1 Introduction

“Understanding and improving well-being requires a sound evidence base that can inform policymakers and citizens alike where, when, and for whom life is getting better” (OECD, 2013:3). This quote from the introduction to OECD’s “Guidelines for measuring subjective well-being” illustrates the increasing importance of well-being among policymakers. There is a substantive body of literature identifying common factors of well-being. A growing strand of well-being research also recognizes subjective and contextual dimensions (Conceição & Bandura, 2008; Diener et al., 2018) by acknowledging that meaning and perceptions of the good life can vary across individuals and societies, and in a particular socio-cultural context.

The basic distinction relating to our study is whether well-being is best measured through objective measures (i.e. directly observed and commonly accepted indicators of welfare) or through subjective perceptions of individuals (Conceição & Bandura, 2008). In this paper, we focus on subjective well-being (SWB), i.e. how individuals feel about their own well-being. While SWB has been studied extensively in the Global North, the factors influencing SWB in the Global South (and specifically in Sub-Saharan Africa) have received a comparatively lower share of attention (Addai et al., 2014; Diener et al., 2018). This is unfortunate, given the recognized importance of SWB in human development (Ketu, 2023; Kulkarni et al., 2023), and its subsequent relevance for any policy and development intervention striving to reduce poverty and improve the quality of life (Beauchamp et al., 2018; Gough et al., 2007).

We respond to this gap by expanding recent research on factors influential to SWB on the micro-level in different contexts of the Global South (e.g., Huang et al., 2024; Pontarollo et al., 2020; Shams, 2014) through a case study of Western province, Zambia. We do so by (1) complex conceptualisation of SWB factors, (2) highlighting the links between SWB and different forms of inequality and deprivation at the household and community levels, and by (3) contributing to the existing empirical evidence on SWB specifically for Zambia.

To our knowledge, SWB in Zambia has been addressed in two studies so far (Holder et al., 2016; Phiri & Abebe, 2016). Holder et al. (2016) assessed how religiousness and spirituality influence SWB among children and adolescents. They found spirituality to be a strong predictor of higher life satisfaction among children. Phiri and Abebe (2016) conducted a qualitative case study of poverty in the Eastern Province of Zambia, with SWB forming only a part of the addressed dimension. Their findings highlight important contextual and relational factors influencing well-being. However, both studies have addressed only some aspects of well-being in children and adolescents.

Our paper examines the factors of SWB among inhabitants (N = 411) from rural areas of Zambia's Western Province (Mongu district). We are particularly interested in linking SWB with different forms of inequality and related deprivation. Inequality has become a prominent concern in the fields of economics and development, evident from the growing body of academic studies and literature addressing this issue (see, e.g., Atkinson, 2015; Milanovic, 2016; Scheidel, 2017; Wilkinson & Pickett, 2018; Wienk et al., 2022). The increased significance of inequality has also been recognisedby the UN’s Target 10.1 of the Sustainable Development Goals,Footnote 1 making within-country inequality goals part of the global development policy agenda for the first time.

We adopt a broader approach to inequality by focusing on its multiple dimensions (i.e., not only income or assets inequality). Moreover, we also work with intra-household inequalities which are often neglected in empirical studies (Chiappori & Meghir, 2015) due to the lack of data at the individual level. Therefore, we aim to address the main research question of whether and how intra-household inequalities (deprivations) and inequalities in social capital (varying social networks) relate to SWB.

The rest of the paper is organized as follows. In the next section, we conceptualize potential determinants of SWB, drawing upon insights from the existing literature. Following this, we describe the data, variables and methods used to explain SWB in our analysis. In the results section, we summarize the main findings, followed by a discussion of their implications, including potential policy recommendations. Finally, the concluding section offers a summary of our findings and suggests avenues for future research.

2 Theoretical Framework

2.1 Conceptualization of Subjective Well-Being

There are various concepts of well-being. The main difference lies in the concepts building on subjective versus objective approaches. Subjective indicators of well-being are based on a person's own experiences and evaluations. Measures of subjective well-being include a person's assessment of their life, feelings or emotional states, and sense of meaning and purpose in life (see OECD, 2013; Clark et al., 2018). The objective indicators for well-being attempt to reflect measurable characteristics of the phenomenon that are independent of people's evaluations. However, the construction of all indicators, including those measuring objective well-being, is influenced by the subjectivity of their creators, who make various normative decisions (such as which dimensions to include, and what weights to assign).

