thrown out of the household and sent packing to her parents.”

In LHs, older women who receive money from their male family members reported not having decision-making power over that money. Younger women’s answers were more variable according to context. Where one young woman was injured and therefore less physically able to participate in household decision making, another is a single accountant and retains complete decision-making power over her life. The remaining two struggle heavily with finances and make decisions with their adult child and older brother, respectively.

In FGDs, there was a notable difference in intrahousehold decision-making patterns as described by older men and younger men. Younger men reported higher levels of joint-decision-making in their households, and generally described this as neutral or positive. In contrast, older men tended to describe a more patriarchal household structure. Several of these older men expressed resentment over cultural changes in recent years that have given their wives more confidence to assert their opinions or give pushback to men’s authority.

Achievements

When asked to describe empowered men and women in their communities, FGD participants most often described someone who can support themselves, and therefore support their family. The emic relevance of self-sufficiency as an empowerment indicator was echoed in LHs, particularly by older women. As one older, disempowered Kiambu woman described:

“I had never kept chickens of my own in my life. Being able to eat my own chicken and even eggs makes me feel empowered.”

Notably, for older women in LHs, self-sufficiency and independence was not viewed as individualistic but as relational; these women tended to attribute their perceived positions on the Ladder of Power and Freedom (Sect. "Pathways to empowerment") to their relative ability to support themselves as well as their families.

The ability to lift one’s family out of poverty is seen as a key achievement and empowerment outcome. Owning a business and/or being self-employed was regarded as a desired empowerment outcome by and for women and men alike. Participants frequently suggested that factors of resources and agency which do not lead to this achievement—poverty alleviation—were useless for empowerment. For example, older Kiambu men were dismissive of education as a resource for empowerment, as they saw it as disconnected from achievements in their lived experience:

“Education is nothing. [...] We have people with education, but their families are living in poverty. [What matters is] is hard work, effort, and knowledge. But education is not important.”

Financial empowerment was valued by women and men alike but conceptualized differently by and for women and men. Financial empowerment traits were more often ascribed to men, but were more frequently mentioned by women. Furthermore, men tended to emphasize financial stability while women focused on individual resources such as land and livestock, which they believe will grant them greater self-sufficiency.

In addition to economic achievements, some of the most prevalent descriptors of empowered people were focused on the person’s role as a model member of the community. When describing empowerment, women and men alike valued being of value to and valued by their communities.

Personal and communal empowerment

In addition to a heavy focus on communal aspects of empowerment, participants also emphasized personal character traits such as being hardworking, dependable, mentally strong, having a positive outlook, and (in the case of men describing empowered women) even being physically attractive. Women, especially, were described with these character-based empowerment traits more often than men. “Having self esteem/believing in self” and “having goals/being ambitious” were especially heavily skewed toward women. Men also placed great emphasis on such traits, although to a lesser degree. As one older Kilifi man explained in his LH:

“It is not just education, it is your determination, wisdom, hard work that can make you be somebody in life. You can have your education but not do anything. You can have your money but not do anything because of lack of knowledge and plan on how to spend it.”

While participants expressed that character traits like determination, wisdom, and hard work are essential to empowerment, they also felt that these traits could be misapplied. For example, a man with determination and self-confidence was described as empowered, while a wife with the same characteristics was disrespectful. In this way, community acceptance or rejection of one’s personal empowerment and empowerment-related character traits can make or break one’s empowerment on a community level. Across FGDs, participants suggest that community acceptance is key to not being disempowered, and having influence and being able to better the community is essential to being empowered.

Pathways to empowerment

When designing their Ladders of Power and Freedom, all focus groups described a process of economic empowerment. In their descriptions, the lowest rung on the ladder represents someone who is impoverished and ‘hustling’, and the top rung of the latter represents someone who has accrued wealth, whether it be through agriculture or a business (such as hospitality in peri-urban Kiambu and sand-harvesting in coastal Kilifi). For many groups, the most empowered step included owning their own business (and, in some cases, their own land). Many groups’ ladders also emphasized the importance of relational and communal elements. People at less empowered steps were described as a drain or burden to their families and communities, whereas those at more empowered steps were described as respected and supportive of their communities. As an illustrative example, the ladder below provides a summary of the ladder steps described in the Kilifi Women’s FGD (Fig. 2):

Fig. 2

Results of the Ladder of Power and Freedom activity in Kilifi older women’s FGD

Most of the ‘mobility factors’ that emerged from this activity were related to resources, including: access to capacity-building seminars and trainings; access to agricultural extension; table banking (a communal savings scheme); the ability to plant a greater diversity of crops and higher value crops; access to credit; access to land; education (and separately, knowledge/wisdom, which many saw as being separate and even antithetical to education); access to technologies and machinery such as a tractor; job opportunities; government support; and having savings. A lesser number of factors were agency-related, such as having voice and autonomy. Others were of a personal and/or communal nature, such as being reliable/being seen as reliable, having a good upbringing, and a strong family relationship.

