Abstract
The COVID-19 pandemic has had a negative effect on children’s mental health worldwide. Existing studies suggest that children with greater levels of hope are more likely to be resilient in the face of disaster. While social support at the family and community level is proposed as an important factor in sustaining and fostering hope, the children of underprivileged households in developing countries tend to lack this support. We investigate whether development projects run by international NGOs are able to fill this gap and help children to remain hopeful during the pandemic. Using original survey data from 834 children in adolescence (aged between 10 and 18) in Kenya and Zambia, we show that children participating in Good Neighbors’ child sponsorship programs and community development projects exhibit higher scores on the Children’s Hope Scale than do non-participating children. These projects appear to foster hope by providing emotional and informational support.
Introduction
Disasters can have a profound psychological impact on the affected population, and children are especially vulnerable to this (Fothergill, 2017; Lock et al., 2012; Masten et al., 2015). During the COVID-19 pandemic, the loss of loved ones, concerns regarding personal safety, and uncertainty surrounding household finances have generated deep anxiety regarding the future and feelings of hopelessness across the globe (Walsh, 2020). This psychological harm can act as an internal constraint on child development (Glewwe et al., 2014; Ross et al., 2021).
Social support at the family and community level can mitigate psychological damage to children in difficult situations and bolster their hope (House, 1981; Lai et al., 2018; Rodriguez & Cohen, 1998; Sahranç et al., 2018). Children from low-income households in developing countries, however, tend to lack this support and experience particular exposure to psychological damage from disasters (Masten & Narayan, 2012; Shammi et al., 2020). Thus, international non-government organizations (NGOs) have in recent decades worked to support underprivileged children in developing countries (Lewis et al., 2020).
Recent empirical studies have confirmed that the child sponsorship programs offered by international NGOs show a positive psychological effect. Children who participate in such programs exhibit higher aspirations, more positive life attitudes, and greater hope than do those who do not participate (Glewwe et al., 2014; Ross et al., 2021; Tirrell et al., 2019). This psychological effect is a key mediator in the higher socio-economic achievement of sponsored children in their adult life (Wydick et al., 2013, 2017), which is in line with earlier research highlighting the importance of psychological capital on human and economic development (Luthans et al., 2006, 2010).
We contribute to this strand of the literature by examining whether the positive psychological effects of these programs would be seen even during the most difficult times. In particular, we investigate whether child sponsorship programs and community development projects (CDPs) run by international NGOs would reduce the negative psychological effects of COVID-19 on children and help dissipate their feelings of hopelessness during the pandemic. We focus on the emotion of hope because it is an important form of psychological capital in times of disaster (Masten & Narayan, 2012) that can strengthen the ability to cope and rebuild (Snyder, 2002; Walsh, 2020).
We survey the children of Kenya and Zambia, two sub-Saharan African countries that have been severely impacted by the COVID-19 pandemic. As discussed in more detail in Section III, our sample includes children who are participating in child sponsorship programs and CDPs run by Good Neighbors, a large international development NGO founded in the Republic of Korea. Our results show that the children who are participating in the Good Neighbors programs exhibit a greater degree of hope than do non-participating children, as measured using the Children’s Hope Scale (Snyder, 2002; Snyder et al., 1997). Our focus group interviews (FGIs) with Kenyan children provided evidence indicating that the positive association between participation in the Good Neighbors program and children’s hope relates to informational and emotional support more than to instrumental support.
This study makes three important contributions to the literature. First, we provide an important and timely attempt to bridge the gap between the burgeoning literature on the psychological effect of child sponsorship programs and the literature on the psychological effects of disasters on children. The former strand of research has yet to empirically incorporate the COVID-19 crisis, and the latter focuses predominantly on children living in developed countries (Fothergill, 2017; Glewwe et al., 2014; Masten, 2021; Ross et al., 2021). Our study is one of only a few that have investigated the potential benefit that child sponsorship programs and CDPs have for psychological support in children in the context of massive and exogenous health and economic shock.
