Determinants of Women Empowerment: Case of Refugee Women Living in Nairobi Kenya

Abstract

This study investigates the determinants of women empowerment among refugee women living in Nairobi, Kenya. First, the study constructs an index to examine empowerment drivers using data from the Refugee and Host Household Survey (RHHS) 2021. A fractional logit regression model was employed in the study. The results obtained show that the incidence of refugee women empowerment among refugees living in Nairobi was six percent. In addition, the study finds evidence that age; the gender of the household head; the education level of the refugee woman; employment status; and the education of the household head play substantial roles in enabling women empowerment. Conversely, marital statuses (divorced/separated/widowed and single/never married) and religious affiliations (Muslim) hinder women empowerment. Efforts geared towards improving wage employment and education are likely to empower refugee women. The study emphasizes the recognition of the role played by women in household income through care work.

1. Introduction

More than 108 million people were forcefully displaced globally due to persecution, conflict, and other human rights violations as of 2022 (United Nations High Commissioner for Refugees (UNHCR) Global Trends Report (UNHCR, 2022)). Such humanitarian crises disproportionately impact the well-being of women, who comprise 51% of the refugees (Han, 2024). Women refugees, unlike their male counterparts, face specific hurdles during migration and resettlement. Some of these challenges include vulnerability to sexual violence and inequalities in access to humanitarian assistance (Jabbar & Zaza, 2016). There is increased exposure to HIV (Human Immunodeficiency Virus) infections among forcibly displaced persons due to the interplay of sexual and gender-based violence, poverty, and limited access to sexual and reproductive health (Logie et al., 2022).

These vulnerabilities exacerbate underlying gender gaps that already interfere with efforts to achieve gender equality and women empowerment. For instance, nearly two-thirds of women are illiterate (United Nations Educational, Scientific and Cultural Organization (UNESCO, 2024)). Further, women in both developed and developing countries are paid 10–30% less than men for the same amount of work (ILO, 2018). Gender equality and women empowerment are essential pillars of sustainable development. Particularly, gender equality is enshrined in the United Nations (UN) Sustainable Development Goals (SDGs) highlighted in SDG 5. The goal seeks to have inclusivity of women and girls for sustained development. According to Leach et al. (2016), gender equality is integral in steering sustainable development as it cross-cuts all aspects of development. Therefore, achieving sustainable development requires addressing the unique challenges faced by marginalized groups, including refugee women, who often encounter compounded social, economic, and environmental vulnerabilities during displacement.

Refugee women play an important role in building resilience during restoration and resettlement in many ways—for instance, in their vital role as caregivers (Khatib et al., 2024). Since forced migration alters household head dynamics when refugee women assume breadwinner roles (Jabbar & Zaza, 2016), studying the interaction of household characteristics with women empowerment is important.

While many studies have measured women empowerment and assessed its determinants, Upadhyay and Karasek (2012) suggest that measurement is context-specific. For example, women empowerment may have different meanings in parts of Africa, where polygamy, the social importance of fertility, and the ideal family size may differ from those of the Asian context that was used in the conceptualization and development of women empowerment measures. Thus, it is relevant to not only examine women empowerment in Sub-Saharan Africa, but also in the societal context of refugees who are exposed to risks such as the loss of documents, insufficient resources, psychosocial health stressors, and gender-based violence.1 Additionally, refugees often face challenges with integrating into formal markets (e.g., accreditation and work) in host countries, regardless of their education level and skill sets (Vuni & Buhendwa, 2023). For instance, a report by Vintar et al. (2022) showed that before COVID-19, only 20 percent of refugees were in employment, including self-employment, compared to 63 percent of the nationals despite the enactment of the Refugees Act of 2021, granting refugees the right to work and freedom of movement.

Refugee women, already vulnerable in the context of forced migration, face disproportionately heightened risks due to intersecting factors such as systemic inequalities in gender roles, social hierarchies, and power relations, which influence their level of agency and socioeconomic status (Trentin et al., 2023). In this study, empowerment is considered in terms of agency over personal health, participation in labor, access to and control of contraception, involvement in domestic household decision-making, access to information, knowledge of HIV prevention and transmission, and attitudes towards gender-based violence and early marriages (Kabeer, 1999; KNBS, 2020). Although the ability to feed the family may empower women, it may also precipitate conflict and violence in private family life. In some cases, intimate partner violence manifests from frustrations related to securing employment, leading to low self-esteem, a loss of confidence, and boredom (Yalim & Critelli, 2023).

The literature presents mixed findings on factors that influence women empowerment (Acharya et al., 2020; Musonera & Heshmati, 2017; Sado et al., 2014; Upadhyay & Karasek, 2012; Sathar & Kazi, 2000). Among other sources, mixed findings have been attributed to approaches such as estimating regressions separately without constructing a women empowerment index (Acharya et al., 2020; Mahmud et al., 2012). Factors such as land ownership, which is used in studying rural women empowerment (Kuma & Godana, 2023), may also not be appropriate for refugee women who have neither formal documents nor assets.

Kenya offers a compelling context in which to study empowerment among women refugees. It is the fifth-largest asylum country in Africa, hosting about 550,817 refugees as of April 2022, with close to half (43%) residing in the Dadaab refugee camp, 41% in the Kakuma refugee camp, and 16% in urban areas (mainly Nairobi)2. Somali refugees account for 53% of the total refugee population in Kenya, followed by refugees from South Sudan, DRC, and Ethiopia at 25%, 10%, and 5.6%, respectively.