There were some notable differences between empowerment pathways by different age groups. Three of the most-mentioned mobility factors—good upbringing/good family, has voice/autonomy, and education—were only mentioned in FGDs of younger men and women, with no mentions from groups composed of older men and women. While most older groups did not mention education as a mobility factor at all, one group of older men in Kilifi actively argued against it as a reliable pathway to empowerment.

FGDs also revealed differences in mobility pathways across gender groups. Participants in three out of four men’s focus groups expressed that one must lift himself out of poverty through self-motivation, hard work, and being less “lazy.” (Young men in Kiambu were an exception: as a group they argued that it is the government’s responsibility to empower the youth in their community through lower taxes, government subsidies, and building local capacity for business and smart farming.) Participants in three out of four women’s groups, on the other hand, took a more collective view of empowerment pathways, mentioning mutual support strategies such as table banking, women’s groups, and peer-to-peer mentorships.

Changes and variations in empowerment pathways

FGD participants expressed that even if a woman has all the same indicators of empowerment and mobility factors as a man, her ultimate level of empowerment will nonetheless be limited due to underlying power structures which serve to benefit men. In other words, the maximum level of empowerment for women is lower than for men. LHs corroborated this finding, even though the sampling was conducted according to self-identified levels of empowerment and included empowered and disempowered people of all genders and ages. Men positioned themselves higher on the 5-step ladder than women, and older age groups positioned themselves higher than younger, with older men positioning themselves the highest of all four study groups, and younger women the lowest (Fig. 3).

Fig. 3

Positions on ladder as self-identified by LH respondents, where 5 is most empowered (the ‘top’ of the ladder) and 1 represents disempowerment

FGDs and LHs showed that empowerment is context-specific – shifting and being negotiated depending on the social space that an ‘empowered’ person is navigating. As one Kiambu older man explained in FGDs, women’s empowerment is limited within the familial/household hierarchy, regardless of her empowerment in other (e.g., professional) domains:

“I am the one feeding her, and she is under my authority, so I am the one controlling her when she comes to my house. Even if she is a president at her workplace, when she enters my compound, she is a wife who has a husband, and she is following the rules that we wrote. When she leaves the gate [of the family compound], that is when she can resume her seniority. But when she enters the homestead, she is under my authority. Her husband is what she follows. Even if she has more salary, that is not a problem to me.”

LHs also show the ways that empowerment can wax and wane over the course of one's life. An older unempowered Kiambu woman who placed herself on the second ladder rung (one step away from total disempowerment) explained:

“Although I have been able to buy myself land and build my house, that was before. Right now, if I was educating my children I wouldn’t manage, because at the moment I find that I am straining even to feed myself.”

Some men, especially older men, viewed that their overall empowerment has decreased due to women’s increasing empowerment in recent years. Even when empowerment was not viewed as a zero-sum game, some men still reported feeling insufficiently targeted and supported in terms of their own empowerment:

“There is just improvement for women, and matters of the youth are being looked at, but the man has been neglected.” (FGD, Kiambu, older man)

“I can say that women are favored because, when we look at the government projects, they just target women. There are no projects for men. In youth employment initiatives, for instance, you find more girls get the jobs than boys.” (LH, Kiambu, older man)

Attitudes toward empowered women and men

FGD participants in both Kiambu and Kilifi expressed that their communities have ambivalent attitudes toward empowered women, describing them as openly admired and treated as a threat/feared in equal turn. By contrast, the most common perception of empowered men is that they’re desired, while empowered women received zero mentions of being desired. Younger women were slightly more likely to be viewed as a threat than older women. Of the different responses of how powerful women are perceived, a slight majority was negative, whereas the majority of different responses for the overall perception of empowered men were positive.

These responses also differed by the gender of the participant; empowered women were much more likely to be admired by other women than by men. On the other hand, they were also likely to be regarded with jealousy by other women. A young woman in Kiambu reflected on the conflicted local attitudes toward empowered younger women, saying:

“This [empowered] young woman is perceived as a social deviant, but she’s secretly admired [for the fact] that she’s broken out of the norm, which is to stay at home, have many babies and languish in poverty.”