Second, many studies that have investigated children’s psychological experiences have relied either on direct surveys of the children alone or only on secondary information obtained from adults (Fothergill, 2017; Peek, 2008). As noted, the mental health of children is significantly affected by family factors. Using both child and adult surveys for each household allows us to control for household-level factors, including the parent’s subjective understanding of COVID-19 and depression.
Third, we consider a recent trend in which international NGOs have begun to place greater importance on community-based projects than on traditional one-to-one child sponsorship. The traditional approach has been criticized for encouraging dependency among sponsored children and discriminating against non-sponsored children, leading to negative ingroup-outgroup sentiment within communities (Pratten, 2007; van Eekelen, 2013). However, the efficacy of community-based interventions has not yet been rigorously investigated. Ours is one of few studies that have empirically compared these two intervention types and provided preliminary evidence on the benefits of this type of community-based project.
The remainder of this paper is structured as follows. In Section II, we review existing studies and present our theoretical expectations. Section III describes our research design, while we report and discuss our findings in Section IV. Section V concludes with policy implications.
Psychological Effects of COVID-19 on Children and the Role of International NGOs
Psychological Effects of COVID-19 on Children
Previous studies have noted that the psychological problems experienced by children due to disaster are largely shaped by surrounding environmental factors. At the family level, household economic uncertainty and hardship and the resulting psychological distress in parents can negatively affect children’s psychological state (McLoyd, 1990). During COVID-19, the prolonged imposition of social distancing measures and school closures further contributed to high levels of parental stress (Jiao et al., 2020; Power, 2020). Studies have shown that parents who experience burnout due to a mismatch between parenting-related demands and the resources available to meet those demands tend to commit child abuse and maltreatment more often, which can physically and psychologically harm their children (Cusinato et al., 2020; Griffith, 2020).
The negative psychological impact of disasters also interacts with community-level factors. Because communities have an important role to play in providing children with economic, social, and human capital (Masten et al., 2015), disruption to the surrounding community usually has an adverse impact on children’s lives. For example, school closures during the COVID-19 pandemic and the resulting restriction of daily social activities and interactions can lead to an increased likelihood that children experience depression, anxiety, phobias, and loneliness (Courtney et al., 2020; Larsen et al., 2021). Further, school closures can lead to higher rates of child abuse, maltreatment, and violence in the home, which have devastating consequences for children’s mental health (Van Bavel et al., 2020).
By contrast, the robust social support provided by the family and community is an important mitigator for psychological distress in the face of a disaster (Brewin et al., 2000; Cao et al., 2020; Helgeson & Lopez, 2010; Ozer et al., 2003; Ye et al., 2020) and helps strengthen children’s psychological capital (Lai et al., 2018; Sahranç et al., 2018). Social support includes all types of accessible support from individuals, groups, and community ties (Cooke et al., 1988), and researchers often divide this pool into three categories: instrumental (e.g., the provision of money, time, and material support), emotional (e.g., the provision of empathy, affection, love, trust, and encouragement), and informational (e.g., advice, guidance, or useful information provided to help resolve a problem during stressful situations) (House, 1981; Rodriguez & Cohen, 1998).
At the family level, positive parent−child interactions and good communication among family members can provide children with strong emotional support, reducing their risk of developing psychosocial disorders during stressful periods (Jiao et al., 2020; Tso et al., 2020). At the community level, the financial and physical resources of the community can provide instrumental support while local education institutions and faith-based organizations can provide informational and emotional support. Community institutions of this type provide sites for children to form and strengthen their social networks and provide each other with support (Bokszczanin, 2012).
Disasters will likely cause the most psychological damage to children in whose environment robust social support is absent. Given that social support from the family and community is a major factor in resource endowment, children from economically disadvantaged households and communities in developing countries are likely to be the most psychologically vulnerable group during the COVID-19 pandemic. The following section discusses the role of programs run by international NGOs in providing these children with a much-needed psychological buffer.