Refugees predominantly live in Eastleigh, Nairobi. Due to financial pressure to survive in this protracted refugee business hub, women must seek paid employment (mainly subsistence activities such as small-scale businesses) and become de facto household heads, owing to their husbands’ death or unemployment (Ritchie, 2018). Restrictive regulatory policies and bureaucratic processes such as obtaining business permits and licenses make it challenging for refugee women to operate within the city, exposing them to exploitation, especially by police and city county officials (Jaji, 2009).

With such hurdles in operating businesses and limited labor market pathways for highly skilled refugee women and men to have their credentials certified and validated, most refugees resort to informal entrepreneurship ventures for economic well-being and integration (Khatib et al., 2024). In addition, while some of the refugees may be highly skilled and educated, Ritchie (2018, 2022) studied Somali refugee women in urban Kenya and found that most were illiterate or semi-illiterate and tended to be single or divorced with children.

Therefore, this study explores how individual characteristics and other determinants affect empowerment. Specifically, the study sought to answer the following questions; first, what is the incidence of empowerment among refugee women in Nairobi, Kenya? Second, how do the characteristics of refugee women compare to those of nationals? Finally, what factors significantly determine the empowerment of refugee women in Nairobi, Kenya? Factors such as education, age, and marital status may vary in their explanatory powers. According to Sado et al. (2014), factors associated with sources of empowerment (education, employment, and wealth status) have higher explanatory power relative to those related to the setting of empowerment (family structure and age). Since differences in cultural contexts affect the measurement of women empowerment, per Upadhyay & Karasek (2012), it is important to construct an empowerment index embedded in the context of urban refugees and examine context-specific influencing factors.

This study uses the Refugee and Host Household Survey (RHHS) 2021 data that were collected in Nairobi to study factors that help to explain refugee women empowerment. The study finds evidence that the age of the household head, gender of the household head, education level of the refugee woman, employment status of the household head, and education of the household head play significant roles in enabling women empowerment. However, being single, widowed, separated, or divorced, as well as being Muslim, lowers empowerment.

The rest of the paper is organized as follows. Section 2 presents the literature review, while Section 3 discusses the methodology and data. Section 4 provides the analytical results, and Section 5 consists of the study’s conclusions and policy recommendations.

2. Literature Review

2.1. Theoretical Review

Similar to Doepke and Tertilt (2019), this study draws from the theory of household decision-making and the division of labor to explain the development implications of gender-specific aspects such as empowerment since forced displacement often deconstructs gender roles in the household. The theory is key in showing the grounding structures of households and the behaviors of the individuals that form the household. Prior to displacement, the household dynamics place men as the breadwinners and women as caregivers. However, women may become breadwinners rather than depending on their spouses for their livelihood, as they did in their original communities (Jabbar & Zaza, 2016). Therefore, refugee men and women may have to renegotiate their roles in their new settings, leading them to assume different roles and responsibilities inside and outside of their family norms and redistribute power (Benson, 1994; Yalim & Critelli, 2023). This creates opportunity for empowerment as women gain financial independence and can assume overall decision making in the household.

Further, the study also relates to the widely cited women empowerment framework by Kabeer (1999), which segregates empowerment into the dimensions of resources (preconditions), agency (process), and achievement (outcomes). Resources represent the material, human, and social assets that serve as the preconditions for empowerment, agency refers to the processes by which individuals exercise choice, voice, and control in their lives, encompassing both observable actions and internal motivations such as aspirations and self-efficacy, while achievements reflect the tangible outcomes of empowerment, such as increased income, improved health, or enhanced decision-making capacity within households (Kabeer, 1999). These dimensions are built from indicators constructed using data from women empowerment questionnaires. For instance, an economic indicator would be developed from questions/indicators such as “if a woman is in paid employment or not”.

2.2. Empirical Review

Previously, there has been a relatively narrow approach to measuring women empowerment in humanitarian settings, such as through biomedical lenses, e.g., sexual and reproductive health and rights of all women and girls. Therefore, other aspects of refugee women empowerment have often not been included in measurement efforts (Goulart et al., 2021). Studies focused on refugee women empowerment are limited. In the United Kingdom, Siddiquee and Kagan (2006) found that, among refugee women, increased competency in internet communication and technology skills promoted access to knowledge and resources and enabled the development of greater autonomy and a better living experience.

In another refugee study, some religious practices (e.g., hijabi dress) represented a hindrance in the implementation of a UNHCR sports project for Somali refugee girls. Some cultural and religious practices have been associated with lower literacy levels, cultural fatalism, and limitations in information and female participation (UNHCR, 2008). Sado et al. (2014) concluded that factors associated with sources of empowerment (education, employment, and wealth status) have higher explanatory power relative to factors related to the setting of empowerment (family structure and age). With limited studies specifically focused on determinants of urban refugee women empowerment, this paper links to the broader literature on women empowerment. In Nepal, Acharya et al. (2020) suggested that social demographic factors such as age, employment, and number of living children positively increased women’s autonomy in making decisions on their healthcare, major household purchases, visits to relatives, and purchases of daily household needs. Women’s schooling had varied effects on the four outcome variables. There was a significant relationship between decision-making on one’s own health for women with primary, some secondary, and higher education qualifications; the effect was more significant for the women with the highest education attainment. However, only primary and some secondary school attainment had a significant positive effect on decision-making regarding household purchases and visits to relatives.

Overall education attainment has been associated with the use of child and maternal health services and better health outcomes (Jejeebhoy, 2000). Upadhyay and Karasek (2012) find a positive influence of education, lending support to Kishor and Subaiya (2008), who found that education is positively associated with both joint and individual decision-making. Two Pakistan-based studies reveal varied results on the effect of the husband’s education on empowerment; Sathar and Kazi (2000) found it detrimental to female empowerment, while Khan and Noreen (2012) found no relationship.