Older men in Kiambu strongly felt that women’s empowerment is a threat to the social fabric. Some saw gender equality as a dangerous intervention from outside influences that poses an inappropriate affront to their culture, as illustrated by the following FGD quotes:

“Women should stay the way God created them. It is not that they will be oppressed, [but] there is always that order of levels. And now they want to lawfully take from us the authority that God gave us. Please, they should move slowly to prevent the world from falling apart, because I can see they will make the world fall apart.”

“A woman is a woman, and she is supposed to remain a woman. I don’t see reason in what people say that ‘what men can do, [a woman can do better].’ Those [ideas] are for white people, and they are deceiving us.”

While some participants saw women’s increased autonomy and decreased reliance on men as a threat to the social fabric, others saw this as a beneficial outcome which allows women to have more autonomy and time to dedicate to more valuable pursuits than taking care of a husband. Younger men were overall more supportive of women’s empowerment than older men, often expressing that empowering women can have community-level benefits.

Negative feelings toward empowerment were not limited to empowered women. Empowered men were also described with suspicion or mixed feelings, albeit to a lesser degree. Some FGDs expressed that empowered men had forgotten about their communities or had left them behind. Young men in Kilifi expressed that women use their empowerment differently than men, in a way that is more uplifting for the community:

“In our area, if a girl is empowered well, she remembers home. But most of the time, the male youth, if he is empowered well… if he goes, he is gone. He does not go back home.”

For women and men alike, empowerment traits do not always translate to being admired or accepted in the local community. In FGDs, descriptions of “empowered” people and descriptions of “admired” people in their communities often diverged significantly. For example, having financial stability and/or a stable job was relevant for an empowered woman, but not for an admired woman, whereas having and/or valuing education was relevant for empowered men but not admired men. Caring for the family was severely skewed toward admiration for women (and fairly evenly spread for men). Being well behaved/disciplined was also highly skewed toward admired women, and no mentions of behavior/discipline were applied to older men at all (whereas it was mentioned in relation to younger men and women of all ages).

The way that empowerment is “used” was described as key to whether empowered men and women are also admired and/or accepted. Older women in a Kilifi FGD expressed that they see two kinds of empowered women in their community: those who use their empowerment to help others and uplift the community as a whole, and those who are prideful and aloof. Young men in Kiambu also expressed that young men’s attitudes toward an empowered young woman depend on her behavior and her obedience. One older Kiambu man expressed in his LH:

“This issue of empowerment will break homes. But in my opinion, just empower women so that we can know those who are good and those who are bad in character. You can perceive someone to be a good person but it’s because she doesn’t have money -- that’s why she behaves differently. When she gets money, you will know her real character.”

Quantitative measures of empowerment from the pro-WEAI

The Project-level Women’s Empowerment in Agriculture Index (pro-WEAI) used externally defined and standardized indicators to quantitatively evaluate women’s and men’s levels of empowerment in the study communities. The results from the pro-WEAI are included in Tables 2 and 3. The pro-WEAI score in the surveyed communities is 0.89. The 3DE score, which is the average score over the three dimensions of agency (intrinsic, instrumental, and collective), is 0.88 for women and 0.86 for men suggesting comparable levels of empowerment between men and women. At the time of the survey, 64% of women in the surveyed communities and 63% of men were classified as empowered. Moreover, 75% of dual-headed households had achieved gender parity based on the gender parity sub-index.

Table 2 Pro-WEAI score

Table 3 Headcount ratios and relative contributions of each indicator to disempowerment

Table 3 shows the share of disempowered men and women according to each indicator (regardless of whether they are classified as disempowered based on the entire index) and the relative contribution of each indicator to total disempowerment. A gender gap in intrinsic agency in favor of men is visible. The gap is mainly driven by the indicator on attitudes about intimate partner violence. Over 30% of women are disempowered in this indicator compared with 18% of men. Based on the pro-WEAI, a similar share of men and women lack intrinsic agency in terms of autonomy over income and self-efficacy.

In addition to having lower overall intrinsic agency than men, female heads of household are also more disempowered than men in terms of instrumental agency according to pro-WEAI measures. The instrumental agency domain includes several indicators focused on decision-making (over productive activities, credit and financial assets, and income). In contrast, a smaller share of women than men are disempowered in terms of decisions over productive activities. This could be linked to women’s greater role in agricultural production relative to men, who are more likely than women to have off-farm wage jobs. Only about 8% of women and 10% of men report that they lack control over the use of income.