Child Sponsorship Programs and Community Development Projects
As discussed in the previous section, the lack of social support at the family and community level can exacerbate the negative psychological impact of COVID-19 on the children of economically disadvantaged families in resource-scarce developing countries. We believe that the intervention of international NGOs can make a difference here by providing much-needed social support. We are particularly interested in how such support can dissipate feelings of hopelessness among children and encourage hope. Hope is a vital form of psychological capital during disasters (Masten & Narayan, 2012) that can strengthen efforts to cope and rebuild (Murphy, 1987; Taylor, 1989; Walsh, 2020; Werner & Smith, 2001). Previous studies have suggested that a range of types of social support can foster the development of hope and other positive attitudes (Kemer & Atik, 2012; Thoits, 2011; Watson et al., 2019).
We focus on the roles of two distinct yet overlapping types of intervention used by international NGOs: child sponsorship programs and CDPs. The positive effects of the former intervention on psychological development have been demonstrated (Glewwe et al., 2014; Ross et al., 2021; Tirrell et al., 2019), but few studies have been conducted on CDPs. This research gap is significant, as recently, greater emphasis has been placed on community-wide interventions. This trend reflects past criticism of child sponsorship programs. In particular, it is said that the tendency of child sponsorship programs to generate discrimination against non-sponsored children might lead to ingroup−outgroup conflict within the community (Pratten, 2007; van Eekelen, 2013). The move toward CDPs also reflects expert emphasis on community-level social capital (Shade, 2006).
In response to this trend, Good Neighbors has adopted a more community-oriented approach since 2008, with a one-to-one child sponsorship program maintained within CDP areas. As we describe in Section III, these community-level interventions provide children with multidimensional social support, including informational (e.g., COVID-19 awareness education), emotional (e.g., various club activities and counseling services), and instrumental (e.g., the provision of sanitary goods) support. This support likely helps reduce anxiety among children and to develop a positive outlook, in line with past research that posits community spirit as a source of psychological capital and hope as a community resource (Godfrey, 1987; Te Riele, 2010). Based on this, we present the following hypothesis:
Hypothesis 1: Children in a CDP area exhibit higher hope than children in a non-CDP area.
It is also important to note that within a CDP area, children who are selected to participate in one-to-one sponsorship programs are entitled to exclusive benefits, often in the form of instrumental support. As discussed in Section III, material assistance from Good Neighbors during the pandemic, such as study materials and food aid, focuses on sponsored children. These targeted benefits may have an independent positive effect on hope by reducing material deprivation. This leads to the following hypothesis:
Hypothesis 2: Within a CDP area, sponsored children exhibit higher hope than non-sponsored children.
Research Design
Case Selection and Good Neighbors in Kenya and Zambia
We chose Kenya and Zambia as intensity cases based on two criteria. Intensity cases are those that strongly experience the phenomenon of theoretical interest (Shakir, 2002). First, we focus on countries with adverse national conditions for tackling the global pandemic, as this is where the positive contribution of the international NGOs is most strongly needed. Kenya and Zambia are two of the three countries most severely hit by the COVID-19 pandemic in sub-Saharan Africa. The proportion of confirmed cases to the population in both countries at the end of 2020 was higher than 0.05%. Second, as the main purpose of this study is to determine the effectiveness of programs run by international NGOs, we focus on countries where these programs are being actively implemented. Kenya and Zambia are the largest beneficiaries of the Good Neighbors programs. As of 2020, Good Neighbors has launched six CDP areas in Kenya, sponsoring a total of 7,120 children; in Zambia, Good Neighbors has sponsored 5,015 children in four CDP areas.
In a CDP area, a range of projects is conducted to promote community development in terms of education, health, water, sanitation, and hygiene, income generation, advocacy, and community partnership. Both sponsored and non-sponsored children benefit from community-level projects. Sponsored children enjoy additional exclusive benefits, primarily in the form of instrumental support. Sponsored children are selected using a sophisticated process based on their needs using detailed information.
During the COVID-19 pandemic, both Good Neighbors Kenya and Good Neighbors Zambia have monitored children in CDP areas by visiting their homes as well as providing counseling via regular telephone consultations and household visits. They have both also distributed emergency hygiene kits and provided villagers with health-related education on COVID-19 to promote awareness and prevention activities. For the sponsored children, they additionally provide basic living and school-related expenses, as well as emergency food aid.