Family sizes have mixed effects on women empowerment. Most studies associate empowerment with lower fertility (smaller family sizes) (Upadhyay & Karasek, 2012; Moursund & Kravdal, 2003). In Ethiopia, more educated women were more likely to communicate with their spouses about family size (Hogan et al., 1999). However, greater household decision-making and a woman’s right to refuse sex were associated with larger family sizes in a Namibia-based study (Upadhyay & Karasek, 2012); these paradoxical findings may mean that empowered women were still fulfilling societal expectations concerning fertility.

Acharya et al. (2020) noted that Nepalese women with more living children were more likely to take part in decision-making, while Musonera and Heshmati (2017) similarly identified that Rwandan women with five or more children—relative to those with one or two—were more empowered in household decision-making on healthcare, large household purchases, agency over the husband’s earnings, and visits to family or relatives. Marriage and childbearing are highly regarded in many societies, earning women respect, rights, and freedoms. Marital status was also found to influence rural women empowerment in Ethiopia. These findings are consistent with societal contexts such as the customary tenure system in Sub-Saharan Africa in which marriage has been an avenue for land access (Kuma & Godana, 2023). However, resource access may not always yield decision-making roles (Mahmud et al., 2012).

Compared to women in extended families, women in nuclear families tended to have greater autonomy and more responsibilities (Desai & Banerji, 2008). Autonomy in household decision-making was accrued as women grew older for extended families in Nepal, Bangladesh, and India. In contrast, newly married women in extended families tend to perform household duties under the supervision of their mother-in-law, who is the primary decision-maker (Acharya et al., 2020). Conversely, in Albania, being in an extended family was associated with increases in women’s decision-making autonomy and self-esteem (Sado et al., 2014).

Variations in the influence of age have also emerged. In Ethiopia, being older decreased economic empowerment in rural areas. Focus group discussions suggested that economic activities in rural areas were more labor-intensive, making them more challenging for older women (Kuma & Godana, 2023). Including age squared in the regression showed that empowerment portrayed a curvilinear relationship with age. The estimated maximum occurred at 26 years, implying that decision-making was highest for women in their mid-twenties compared to younger and older women (Mahmud et al., 2012).

Household wealth had alternative effects on different empowerment indicators. For instance, there was no association between household wealth and a woman’s mobility. Furthermore, women in wealthier households had less say in household decision-making but may have access to cash for expenditure (Mahmud et al., 2012). Therefore, resource access may not necessarily empower women in decision-making.

Paradoxical findings may be attributed, in part, to methodological approaches. For instance, Mahmud et al. (2012) and Acharya et al. (2020) estimate regressions for each empowerment dimension, rather than constructing an empowerment index. In addition to creating the index, the present study incorporates a broader array of information, compared to Musonera and Heshmati (2017) and Upadhyay and Karasek (2012), using 15 indicators grouped into five dimensions. Further, since differences in cultural contexts affect women empowerment (Upadhyay & Karasek, 2012), it is important to construct the empowerment index in the context of refugees and to examine its determinants.

3. Methodology and Data

3.1. Data

The study used cross-sectional data from the Refugee and Host Household Survey (RHHS) 2021, collected in Nairobi, Kenya. Data were collected using computer-assisted personal interviews. The unit of analysis of the dataset is households. The data consisted of three strata: “core host”, which consisted of Kenyan nationals living in areas where at least 10 percent of the refugees reside; “wider host”, which consisted of nationals living in other areas; and “refugees”. The refugee sampling frame was obtained from the UNHCR’s database on asylum seekers and refugees, considering those registered in their system and living in Nairobi. Stratification was carried out by country and county of origin for refugees and host-country respondents, respectively. Due to the sensitivity of the information requested, the data from this section of the questionnaire were collected by trained female enumerators. Women aged 15–49 years were randomly selected in each household to respond to the women empowerment section of the questionnaire. The data sampled 4853 households, of which 2420 were refugee households. Of the 1579 refugee women randomly sampled, 1532 consented to respond to the women empowerment section, for a 97 percent response rate.

3.2. Construction of the Refugee Women Empowerment Index

To construct the women empowerment index, the study classified five domains under which indicators used to measure women empowerment would be placed. The classification of the domains was guided by the KNBS (2020) guide on the construction of women empowerment. Based on Kabeer (1999), the domains were anchored in the resource and agency components. The five domains are attitudes towards socio-cultural norms, access to human and social resources, decision-making over general household decisions, decision-making over sexual and reproductive health and rights, and economic empowerment. Equal weights were assigned to each domain to total 1. The weights were then assigned equally to the indicators in each domain; see

Table 1. Description of domains and indicators used in the construction of the refugee women empowerment index.