Gender differences in work balance contribute most notably to the gender gap in instrumental agency. More than half of women report inadequate levels of work balance compared with 36% of men.

Male household heads are more disempowered than their female counterparts in terms of collective agency, mainly because men are less likely to participate in groups than women. Only eight percent of women compared with 28% of men are disempowered in group membership; similarly, 22% of women compared with 37% of men are disempowered in terms of membership in influential groups.

Discussion

Intersections of empowerment and identity

Kabeer (1999) argues that resources, agency, and achievement are intimately linked, and the necessary triangulation of these factors for empowerment was reflected by participants in FGDs and LHs. Resources (most notably money, land, livestock, and education) were seen by women and men of both age groups as essential for empowerment, but useless without agency to control those resources or channel them into achievements (most notably poverty alleviation and the ability to support one’s family). Those abilities were shown to be highly gendered and mediated by life stage and generation.

While women’s access to money and ownership of land and livestock is increasing in younger generations, for example, norms and customs result in men’s greater access to and control over these key resources. This finding is in line with the broader literature on women’s land rights, which demonstrates that, at a global level, “despite women’s complex relationship with productive resources, they tend to remain outside the associated decision-making processes, for cultural or other reasons” (Budlender and Alma 2011, p. 11). Participants suggested that being a landowner or a homeowner grants you decision-making power regarding those assets, however these assets are almost entirely owned and controlled by men. As such, women’s agency is more limited than men’s, and their ability to achieve key aims such as self-sufficiency and supporting their families is also limited. However, such opportunities and abilities can change over time according to women’s life stage, showing that while gender is a major condition for resources, agency, and achievements, gender groups are ultimately heterogeneous, and individuals’ experiences are often divergent. These findings corroborate recent trends in development discourse that emphasize the importance of intersectional nuance for more inclusive and effective development initiatives (Corrêa 2010; Gill and Pires 2019; Leder and Sachs 2019; Trivelli and Morel 2021).

Conceptualizations of empowerment also proved to be highly gendered. More of the attributes ascribed to men than to women concern wealth and assets, whereas the top traits for women were largely personal character traits. Furthermore, the top attribute for women which does concern wealth and assets (makes money/contributes) as discussed in FGDs had more to do with contributing money to household needs than with a woman’s ownership/control of assets for her own decision-making and personal empowerment. However, while economic descriptors skewed toward men, these comments were mostly made by women, suggesting that this characteristic is more valued overall by women than men as an indicator of empowerment for the men in their community. This finding is aligned with social norms that prescribe men’s role as the financial provider of the household despite shifting roles and practices (Johnson 2004; Wamue-Ngare and Njoroge 2011). In addition, when men and women talked about economic factors of empowerment, men tended to emphasize wealth more generally while women focused on specific resources. This may be because men already own these assets and are striving for higher-level achievements of wealth, whereas women are still striving to gain access to and control over basic resources such as land (Po and Hickey 2018) and livestock (Njuki et al. 2013).

LHs echoed the finding that men’s empowerment is more closely aligned with economic empowerment pathways, and showed that young women place a high value on non-economic resources. Several young women pointed to information, community groups, and training as integral to their pathways; in contrast, three of four younger men listed having work and funds in their pathways to having more influence in their communities. LHs show that decision-making dynamics around money in female-headed households are influenced by a greater variety of factors, while male headed households reported clearer trends. In other words, a number of factors must align for women to exercise agency, whereas for men the barriers are fewer and more easily identified. In part, this is because the presence (or absence) of men in a woman’s life can greatly change her access to money and decision-making over income and assets. Women’s LHs showed that even in female-headed households, men often have higher agency and decision-making power over financial resources. On the other hand, a man’s control over money is less likely to be impacted by the presence (or lack thereof) of a woman.