In fact, the objective and subjective approaches to measuring well-being are interrelated. The most notable indicator of objective well-being, the Human Development Index, builds on the capability approach. From its perspective, well-being is about capabilities to achieve things and way of life people have a reason to value (Sen, 1993). To put it differently, this conceptualization of objective well-being includes the various capabilities and their functions that result, among other things, in subjective well-being (Clark, 2005;). Conversely, when respondents are asked about subjective well-being (for example, life satisfaction), they are likely to consider factors such as their living standard or health, i.e., traditional dimensions of objective well-being. This is reflected by White (2010:162), who argues that “well-being emerges in the interplay of ‘objective’—that is, externally observable and independently verifiable aspects of people’s circumstances, and their ‘subjective’ perceptions and assessments of these”.

In this paper, we acknowledge the interconnectedness of subjective and objective-well being. Subjective evaluation usually includes one or more of the following three dimensions: (a) overall reflection on one’s life and related achievements (life satisfaction); (b) affect—feelings and emotions (happiness, sadness, anger); c) eudaimonia—evaluation of meaning and purpose in life. This study’s dependent variable of SWB is measured through overall life satisfaction. Although eliciting life satisfaction may be impacted by reliability issues, it is an established approach to assessing well-being and predicting future life choices (Kahneman & Krueger, 2006). To account for the close relationship between subjective and objective well-being, we include traditional dimensions of objective well-being as independent variables.

2.2 Factors Influencing SWB

The existing body of research on SWB suggests that it is influenced by a combination of individual internal and external determinants. Diener et al., (2018:255) distinguish between three main groups of theories aiming to explain different levels of SWB: (1) biological and genetic traits theories, (2) cognitive explanations, and (3) satisfaction with achieving goals and needs. Overall, a large part of SWB or life satisfaction can be explained by the former two groups of determinants. Nonetheless, there are arguably important context-dependent root factors such as living standards, sense of freedom, and different forms of inequality, that explain variance in SWB levels across and within societies, mainly by affecting the satisfaction with achieving goals and needs (Conceição & Bandura, 2008; Cordero et al., 2021). Therefore, the paper focuses on exploration of the third determinant and its associated root factors.

Some of the most prominent root factors influencing SWB include basic dimensions of objective well-being such as financial and material conditions, education level, and health (Elgar et al., 2011; Howell & Howell, 2008; Ngamaba et al., 2020; Voukelatou et al., 2021). The relationship between income and SWB has been addressed by Mentzakis and Moro (2009), who found that increasing absolute income levels improves SWB, though only up to a certain threshold and less than relative income levels (more recently e.g. Reyes-García et al., 2016). Higher levels of education tend to enhance SWB indirectly through higher income and improved health. However, education also often raises aspirations that, when unmet, can negatively impact SWB (Kristoffersen, 2018). While the precise nature of the relationship between living standards, education, and health (and other socioeconomic factors such as age and sex) remains a subject of ongoing debate, these variables consistently emerge as some of the most influential predictors of SWB (e.g. Huang et al., 2024; Kulkarni et al., 2023; Nikolaev, 2018).

Satisfaction with achieving goals and needs is closely related to the matter of achieving aspirations (Ibrahim, 2011) either in one area of life (such as ideal education or job), or in life generally. The latter can be viewed as a complex of multiple aspirations relating to an imagined life (Copestake & Camfield, 2010). In this sense, well-being (or, alternatively, ill-being) could be a result of an aspiration gap, i.e., the difference between the current situation and unfulfiiled desires (Ray 2006), or aspiration failure, i.e., the lack of capability to achieve aspirations and resulting outcomes (Bernard et al., 2011). The aspiration gap can lead to feelings of diminished control or freedom in life. Conversely, it can be argued that a lack of control or freedom can contribute to an inability to fulfill personal desires, thereby perpetuating the aspiration gap.

Inequalities influence satisfaction with achieving goals and needs through two pathways. First, they influence individuals' aspirations. Second, inequalities inherently limit opportunities for those positioned unfavorably in the distribution, such as individuals with poor health, lack of education, or financial resources, thus fostering a sense of deprivation. Generally, studies on subjective well-being confirm the importance of relative comparisons (and therefore distribution) as there is a consensus that on the national level, wealthier people manifest higher levels of happiness that those on the lower part of the income spectrum At the same time, economists still argue up to what extent increasing a country's average income makes people happier (see, e.g., Stevenson & Wolfers, 2012; Clark et al., 2018). Goff et al. (2018) investigated the relationship between subjective well-being inequality and the average reported level of well-being across over 160 countries, utilizing data from three international surveys, and consistently observed a negative correlation.