Domain (Ii)	Indicator (Iij)	Indicator Description	Weights
Attitudes towards socio-cultural norms (I1)	Knows the acceptable age for a woman/girl to be married (I11)	1 if the woman selects 18 years and above; 0 otherwise	w1 = 1/15
	Does not justify wife-beating (I12)	1 if the woman does not justify wife beating; 0 otherwise	w2 = 1/15
	View that FGM should be stopped (I13)	1 if the woman agrees FGM should be stopped; 0 otherwise	w3 = 1/15
Human and social resources (I2)	Exposure to the media (I21)	1 if the refugee woman has exposure to the media at least once in a week; 0 otherwise	w4 = 1/20
	Knowledge of contraception (I22)	1 if the refugee woman knows at least one method of contraception; 0 otherwise	w5 = 1/20
	Knowledge about where to access to contraceptives (male and female condoms) (I23)	1 if the refugee woman knows where they can access contraceptives; 0 otherwise	w6 = 1/20
	Knowledge about HIV/AIDs prevention and transmission (I24)	1 if the woman knows at least seven causes and preventive measures of HIV/AIDS; 0 otherwise	w7 = 1/20
Decision-making over general household matters (I3)	Decision-making power over their partner’s/husband’s income (I31)	1 if the woman says yes to being able to make decisions over their partner’s/husband’s income; 0 otherwise	w8 = 1/20
	Decision-making power on significant household purchases (I32)	1 if the woman says yes to participating in making decisions over significant household purchases; 0 otherwise	w9 = 1/20
	Decision-making over personal health (I33)	1 if the woman says yes to making decisions over personal health; 0 otherwise	w10 = 1/20
	Decision-making power on visiting relatives (I34)	1 if the woman says yes to making decisions over relatives visiting; 0 otherwise	w11 = 1/20
Decision-making over sexual and reproductive health and rights (I4)	Can say no to sex (I41)	1 if the woman says yes to being able to freely say no to sex; 0 otherwise	w12 = 1/10
	Is free to ask their partners to use a condom during sexual intercourse (I42)	1 if the woman says yes to being able to ask their partners to use a condom during sexual intercourse; 0 otherwise	w13 = 1/10
Economic domain (I5)	Receives compensation for taking care of children below 14 years of age (I51)	1 if the woman receives compensation for taking care of children below the age of 14 years; 0 otherwise	w14 = 1/10
	Refugee woman is in continuous paid employment (I52)	1 if the woman is in continuous paid employment; 0 otherwise	w15 = 1/10

.

Wife beating was included under attitudes towards socio-cultural norms, as research highlights wife-beating to be a deeply rooted social-cultural norm (Ramage, 2018). Similarly, this research acknowledges the unique challenges that women face in accessing and making decisions over sexual and reproductive health and therefore, treats this as a domain. The selection of indicators was based on the women empowerment module provided in the Refugee and Host Household Survey (RHHS) 2021 questionnaire. See Table 1 for the description of domains.

The individual refugee woman empowerment index (RWEI) was constructed as follows:

$$RWEI={\sum }_{k=1,i=1,j=1}^{15,5,n}{w}_k{I}_{ij}$$

(1)

where ${\sum }_{k=1}^{15}{w}_k=1$.

3.3. Conceptualization and Empirical Model

Our conceptualization implies that the refugee women empowerment pathway is through factors associated with demographic characteristics and factors regarded as empowering factors such as education (Jejeebhoy, 2000; Mahmud et al., 2012; Upadhyay & Karasek, 2012). The empowerment process is exhibited in five dimensions: attitudes towards social-cultural norms; human and social resources; decision-making over sexual and reproductive health and rights; and decision-making over general household matters and economic resources. The study hypothesizes that household demographics, including marital status, sex of the household head, religion, and age, affect refugee women empowerment.

The fractional logit regression model was used to estimate the determinants of women empowerment among refugee women living in Nairobi, Kenya. The fractional logit model allows for appropriate modeling in cases of a continuous outcome variable bounded on the interval (0, 1) (Wooldridge, 2011). The nature of the outcome variable (Refugee women empowerment), which is an index taking the values between 0 and 1, resulted in the choice of the model. To specify the model, we assign the outcome variable refugee women empowerment Y, where Y is continuous on the interval between 0 and 1. Therefore, if $0\le Y\le 1$ is explained by a $1\times k$ vector of explanatory variables denoted by X, where $X=\left(x_1,x_2,\dots ,x_k\right),$ then the estimation of $E\left(X\right)$ can be modelled using a Generalized Linear Model (GLM), where, in this study, a logit link function is adopted. Therefore,

$$E\left(X\right)={\displaystyle \frac{e^{X\beta }}{{1+e}^{X\beta }}}$$

(2)

where $Y$ is the continuous outcome variable of refugee women empowerment, $\beta =({\beta }_1,{\beta }_2,\dots ,{\beta }_k)$ is a $1\times k$ vector of estimated coefficients for the independent variables, and $X=(x_1,x_2,\dots ,x_k)$ is the $1\times k$ vector of the explanatory variables.

See

Table 2. Description of control variables.

Variable	Description
Dependent Variable
Refugee women empowerment	Continuous variable ranging from 0 to 1
Independent Variables
Age of the selected woman	Continuous variable
Gender of the household head	Dummy variable = 1 when female; 0 otherwise
Education level of the selected woman
No formal education	Dummy variable = 1 when respondent has no formal education; 0 otherwise
Primary education	Dummy variable = 1 when respondent has primary education; 0 otherwise
Secondary education	Dummy variable = 1 when respondent has secondary education; 0 otherwise
University education	Dummy variable = 1 when respondent has university education; 0 otherwise
Marital status
Monogamous marriage	Dummy variable = 1 when respondent has a monogamous marriage; 0 otherwise
Polygamous marriage	Dummy variable = 1 when respondent has a polygamous marriage; 0 otherwise
Divorced/widowed/separated	Dummy variable = 1 when respondent is divorced/widowed/separated; 0 otherwise
Single/never married	Dummy variable = 1 when respondent is single/never married; 0 otherwise
Religion
Christian	Dummy variable = 1 when respondent is Christian; 0 otherwise
Muslim	Dummy variable = 1 when respondent is Muslim; 0 otherwise
Other religion	Dummy variable = 1 when respondent is other religion; 0 otherwise
Employment status of the household head	Dummy variable = 1 when the household head is employed; 0 otherwise
Education level of the household head
No formal education	Dummy variable = 1 when household head has no formal education; 0 otherwise
Primary education	Dummy variable = 1 when household head has primary education; 0 otherwise
Secondary education	Dummy variable = 1 when household head has secondary education; 0 otherwise
University education	Dummy variable = 1 when household head has university education; 0 otherwise
Household receives gifts from outside Kenya	Dummy variable = 1 when household receives gifts from outside Kenya; 0 otherwise.

for the description of the variables used in the modeling.