Age proved to be an important factor in empowerment, but in divergent ways. Data indicated that ‘age’ can be broken down into two different influential factors that deserve a closer look: generation and life stage. Attitudes and perceptions about empowerment factors, empowered people, and gender equality varied significantly between generations, while people’s own lived experience of empowerment was highly mediated by their current life stage. For example, three of the most-mentioned mobility factors—good upbringing/good family, voice/autonomy, and education—were only mentioned in FGDs of younger men and women. Interestingly, none of these are economic mobility factors, possibly suggesting that the younger generation’s perception of empowerment pathways goes beyond economic factors. What’s more, younger participants (both women and men) were generally more accepting of women’s empowerment and more hopeful about changing gender roles. In contrast, older generations tended to view women’s empowerment as destabilizing and/or as limited in its extent due to patriarchal structures which were described as natural and/or functionally permanent. Within each generation, however, there was significant variation in individuals’ empowerment pathways related to life stage (e.g., being single, being married with young children, being married with older children, being married/unmarried with grown children, and being widowed). Women’s ability to access and exploit resources, agency, and achievement was largely mediated by their relationships with men in their current stage of life. In this way, empowerment is a continual and dynamic process rather than a static achievement.

Norms, negotiation, and the politics of empowerment

Relational empowerment, respectability, and backlash

One of the standout findings from the qualitative data was that participants placed a high value on personal character traits as essential to empowered people and as important factors in empowerment pathways. These traits were not important on their own or on an individual level, however, but in how they were used in a communal context. The finding that both highly personal and highly communal aspects of empowerment are deeply interconnected, and that together they are essential to Kenyan emic definitions, agrees with a 2019 study byGaliè and Farnworth. They found that Kenyan men and women dairy farmer participants “argued that an individual's empowerment is mediated through intrinsic (inborn) characteristics, with particular importance given to how these characteristics promote the ability to relate to others” (p. 15). This study’s findings also correspond with the 2019 study’s findings that Kenyans feel that empowerment characteristics can be ‘misused’ by certain individuals. For example, a man with determination and self-confidence was described as empowered, while a wife with the same characteristics was disrespectful, and therefore “not acknowledged as empowered by the rest of the community.” (p. 15). Galiè and Farnworth conceptualize this form of (dis)empowerment through communal perception and acceptance as “relational” empowerment. In this study, participants in both counties placed a high value on relational empowerment, both as a potential process of disempowerment and as an essential component of effective emic empowerment.

Furthermore, this study echoes Galiè and Farnworth’s (2019) findings that the politics of relational empowerment is deeply gendered. In cases where the same empowerment characteristics are praised in men and discouraged in women, they argue that “an involuntary form of empowerment emerges: a complex interplay between a woman holding innate characteristics of empowerment and the compatibility of these characteristics with locally sanctioned gender roles.” In this way, women’s need to ‘negotiate’ their own relational empowerment may be higher or more difficult than that of their male counterparts. Furthermore, women’s empowerment was more often associated with these ‘personal’ or ‘inborn’ empowerment traits which are monitored and adjudicated by the community, while men’s empowerment is more often associated with more straightforward (and thereby more difficult to negate) economic indicators. Indeed, our findings show that while relational aspects of empowerment were key for all genders and ages in Kenya, men seemed to be primarily concerned with relational empowerment in the positive sense (a pathway to becoming more empowered), while for women it was more often seen in a negative sense (a barrier to becoming empowered or a means of being disempowered).

Findings from FGDs and LHs showed that women’s empowerment is viewed as acceptable insofar as it does not upset traditional societal roles and values. This effectively places a ceiling on women’s empowerment. As FGDs indicated, even if an empowered man and woman have exactly the same characteristics of empowerment, they are not seen as equals due to the deep-seated power structures wherein the man leads and the woman supports. Women who are trying to challenge these roles are looked down upon and seen as unhelpful or even destructive to the community, and may, as Kabeer (1999) describes, “pay a high price for their autonomy” (p. 457). However, norms are constantly being challenged, adapted, and negotiated as evidenced by the younger generation displaying more hope and less fear than their older community members with regard to gender equality.

Externally defined empowerment indicators fail to fully capture ground realities

Comparing the quantitative results from the pro-WEAI (which represent externally defined empowerment indicators) with the qualitative data (which sought to provide an emic understanding of empowerment in the given context) yielded contradictions in terms of overall empowerment levels of women and men. Pro-WEAI findings that women and men in the study are empowered at a similar level, and that 75% of households have gender parity, are somewhat challenged by insights from the qualitative data.