While different societies may be similar in terms of specific dimensions of inequality, such as health and income, the overall level of inequality can be amplified by the extent to which these particular dimensions of inequality overlap within a society (Syrovátka & Schlossarek, 2019). Moreover, this notion of overlapping inequality can be applied across various levels of measurement. Arguably, experiencing disadvantage across multiple levels—be it at the national, community, or household level—is more detrimental than facing it in just one or two of those contexts. Due to lack of data, inequality at the household level tends to be neglected (Chiappori & Meghir, 2015; Klasen & Lahoti, 2021). This is unfortunate, as, for example, Banda et al. (2017) have shown that gender inequality within Zambian households has negative implications for women’s well-being.

This brings us to the importance of social dimensions of SWB. SWB is inherently social and relational: aspirations and the capability to achieve a good life are forged within socio-cultural norms and relations (Appadurai 2004; Ray, 2006; Leung et al., 2011). Social capital is an important determinant of poverty, SWB, and happiness Sarracino, 2010; Rojas, 2018, Kim et al., 2021). Especially in contexts where inequality is high and the provision of public services and assets is limited, social networks and kinship ties constitute a valuable resource (Elgar et al., 2011; Helliwell et al., 2020). Huang and Fang (2021) suggest (income) inequality may be mitigated by “neighborhood social capital”. In the context of Zambia, studies have shown the importance of social capital in livelihood strategies and decision-making (Njuki et al., 2008; van Bastelaer & Leathers, 2006).

In the following section, we describe data collection, variables included in this study, our main hypotheses and data analysis.

3 Data and Methods

3.1 Data Collection

The survey was conducted in the area of Muoyo town and Mukukutu village south of the city of Mongu (Fig. 1). Muoyo was selected to represent a “typical” regional township with comparably good infrastructure, located close to the strategic M10 road, which is surrounded by a cluster of a larger number of settlements that are socio-economically and demographically similar to each other. Mukukutu represents a rural settlement with a high representation of small farmers, poor transport accessibility, and poor infrastructure, located on the edge of a floodplain that is extensively used for the cultivation of irrigation-intensive crops such as rice.

Fig. 1
figure 1

Location of the survey area

Full size image

Within the Muoyo-Mukukutu region, defined by the above-mentioned municipalities, stratified sampling was carried out by dividing each of the two municipalities into four geographically defined areas. Next, quotas were set for the representation of age categories, sex, and household heads (versus ordinary household members) for data collection purposes so that the structure of the resulting sample would roughly match available data or estimates of the actual socio-demographic structure of the region. The data enumerators were also given guidelines to construct a sample characterized by qualities similar to those derived from a stratified random sampling procedure. Following data collection, it was observed that all quotas were met except heads of households quota. The expected share of household heads for the Muoyo-Mukukutu area was lower than what we acquired. The discrepancy was probably caused by combination of sampling bias and multiple claims to headship (for more details about the discrepancy and related considerations, consult Schlossarek et al., 2024a, 2024b). Furthermore, certain groups within the population may have been underrepresented or excluded from the study. For example, individuals who were not available during the data collection period or who refused to participate could introduce another sampling bias.

The questionnaire pilot was followed by intensive training of the five data enumerators and then by the actual collection of the data, which took place from 18 to 28 July 2022.

The questionnaire covered a range of topics; notably SWB, human development indicators, various dimensions of social capital, questions covering intra-household inequalities, and the demographic characteristics of the respondents’ households (however, we highlight that the unit of the subsequent analysis was the individual, not the household). The data collection resulted in a sample size of 411 respondents, 229 from Muoyo and 182 from Mukukutu. Raw data are available at Schlossarek et al., (2024a, 2024b).

3.2 Variables Used in the Analysis

The aim of the survey was to identify the factors that influence the SWB of the respondents. This was measured using a question in which respondents rated their satisfaction with life on a scale of 1 (worst possible life) to 10 (best possible life). Specifically, we used the question measuring evaluation of present life in the Gallup World Poll (Helliwell et al., 2023). A list and description of all variables included in the regression model are given in Table 1.