4. Results and Discussion

4.1. Descriptive Results

A descriptive analysis of the indicators used in constructing the women empowerment index was undertaken, including testing for mean differences between refugees and Kenyan nationals; see

Table 3. Descriptive results for the indicators used in the measurement of empowerment.

Variable	Refugees			Nationals
	n	Mean	Standard Deviation	n	Mean	Standard Deviation	t-Statistic (Mean Difference Between the Two Samples)
Empowerment at 80 percent of the total weighted indicators (Nationally recommended threshold)	1579	0.06	0.23	1658	0.22	0.17	12.39 ***
Attitudes towards socio-cultural Norms
Knows the acceptable age for a woman/girl to be married	1529	0.92	0.24	1588	0.99	0.11	7.96 ***
Does not justify wife beating at all	1531	0.81	0.39	1591	0.83	0.37	4.06 ***
View that FGM should be stopped	1506	0.94	0.24	1584	0.97	0.18	1.97
Access to human and social resources
Exposure to the media	1531	0.63	0.48	1591	0.92	0.27	17.36 ***
Knowledge of family planning methods	858	0.90	0.30	1092	0.96	0.18	5.18 ***
Has access to family planning	1517	0.56	0.49	1590	0.91	0.29	22.25 ***
Knowledge of HIV/AIDS	623	0.80	0.39	495	0.90	0.30	4.19 ***
General Household Decision-making
Decision-making power over their partner’s/husband’s income	856	0.63	0.48	1092	0.62	0.49	−1.39
Decision-making power on significant household purchases	858	0.78	0.42	1092	0.84	0.37	0.34
Decision-making power on visiting relatives	854	0. 76	0.43	1091	0.84	0.36	1.86
Decision-making power on their personal health	857	0.80	0.40	1092	0.89	0.30	3.72 ***
Household Sexual and Reproductive Health Decision-making
Can say no to sex	857	0.46	0.50	1087	0.63	0.48	8.19 ***
Is free to ask their partners to use a condom during sexual intercourse	852	0.25	0.43	1084	0.50	0.50	12.18 ***
Economic Empowerment
Receives compensation for taking care of children below 14 years of age	163	0.12	0.32	193	0.05	0.23	−2.31 **
Woman is in continuous paid employment	1579	0.41	0.49	1658	0.51	0.5	6.58 ***

Note: *** shows significance at p < 0.01, while ** shows significance at p < 0.05.

.

Table 3 shows that 6 percent of the refugee women respondents living in Nairobi are empowered at 80 percent of total weighted indicators, compared to 22 percent of the nationals living in Nairobi. The difference in the means is significant (p < 0.01). Nationally, 29 percent of Kenyan women were empowered at 80 percent of the weighted indicators using the Kenya Demographic and Health Survey (KDHS) data collected in 2014.

According to Zegeye et al. (2022), attitude towards wife beating is a proxy for a woman’s perception of their status. Therefore, women who cannot justify wife beating are assumed to possess a greater sense of self-esteem and awareness, which reflects positively on their level of empowerment (Khan & Islam, 2018). In the sample, 81 percent of the refugee women living in Nairobi do not justify wife beating, compared to 83 percent of nationals. In Kenya, FGM is illegal, as outlined in the FGM Act 2011 (Towett et al., 2014).

However, the practice is still deeply rooted in some African cultures. As shown in Table 3 above, 94 percent of the refugee women respondents agree that FGM should be stopped, while the corresponding figure for nationals is 97. According to Sheikh et al. (2023), the prevalence of FGM in Nairobi was less than 20 percent. The performance on this indicator can be attributed to exposure to urbanization and modernization. Another norm that is still practiced in some cultures in Kenya and Africa in general is early marriage. Regardless of gender, marriage to a person below the age of 18 is considered a violation of human rights and is illegal in Kenya (Warria, 2019). In the sample, 92 percent of the refugee women know the right marriage age, compared to 99 percent of the nationals.

There is a significant difference (p < 0.01) in the exposure to media experienced by refugees (63 percent) and nationals (92 percent). Media exposure is critical for education and social awareness, and this difference highlights the greater isolation of refugees, which may limit their access to empowerment information and resources such as knowledge of and access to family planning; Ochako et al. (2015) identified a significant positive association between the delivery of targeted messages on family planning through media channels and the uptake of family planning. Despite the result that 90 percent of refugee women in the sample know about family planning, only 56 percent have access to it, which is much lower than the 91 percent reported by national respondents. According to Gitonga and Gage (2024), most refugees residing in urban areas lack legal status and therefore are not covered by health policies such as access to free contraceptives. In addition, 80 percent of the refugee women know about HIV/AIDS, compared to 90 percent of the national respondents.

Peterman et al. (2021) identify women’s involvement in decision-making as a key indicator of empowerment. While this proxy is widely accepted, challenges remain in its quantification, particularly in terms of the prioritization of what decisions are representative of empowerment. Misra et al. (2021) provide a basis for measuring decision-making by highlighting that women should take part in economic decisions and decisions about their social well-being. In this study, women’s involvement in decision-making is classified into two domains: decision-making about general household issues and sexual/reproductive health. The mean difference in decision-making over personal health is significant (p < 0.01). As shown in the table, 80 percent of the refugee women make decisions regarding their personal health, compared to 99 percent of the nationals.