Quantitative and qualitative results agreed in terms of some individual empowerment indicators and disagreed for others. For intrinsic agency, findings were in agreement, with both quantitative and qualitative results surfacing the fact that intimate partner violence poses a threat to women’s empowerment. There was less agreement in terms of instrumental agency. The pro-WEAI suggested that women are more empowered than men in terms of decision-making over productive activities, but the qualitative findings paint a more nuanced picture. While women may engage in production decisions, they may not be in control of the income derived from this production, a finding that comes out clearly in the qualitative data but not in the quantitative data. Unlike the qualitative data, the survey questions do not ask respondents to state who in the household has the final say in instances of disagreement between spouses. For collective agency, the pro-WEAI found that men have lower overall levels than women. This was also reflected in the qualitative data; women’s groups tended to view collective organizing such as table banking, women’s groups, and peer-to-peer mentoring as integral to empowerment pathways, whereas men focused on more individual approaches such as “hard work” and being “less lazy.” The importance of collective agency was further underscored by the high value placed on “power through”, and being valued by and of value to the community.

Furthermore, the qualitative data revealed the importance of relational empowerment in the Kenyan culture context, which is not accounted for in the pro-WEAI. If these locally important factors were included in the instrument, the findings likely would have been quite different, and more reflective of women’s and men’s ground realities in Kenya. The pro-WEAI does include a qualitative component to surface exactly these kinds of nuances and contradictions and is one of the most adaptable and context-conscious tools of its kind. However, the qualitative component serves as a complement to the externally defined index rather than a core component.

Men’s involvement

In FGDs and LHs, women and men alike reported that women and youth are more often targeted for trainings and empowerment schemes than men. As a result, the latter perceived that their own levels of empowerment have decreased, exacerbating feelings of resentment and/or distrust toward women’s empowerment (which are deeply rooted in cultural and religious ideology, making them particularly sticky). It can also disincentivize empowerment for women, who have much to lose in terms of desirability, respectability, and acceptance in the eyes of their husbands, families, and communities. These findings are in line with a previous study about shifting gender norms in Kiambu county (Wamue-Ngare and Njoroge 2011), which found that changing gender roles and women’s increased empowerment at the local level have resulted in a scenario in which men feel “unable to live up to the unwritten cultural norms and values of behaviour.” As such, “most men in Kiambu have succumbed to feelings of inferiority, uncertainty and frustrations, all with dire consequences on the family” (p. 11). Men’s feelings of neglect also signal that there may be men struggling with dimensions of disempowerment (e.g., poverty, rurality, disability) who have been insufficiently targeted by empowerment initiatives and may need improved supports for their own empowerment. Thus, a unilateral focus on women’s empowerment may be insufficient as an approach for gender equality as well as empowerment for all marginalized community members.

Conclusion

Viable empowerment pathways for men and women remain gendered in Kenya, with notable differences for different age groups. These differences often reflect or serve to reinforce patriarchal norms (e.g., economic empowerment is predominantly associated with men, while relational empowerment is more heavily associated with women). Furthermore, empowered men were mostly perceived positively by their communities, whereas attitudes toward empowered women were ambivalent. This acceptance or lack thereof has a considerable impact on empowerment pathways in Kenya, where relational empowerment is as important of a factor in empowerment as resources, agency, and achievements. The results of this community judgment can also be used to reflect or perpetuate patriarchal norms.

Negative attitudes toward women’s empowerment are related to perceived threats toward underlying power structures which shape these societies, as well as with men’s feelings of neglect and the idea that if women are empowered men will be disempowered. These feelings indicate that risk of backlash may be high for some women, and that men’s engagement in empowerment initiatives is paramount to their own empowerment and wellbeing, as well as that of the women in their communities.

An intersectional approach that takes into account the diverse needs of men and women of different social groups is paramount to the safe, effective, and lasting empowerment of marginalized community members. We contend that this should include a more nuanced understanding of age which takes into account the importance of and differences between ‘life stage’ and ‘generation’, the distinction between which is usually lost under the simplified label of ‘age.’

The diversity of emic understandings of empowerment surfaced in this study and the ways that they impact and are impacted by intersectional identities underscores the need for nuanced approaches to empowerment. These approaches must center local communities as the protagonists of their own development, and empower them to define their own goals and pathways. In this way, communities may be more incentivized to pursue empowerment that is relevant and valuable to them, and feel safe in doing so. Efforts to support women’s empowerment, in particular, need to be conducted with the larger social context in mind in order to mitigate risk of backlash and to create an enabling environment that encourages women to pursue their own empowerment without fear of disproportionate tradeoffs such as social alienation. This could include employing gender-transformative approaches that target changing deep-rooted social norms in addition to empowerment outcomes, as well as forging greater engagement with men and especially including disempowered men in empowerment interventions.