Table 1 Variables used in the model explaining the respondents’ SWB
Full size table

Explanatory variables include two variables measuring demographic characteristics. These are sex and age. Other three variables (living_standard, health, education) summarise the three basic dimensions of human development (HD) as reported, for example, in the Human Development Index (UNDP, 2022). To assess the relationship between well-being and aspiration gap, we include a variable addressing respondents' perceived freedom (freedom) to choose what to do with their lives.

Another set of three variables operationalizes various deprivations of the respondent within their households in the areas of education (depriv_educ), access to resources (depriv_living), and decision-making (depriv_power). Next, we list variables approximating the respondent's social capital. Specifically, they focus on the following dimensions: informal relations (no_friends), formal relations (group_member), the existence of contacts that can arrange a valued service for the respondent (friend_car), and contacts to the government representatives (villhead_knows, gov_knows).

3.3 Research Hypotheses

While the overarching goal of the paper is to identify factors influencing the SWB of the respondents, our hypotheses are focused on the examination of intra-household inequalities and the role of social capital in communities as factors associated with the respondents’ SWB. The first set of hypotheses relates to intra-household deprivations. We address not only the respondent's education but also the household's highest-achieved education. When the respondent's educational attainment is comparatively lower, it can have a dual effect: a positive spill-over effect of knowledge and good leadership, thereby augmenting the respondent's well-being, or conversely, an aspiration gap, leading to a detriment in the respondent's well-being.

Similarly, a parallel dichotomy can be discerned in the case of power inequalities within a household. A decision-making role within the household can undoubtedly fulfil an individual's aspirations. On the other hand, it can also precipitate feelings of pressure and stress. Lastly, we hypothesise that limited access to household resources relative to other household members inevitably leads to a diminishment in the well-being of the individual subjected to this disparity. This hypothesis underscores the adverse consequences of intra-household resource imbalances and highlights the significance of material resource allocation in shaping individual welfare.

The second set of hypotheses consists of four assumptions which focus on the interplay between social capital and SWB. Firstly, we assume that membership in organized groups improves social capital and fosters societal inclusivity, consequently leading to an increased level of SWB. Secondly, we hypothesize that an absence of friendships exhibits a negative relationship with SWB. Thirdly, when hypothesizing about knowing people who can facilitate valued services for the respondent, a notable dichotomy emerges. These networks can serve as a substitute for limited resources, potentially mitigating the adverse repercussions of resource constraints and thereby increasing the respondent's well-being. On the other hand, direct interaction with individuals who occupy clearly superior socio-economic positions relative to the respondent may intensify the sense of deprivation stemming from socio-economic disparities. This can diminish the respondent´s SWB. Lastly, we hypothesize that close relationships with statutory and customary government representatives exhibit positive correlations with SWB. This association can be channelled through heightened perceptions of security, increased social prestige leading to an enhanced societal standing, and greater inclusion and influence within the community. Figure 2 shows the analytical approach used to test our hypotheses.Footnote 2

Fig. 2
figure 2

Visualisation of the analytical approach. Note: variables used for the concepts operationalization are in the parentheses. Highlighted in bold are variables used in the main regression models (see Table 3 below). Variables in italics are used in the sensitivity analysis presented in the Appendix

Full size image

3.4 Methods of Analysis

We begin our quantitative analysis by conducting a thorough examination of descriptive statistics. Next, we explore bilateral relationships between our explanatory variables and SWB. Depending on the characteristics of the explanatory variables, we analyze the relationship either by comparing means of subjective-wellbeing (categorical variables with two categories), Goodman and Kruskal’s gamma (ordinary variables with more than two categories), and Pearson correlation coefficient (quantitative variables). For comparisons of means, we report p-values of two-tailed Welch’s t-tests (unequal variances t-tests); for Goodman and Kruskal’s gammas and Pearson correlation coefficients, we report p-values indicating whether they are significantly different from zero.

The key part of the research is based on regression analysis which is used to assess how our explanatory variables are associated with SWB while controlling for possible confounders. Our dependent variable, SWB (or life satisfaction), is measured on a discrete scale going from 1 (worst possible life) to 10 (best possible life). With regard to the nature of our dependent variable, we use the Poisson regression technique which is often used to model discrete count variables.