There is a significant difference (p < 0.01) in being able to say no to sex among refugees (46 percent) and nationals (63 percent). Similarly, there is a significant difference (p < 0.01) in being able to ask their partners to use condoms during sex by refugees (25 percent) and nationals (50 percent). According to Seidu et al. (2021), polygamy, sociocultural factors, and education levels influence women’s ability to negotiate for sexual preferences freely, and women in urban areas are likely to have more control over their sexual and reproductive health and rights than rural women.

In many studies on women empowerment, economic empowerment is a key domain for measuring the success of women empowerment efforts. In this study, economic empowerment was measured using two indicators: evidence that a woman is in continuous paid employment and when a woman receives compensation for childcare. As shown in Table 3, 12 percent of the refugee women respondents receive compensation for the child care of children, compared with five percent of nationals.

Gender roles, especially childcare, are strongly associated with women in African cultures, preventing women from participating fully in the labor market (Jayachandran, 2020). Childcare is seldom viewed as formal work but rather as a responsibility tasked to women. In this regard, women’s economic participation is limited. In addition, 41 percent of the refugee women in the sample are in continuous paid employment, compared to 51 percent of nationals. Vuni and Buhendwa (2023) discussed the hurdles faced by refugees in the quest to obtain work permits, as compared to obtaining business licenses, finding that most refugees reported that the processes are tedious and complicated, to an extent that few refugees can obtain work permits.

Table 4. Descriptive results for the control variables.

	Refugees			Nationals
Control Variables	n	Mean	Standard Deviation	n	Mean	Standard Deviation	t-Statistic (Mean Difference Between the Two Samples)
Age of the selected woman	1532	28.57	8.04	1591	29.30	7.33	1.99
Household head gender	1532	0.53	0.50	1591	0.34	0.47	−11.37 ***
Education level of the selected woman
No formal education	1512	0.31	0.47	1589	0.027	0.16	−20.80 ***
Primary education	1512	0.38	0.43	1589	0.22	0.41	−9.99 ***
Secondary education	1512	0.29	0.47	1589	0.66	0.48	22.12 ***
University education	1512	0.03	0.27	1589	0.08	0.33	8.46 ***
Marital status							−
Monogamous married	1531	0.53	0.50	1591	0.68	0.46	8.06 ***
Polygamous married	1531	0.02	0.13	1591	0.004	0.68	−3.31 ***
Divorced/Separated/Widowed	1531	0.14	0.35	1591	0.093	0.29	−3.99 ***
Single/Never married	1531	0.30	0.46	1591	0.21	0.40	5.12 ***
Religion							−
Christianity	1531	0.61	0.49	1589	0.93	0.25	19.88 ***
Muslim	1531	0.35	0.48	1589	0.04	0.19	−21.24 ***
Other	1531	0.04	0.19	1589	0.04	0.16	−0.14
Employment status of the household head	1532	0.71	0.45	1591	0.91	0.29	16.14 ***
Household receives gifts from outside Kenya	1532	0.01	0.07	1591	0.002	0.05	−1.58
Education level of the household head
No formal education	1516	0.33	0.47	1589	0.03	0.16	−24.91 ***
Primary	1516	0.25	0.43	1589	0.21	0.41	−3.39 ***
Secondary	1516	0.34	0.47	1589	0.64	0.48	18.52 ***
University	1516	0.08	0.27	1589	0.12	0.33	4.89 ***

Note: *** shows significance at p < 0.01.

below shows the descriptive results for the control variables used in the regression model.

A significant difference (p < 0.01) is observed in the gender of the household heads in the sample, with 53 percent of the refugee households headed by women, compared to 34 percent for nationals. According to Doocy and Lyles (2018), many female-headed families arise during displacement, as men face challenges such as being unable to free conflict areas. Another significant (p < 0.01) difference pertains to education; 31 percent of the refugee women in the sample have no formal education, compared to 2 percent of the nationals. Similarly, 22 percent of refugee women and 66 percent of the nationals have achieved secondary education. The largest category of refugee women falls under primary education (38 percent). Marital status also varies significantly (p < 0.01) between the nationals and the refugee women. Among the respondents, 53 percent of refugee women and 68 percent of nationals are in monogamous marriages, while 14 percent of the refugee women are divorced, separated, or widowed, compared to 9 percent of the nationals.

There is a significant (p < 0.01) difference in religious affiliation; while 93 percent of Kenyan respondents identify as Christian, 61 percent of refugees share this affiliation. Conversely, 35 percent of refugees identify as Muslim, compared to 4 percent of nationals. This diversity could have implications for social integration and community cohesion in the host country. The employment status of household heads also shows a significant difference, with 91 percent of national household heads employed, compared to 71 percent of refugees. There is a significant disparity in the education levels of household heads between refugee women and nationals. From the sample, 33 percent of the household heads in refugee households have no formal education compared to 3 percent of the nationals’ household heads. Similarly, 64 percent of the nationals’ household heads have completed secondary education compared to 34 percent of the refugee household heads. According to Betts et al. (2024), the level of education of household heads in refugee settings explained variations in the statuses of refugee. The study showed that with higher levels of education, the refugee households had greater access to social protection and social capital.

4.2. Fractional Logit Regression Results

To investigate the enablers and impediments of women empowerment, a fractional logit regression model was adopted where the refugee women empowerment index was regressed on individual and household level socioeconomic variables. We present the coefficients and marginal effect; the results are reported in

Table 5. Regression Estimates and marginal effect for factors enabling women empowerment among refugee women in Nairobi Kenya.

Variables (Reference/Base Category)	Coefficient (Linearized Standard Error)	Margins (dy/dx) (Standard Error)
Age of the selected woman	0.009 *** (0.001)	0.002 *** (0.004)
Household head (Male)
Female	0.16 *** (0.03)	0.04 *** (0.008)
Education level of the selected woman (No formal education)
Primary education	0.06 (0.04)	0.01 (0.009)
Secondary education	0.09 ** (0.05)	0.02 ** (0.01)
University education	0.17 (0.09)	0.04 (0.02)
Marital status (Monogamous married)
Polygamous married	−0.23 ** (0.10)	−0.06 ** (0.03)
Divorced/Separated/Widowed	−0.97 *** (0.04)	−0.23 *** (0.008)
Single/Never married	−0.94 *** (0.03)	−0.22 *** (0.007)
Religion (Christianity)
Muslim	−0.17 *** (0.03)	−0.04 *** (0.007)
Other	0.02 (0.07)	0.005 (0.02)
Employment status of the household head (Not Employed)
Employed	0.16 *** (0.03)	0.04 *** (0.007)
Household receives gifts from outside Kenya (No)
Yes	0.09 (0.11)	0.02 (0.03)
Education level of the household head (No formal education)
Primary	0.10 ** (0.04)	0.02 ** (0.01)
Secondary	0.05 (0.04)	0.01 (0.02)
University	0.09 (0.07)	0.02 (0.01)
Number of observations	1508
Number of strata	1
Number of Primary Sampling Units (PSUs)	2	F−statistic (p−values) = 0.00

Note: *** shows significance at p < 0.01, while ** shows significance at p < 0.05.

. See Table 5.

The marginal effects results provide insights into how the individual and household characteristics influence the probability of empowerment among refugee women. The results suggest a positive relationship between a refugee woman’s age and her empowerment; for each additional year, empowerment increases by 0.2 percentage points (p < 0.01). This trend could be attributed to the accumulated life experience, greater access to resources, and increased social capital that come with age (Stromquist, 2015). Older women might have more opportunities to engage in community activities and leadership roles, enhancing their empowerment (Ansari et al., 2012; Batool & Jadoon, 2018).

The identity of the household head also matters; a female-headed household increases the probability of empowerment by four percentage points (p < 0.01). This finding suggests that refugee households led by women are more likely to foster environments where women have access to opportunities. Female household heads might prioritize the well-being and empowerment of other women in the household, promoting a culture of support (Kes & Swaminathan, 2006). The attainment of a secondary level of education increases the probability of a refugee women’s empowerment by two percentage points (p < 0.05) relative to refugee women who have no formal education. Secondary education provides women with essential skills and knowledge, enabling them to participate more actively in economic and social activities (Malhotra et al., 2009). It also broadens their understanding of life issues, enhancing their ability to advocate for their rights and interests and contributing to their overall empowerment.

Notably, being in polygamous marriage for refugee women lowers the probability of being empowered by six percentage points (p < 0.05) compared to refugee women in monogamous marriages. However, the situation is worse off for divorced, separated, widowed, or separated women and single or never married refugee women, whose probability of being empowered decreases by 23 and 22 percentage points, respectively (p < 0.01) compared to refugee women in monogamous marriages. As refugee women, these marital statuses are associated with greater social and economic vulnerabilities, reduced access to resources, and diminished social support networks (Aoláin, 2011). This suggests that the refugee women holding such statuses have fewer economic opportunities and less social support compared to their married counterparts.

The results indicate that Muslim refugee women have a four-percentage-point-lower probability of being empowered compared to Christian refugee women (p < 0.01). This result suggests that Muslim women in Nairobi’s refugee community face cultural or religious constraints that hinder their empowerment. These could include restrictive gender norms and limited access to education or employment opportunities (Abdi, 2016). Efforts to enhance women empowerment in this context should consider culturally sensitive approaches that respect religious practices while promoting gender equality and women’s rights.

The employment status of the household head increases the probability of empowerment by four percentage points (p < 0.01) when the head is in an employed status compared to when they are not employed. Similarly, household heads with primary education are associated with a two-percentage-point increase in the probability of women’s empowerment compared to those with no formal education.

5. Discussion

The findings from the study highlight the significant influence of individual and household characteristics on the empowerment of refugee women living in Nairobi, Kenya. Comparing the characteristics of refugee women to those of nationals provides valuable insights into the unique circumstances and challenges faced by refugee women. Refugee women empowerment is critical in achieving gender equality, an important United Nations SDG. The issue of refugee women empowerment also aligns with other SDGs: SDG 4 on access to quality education, SDG 8 on decent work and economic growth, and SDG 10 on reduced inequalities.

The study finds that the incidence of women empowerment among refugee women living in Nairobi is much lower (6 percent) compared to that of nationals living in Nairobi (22 percent) and well below Kenya’s national expectations. It is clear that there is a substantial gap in empowerment among Nairobi-based refugee women. While some studies like that of Vuni and Buhendwa (2023) identify the legal framework and bureaucratic hurdles like obtaining business permits as key issues preventing the economic participation of refugee women, other studies like that of Ritchie (2018) show that Somali refugee women living in urban areas of Nairobi have developed mechanisms to increase their economic participation through embracing collective action. However, empowerment is a multifaceted concept involving the interplay of various dimensions, not limited to the economic component. A definition by UNHCR (2001) characterizes empowerment as “a process through which women and men in disadvantaged positions increase their access to knowledge, resources and decision making power and raise their awareness of participation in their communities in order to reach a level of control over their own environment.” This definition reinforces the domains classified in this study to collectively measure empowerment. The gap is noticeable in the areas of economic empowerment. exposure to media, and sexual and reproductive health and rights. Similar findings by Logie et al. (2022) show the challenges faced by forcibly displaced persons in urban low- and middle-income countries in accessing sexual and reproductive health services, including contraception.