We perform four regression models. In the first model, we use basic demographic and human development variables, among which we also include freedom of choice. In the second model, we add variables measuring intra-households inequalities (deprivations). Then, we replace deprivations variables by social capital variables in the third model. The fourth model is the final one and it includes all explanatory variables of our interest, i.e. variables capturing basic demographics, human development development (including freedom of choice), intra-household inequalities (deprivations), and social capital. All calculations are performed using statistical software Stata (StataCorp, 2017).

4 Results

We first look at the summary statistics of our dependent variable. The respondents’ SWB is generally low, with the average score rising just above 3 (3.05) on a scale of 1 (lowest) to 10 (highest), with only one of the 411 respondents choosing the highest score. In contrast, 47 respondents chose the lowest score. As seen in Fig. 3, the answers of most respondents range from 1 to 5.

Fig. 3
figure 3

Subjective well-being of respondents

Full size image

Descriptive statistics of all variables are presented in Table 2. In the last column of the table, we present bilateral associations of our explanatory variables with SWB. It can also be seen from Table 2 that some variables have more missing data points, which lowers the number of observations entering the bilateral analyses (Table 2), as well as the main regression models (Table 3 below).

Table 2 Variables—descriptive statistics and bilateral associations with SWB
Full size table
Table 3 The results of the regression analysis for the life_quality variable (SWB)
Full size table

We identify statistically significant bilateral associations, with a significance level of at least 10%, between SWB and the following variables: health, education, freedom, depriv_living, depriv_power, group_member, and friend_car. As expected, health, education, freedom, depriv_living, and group_member are positively associated with SWB. Depriv_power exhibits a positive relationship, and friend_car demonstrates a negative relationship. The remaining associations are statistically insignificant.

For the key part of the research, we construct four regression models as specified earlier to estimate the effects of our explanatory variables on the dependent variable (SWB). Table 3 presents the results of all models and it also shows that the coefficients and significance are stable across the models.Footnote 3

It can be seen in the table that Model 2 and Model 4 have significantly lower numbers of observations. This is a consequence of using the depriv_power variable in these two models. However, the lower number of observations itself does not affect the results of the regression models—the findings remain the same also in cases when this variable is excludedFootnote 4 or replaced with an alternative variable of a similar meaning.Footnote 5

Table 3 also shows that of the fourteen variables included in the final model (Model 4), ten are statistically significant at least at the 10% significance level (all else being equal—that is, controlling for all the other variables in the regression analysis). Regarding the basic demographic characteristics, sex is significant, indicating that men tend to express higher satisfaction levels than women. In contrast, age has no proven effects.

All three dimensions of human development (living_standard, health, education) are statistically significant (living standard and education at the 1% significance level, health at the 10% significance level) in the expected direction. Better living standards, improved health, and higher education all exert a positive influence on SWB. Similarly, we observe the anticipated relationship with freedom of choice: a higher degree of freedom corresponds to increased SWB (5% significance level).

Deprivation within the household pertaining to access to resources (depriv_living) reduces SWB (the result is significant at the 10% significance level). The model reveals a positive effect of educational deprivation (depriv_educ, significant at the 1% significance level). This suggests that when an individual within the household possesses higher educational attainment than the respondent, there is an increase in the respondent's SWB, holding all other variables constant. Being not involved in pivotal decisions within the household (depriv_power) has a positive effect on SWB in the model (10% significance level).

Social capital built through membership in different formal or informal groups (group_membership) positively affects SWB (10% significance level). Knowing people with the capacity to facilitate valued services for the respondent (friend_car) have a negative effect on SWB (the result is valid at the 1% significance level). Variable no_friends has an expected relation with SWB; however, the result is narrowly insignificant (p = 0.176). Close relationships with statutory (gov_knows) and customary (villhead_knows) government representatives have no proven effects.

5 Discussion

SWB is generally very low in the study area of microregion Muyo-Mukukutu, Western Province of Zambia. In the World Happiness Report 2023 (Helliwell et al., 2023), the most recent score of Zambia at national level is 3.98, which positions the country among the ten least satisfied countries where life satisfaction is reported (the data are available for 137 countries). In the area we surveyed, the score is just 3.05. A similar trend can be seen with freedom of choice: while in the surveyed area only 64% of people are satisfied with the degree of freedom they have, the Zambian national proportion is 79% (Helliwell et al., 2023). This can be possibly related to the generally lower living standards in the region compared to the rest of the country.Footnote 6

Our model indicates that women have lower SWB even when controlled for other factors. The comparatively lower level of SWB for women may be explained by persistent gender-related norms resulting in disadvantages in terms of access to land, resources or alternative livelihood sources in rural Zambia (Shipekesa & Jayne 2012; Evans, 2018). In our model, age is not significantly associated with SWB. Although the exact relationship between age and SWB is still widely debated (e.g., Biermann et al., 2022), most studies find either a U-shaped or linear effect of age on SWB (Blanchflower, 2021; López Ulloa et al., 2013). The insignificance of age in this case study may have several possible explanations, for example the limited sample size, or unaccounted contextual factors that may influence the impact of age on SWB.