In addition, the study finds that refugee women’s age, education level, marital status, and religion, along with the household head’s gender, employment status, and education, are significant determinants of women empowerment. While age, gender of the household head, education level of the refugee women, and employment of the household head increased the probability of empowerment, religion and marital status modalities decreased the probability of empowerment among the refugee women. Similar findings by Abdulai et al. (2024) show that education and gender of the household head were enablers of women empowerment in Mali. The findings by Stromquist (2015) note that older women may have greater opportunities to engage in leadership roles within their communities, which enhances their agency. Similarly, Ansari et al. (2012) found that empowerment increases with age due to improved social networks and greater access to resources. The attainment of secondary education among refugee women increased their probability of empowerment compared to those who had no formal education. A study by Upadhyay and Karasek (2012) highlights the impact of education in enhancing health outcomes and involvement in decision making. Secondary education provides the refugee women with the right skills and knowledge they can use to navigate the host environment in their resettlement and secure employment opportunities (Stromquist, 2006). In addition, Abdi (2016) explores how certain religious practices among Somali refugee women in Kenya limit their mobility and access to education and employment. However, these findings are context-dependent, as religion can also provide social cohesion and support networks in certain situations (Ling & Bouma, 2008).

The methodology and results provide a foundational framework for measuring women’s empowerment among refugee populations. However, the study acknowledges a key limitation in its operationalization of empowerment. While the index incorporates relevant indicators of economic participation and caregiving compensation, it does not comprehensively capture the multidimensional nature of empowerment. This limitation is primarily due to the constraints of the data used, as the available dataset only included the indicators used in this study. The RHHS 2021 questionnaire is extensive but does not include some key issues that pertain to empowerment among refugee women, such as digital empowerment, psychological issues, and legal know-how. The GSMA 20243 report on the mobile gender gap highlights key indicators that can be adopted for measuring women empowerment and inclusion in the digital space. In addition, the aspect of business as an income-generating activity for refugees is not comprehensively covered. The World Bank (2023) report on socioeconomic profiles of refugees highlights that more educated refugees are not likely to be employed. This data’s limitation on business perspective omits essential issues that affect women economic empowerment. These gaps form a basis for future research on measuring and predicting women empowerment. Future studies should adopt a more comprehensive approach to empowerment by incorporating additional factors such as legal rights, mental health, and community engagement, which are integral to the refugee experience and could offer deeper insights into their empowerment dynamics.

6. Conclusions and Implications

In conclusion, the study provides insights into the varied nature of refugee women and their empowerment. Both individual and household factors have been shown to determine the chances of empowerment among the refugee women. Similarly significant gaps between refugee women and nationals in their empowerment and individual characteristics have been clearly demonstrated by this study. These findings also underscore crucial policy implications for promoting women empowerment among refugee women living in Nairobi. Empowering refugee women is not just about achieving the ethical mandate of addressing humanitarian issues, but it is also a strategic approach towards addressing the complexities refugee women face during resettlement. Refugee women play a pivotal role in the family unit in terms of building a strong foundation when resettling. When empowered, they are better placed to support refugee families, reducing overreliance on humanitarian aid. Additionally, empowered women are more likely to invest in their children’s education and health, leading to better long-term outcomes for the entire family and building overall resilience and stability in refugee populations, ultimately leading to more successful and sustainable resettlement outcomes.

Based on the study’s results, three key policy messages arise, focusing specifically on refugee women. The first highlights the critical role of economic empowerment in enhancing the overall well-being, household resilience, and autonomy of refugee women. According to Vuni and Buhendwa (2023), refugees find it easier to access business permits relative to work permits. A holistic approach to economic empowerment that addresses both waged employment and self-employment sources of income is encouraged. The establishment of refugee-specific platforms or centers that educate refugee women on the processes and requirements for obtaining work permits will increase awareness and increase access to formal employment opportunities and business opportunities. These platforms could provide information on the application process, required documentation, and legal rights related to employment. In addition, access to credit and financing for starting businesses, and advocating for enabling environments where refugee businesses are thriving, is key to the successful economic empowerment of refugee women.

The second key policy message concerns decision-making on sexual and reproductive health and rights. The Global Compact on Refugees underscores the importance of SRHR in improving health outcomes among refugee women and girls. Despite the declaration, though, countries remain reluctant to commit to issues of SRHR among refugees. Strategies such as the establishment of sexual and reproductive health clinics and subsidizing the cost of accessing such services in hospitals could promote access and women empowerment in this space. These strategies require a multi-stakeholder approach, where negotiations for such subsidies are spearheaded by the stakeholders within humanitarian groups.

Education emerges as a crucial enabler for refugee women empowerment. The third key policy message heightens the need to re-integrate the refugee population in schools. Several barriers, such as the documentation and allocation of resources to education needs, may pose challenges in accessing education. A framework for the recognition and validation of prior learning qualifications will increase refugees’ access to higher levels of learning. Empowerment holds immense potential for refugee women, offering them the opportunity to transform their lives and contribute meaningfully to their communities. Through targeted policies and support efforts, much more can be unlocked for better resilience among refugee households.

In addition, the study contributes to a broader application of Kabeer’s (1999) framework on resources and agency by applying it to the case of refugee women living in Nairobi Kenya. Applying Kabeer’s framework in the context of refugees provides an important understanding of how resource access and agency are reshaped by displacement. Similarly, the study shows the decision-making dynamics within the household and household characteristics; in cases where households are female-headed, there is a higher probability of empowerment. This underlines how forced displacement disrupts traditional gender roles, demanding a renegotiation of power dynamics within households.