As expected, we observe a positive relationship between objective development indicators and SWB. The basic three dimensions of human development—standard of living, health, and education—are among key determinants of SWB in our study. These results concur with previous research on SWB (Elgar et al., 2011; Howell & Howell, 2008; Kulkarni et al., 2023) and possibly confirm the validity and relevance of standard indices such as the Human Development Index or Multidimensional Poverty Index.

Concerning intra-household inequalities, our hypotheses encompass both positive and negative associations for education and power deprivation. Additionally, we argue that disparities in access to household resources exhibit a negative correlation with SWB. Our results confirm the hypotheses across all three dimensions under examination. Firstly, while it could be assumed that having a lower educational level evokes feelings of deprivation, our study shows the exact opposite. A more educated individual in the household seems to create a positive spillover effect, benefiting other members rather than causing feelings of deprivation. Indeed, better education has been linked to increased SWB through higher income and overall livelihood opportunities (see Kristoffersen, 2018 for a review), which may positively affect the household as a whole.

Secondly, individuals with limited influence over decision-making within the household manifested high-levels of SWB. This is in contrast to findings of Srivastava et al. (2021a). One plausible explanation lies in the reduced sense of pressure and stress that might be associated with decision-making and lower level of responsibility for the well-being of other household members.Footnote 7 Consequently, we can argue that the focus should potentially shift from "power deprivation" to highlighting the deprivation stemming from an overwhelming sense of responsibility. The responsibility deprivation logic may also partially explain our results concerning the educational attainment of different household members, as the expectations and pressures related to providing for the household may rise with an individual's education.

Thirdly, and as anticipated, restricted access to household resources exhibits a detrimental impact on SWB. When other individuals within a household are perceived as being materially better off, it leads to feelings of relative deprivation for the less affluent member. Additionally, it evokes a perception of inequity or disadvantage (Bolt & Bird, 2003). These results are in line with previous research addressing the role of intra-household resource distribution and decision-making (e.g., Bárcena-Martín et al., 2020). Interestingly, the results point to the complex (and possibly indirect) effects and mechanisms behind the different overlapping dimensions of inequality discussed in the first part of this paper. Our results show that while deprivation in terms of educational attainment seems to create positive spillovers and enhance paritcipants’ SWB, perceived material deprivation negatively affects SWB.

With regard to the social capital and relational aspects, we introduced earlier several hypotheses covering five dimensions in total. We confirm that group membership increases people's SWB. This can be attributed to factors such as a heightened sense of belongingness, augmented social support networks, enhanced interpersonal relationships, a shared sense of purpose and objectives with fellow group members (Elgar et al., 2011; Helliwell et al., 2020; Moose, 2010; Rojas, 2018; Sarraccino, 2010). However, we did not find a statistically significant relationship between the absence of friendships and SWB, which is not in accordance with previous research findings (Helliwell & Putnam, 2004; Leung et al., 2011).

We identify a somewhat surprising relationship between networks providing the capacity to facilitate valued services for the respondent and SWB. It seems that deprivation caused by direct interaction with individuals who occupy clearly superior socio-economic positions relative to the respondent outweighs the positive effect of substituting limited resources by having somebody who is able to provide help when needed.

In terms of formal social relationships, the model indicates no effect of having close relations with statutory (gov_knows) and customary (villhead_knows) government representatives. Several factors may account for this outcome, including resource limitations and inequality, insufficient political influence, corruption or low trustworthiness of these representatives (see for example Córdova & Layton, 2016; Jovanović, 2016; Tay et al., 2014). The impact of 'knowing someone' depends on the quality and nature of these relationships. When there is low trust or the relationships lack quality, it may lead to a diminished belief in the ability of leaders to enhance the well-being of the respondents (see also Addai et al., 2014). Consequently, having connections with statutory or customary representatives does not automatically translate into direct benefits in terms of one's well-being. Although this finding is somewhat surprising, particularly in light of previous research that examined various facets of social capital and trust, a recent study by Glatz and Schwerdtfeger (2022) suggests that the causal relationship between SWB, social capital, and trust may be more intricate.

Finally, we would like to note a few limitations of this study. Although we believe that the Muoyo-Mukukutu area is not considerably different (in terms of relations between SWB and factors that determine it) from other rural areas of Zambia (and other countries of Sub-Saharan Africa on a similar level of socio-economic development), the results presented in this paper are applicable primarily for the surveyed area. Our results may be subject to influences such as local habits, traditions, or environmental nuances not fully recognized by researchers. We also acknowledge that (potential) presence of sample bias within our dataset poses an inherent risk of unexpected influences on our study outcomes.

We note that the study’s cross-sectional design limits the ability to establish causal relationships between variables. Naturally, such a design fails to capture the dynamic nature of phenomena over time. This temporal limitation restricts us from discerning how variables evolve or interact longitudinally, potentially overlooking important developmental trajectories or fluctuations. In general, cross-sectional studies suffer from compromised validity and generalizability of findings, particularly when investigating complex phenomena such as subjective well-being. A longitudinal study would provide more robust evidence of the relationships between factors analysed in the study and subjective well-being.

6 Conclusion and Recommendations

This paper investigates multiple factors that impact SWB within the Muoyo-Mukukutu region of the Western Province in Zambia. Drawing upon existing well-being literature, our conceptual framework encompasses numerous determinants of SWB, with a specific emphasis on exploring the associations between SWB and various types of inequalities and associated deprivations, notably those pertaining to intra-household dynamics and social capital.

The results highlight the importance of addressing regional disparities in terms of HD and SWB and directing resources and initiatives to uplift disadvantaged areas such as Muoyo-Mukukutu. Our results also confirm the significance of HD dimensions (standard of living, health, and education) in shaping SWB. To this end, our findings support the current trend in development policy aiming at general improvevement of HD, reduction of gender inequalities in all spheres of HD, and other factors that are causing lower SWB among women.

Furthermore, our results suggest that encouraging more educated family members to impart their knowledge and skills to others can amplify the positive spillover effect of education and further contribute to SWB. Concurrently, policies can facilitate opportunities for lifelong learning and skill development within households. Such initiatives can have stronger effects in disadvantaged communities with low levels of education among all household members. Our findings support the idea that where resources are extremely constrained, focusing on the education of a single household member can be a viable strategy for improving SWB. However, this may reinforce the disadvantaged positions of members already facing various forms of discrimination.

Moreover, acknowledging that restricted decision-making power within households can contribute to improved well-being, policymakers may need to reassess strategies to ensure that empowerment efforts do not inadvertently overwhelm individuals with excessive responsibilities. Empowerment initiatives should target and focus on those people with very limited decision-making power within the household. Our result support placing greater emphasis on creating a better environment for sharing and alleviating responsibilities within households that is increasingly becoming a more integral part of empowerment initiatives. Additionally, these interventions can enable household members to make informed decisions regarding fair resource allocation, thereby enhancing SWB, particularly of those with limited access to household resources. Sharing responsibilities and resources may eventually help to reduce or alleviate discriminatory practices such as limited access to land for women.

Finally, policymakers should focus on equitable access to resources and services so that individuals do not feel overly reliant on informal contacts. One effective policy to address this issue could involve continuing and more targeted support of community groups. For instance, the awareness about the benefits of group membership might be strengthened through clearly articulated benefits for objective and subjective wellbeing hence encouraging individuals to actively seek opportunities to join or establish such groups. However, most importantly, group membership opportunities should be inclusive and accessible to all segments of the population, especially to marginalized and vulnerable groups. We note that to prevent the potential failure of the groups, the promotion of the group establishment and membership should be done in a culturally sensitive manner.

We invite other researchers to delve deeper into the subject of intra-household inequalities. Further investigation is needed to analyze the complex mechanisms behind “power deprivation” and deprivation stemming from an overwhelming sense of responsibility within the household. Since the “head of household” variable did not show a statistically significant effect on SWB, it would be valuable to conduct a more comprehensive examination of the roles and responsibilities assumed by individuals within households. Furthermore, it is important to investigate how power deprivation, roles and responsibilities of household heads, and the distribution of educational benefits within household interact.