Survey Design and Implementation
Our data were drawn from original surveys conducted among 834 households in selected villages from CDP and non-CDP areas in Kenya (n = 434) and Zambia (n = 400). A CDP area typically includes multiple villages, although their size varies. In Kenya, we selected 10 urban and 20 rural villages belonging to urban and rural CDP areas, respectively, and 42 villages in non-CDP areas (5 urban and 37 rural villages). In Zambia, we selected 10 villages in CDP areas and six villages in non-CDP areas, all located in rural regions, because all of the CDPs in Zambia are active in such regions. To select the villages, we first identified the villages in CDP areas that were the most severely affected by COVID-19. We then selected villages in non-CDP areas that (1) had similar socio-economic conditions to the selected villages, (2) were located close to the selected villages, and (3) were affected by COVID-19 to a similar extent to the selected villages. The final list of villages in this study is available upon request.
In each village, the sample population was identified using stratified sampling that considered the location of the villages, and the gender of the children. In Kenya, we chose subjects from both urban and rural villages. In Zambia, as noted, the subjects were all from rural villages. In addition, we attempted to choose non-sponsored children from families whose socio-economic conditions were similar to those of the sponsored children.
There were three groups of theoretical interest: non-sponsored children in non-CDP areas (NonCDP), non-sponsored children in CDP areas (CDP:NonSpon), and sponsored children in CDP areas (CDP:Spon). Applying Tukey’s multiple comparison of means, we found that the average household incomes of the three groups were not statistically significantly different. The test results are available upon request.
We conducted two surveys, one to collect information from the head of each household and one to collect information from a child from each household. We asked the children to answer the survey independently without the assistance of adults to identify their true opinions about the situation. In Kenya, the surveys were conducted from March 20 to March 31 and from April 15 to April 22, 2021. The surveys in Zambia were conducted from March 8 to March 28, 2021. The human subject protocol of the research was evaluated and approved by the Institutional Review Board (Approval ID: HIRB-202107-HR-001).
In addition to the self-completed survey, we conducted FGIs with sponsored children (five to six children aged between 13 and 14 years old) in both countries. The interviews were all led by staff at Good Neighbors Kenya and Good Neighbors Zambia. Using these qualitative FGIs, we sought to understand the mechanisms through which Good Neighbors programs have affected the psychological condition of children during the pandemic. Both the surveys and questionnaires used in the FGIs are available upon request.
Operationalization
We adopted a multivariate regression framework to compare the level of hope among the three groups (NonCDP, CDP:NonSpon, and CDP:Spon) while controlling for influencing factors on hope.
Hope was measured using the well-known Children’s Hope Scale (Snyder et al., 1997), a six-item self-reported measure with two dimensions: agency (the odd questions) and pathway (the even questions). We created the Hope Score variable by aggregating the scores for the six items using principal component analysis (PCA). The resulting value equals the weighted average of the component scores, with the weights equivalent to the proportion of variance explained by each component. For our main variable Hope Score, we pooled the Kenyan and Zambian responses and conducted PCA because the loadings were similar for the two countries. The resulting variable is given in the form of a standardized index with a mean of 0 and a standard deviation of 0.4. Our main findings held when we performed PCA separately and constructed a country-specific Hope Score. Our key findings also held when we used simple averages for the aggregation method.
Independent and control variables
Our key independent variables were the two group dummies that indicated participation in the Good Neighbors programs (CDP:NonSpon, and CDP:Spon). The reference category was NonCDP. We expected children in the CDP:Spon and CDP:NonSpon groups to show significantly greater hope than the children in the NonCDP group, and we expected the difference between the hope reported by the CDP:NonSpon and CDP:Spon groups to be significant.
Our regression analysis included a set of control variables. We controlled for child age and gender. Several studies have demonstrated that the psychological impact of disasters is more severe for younger children (Norris et al., 2002; Shannon et al., 1994). There appear to be differences by gender in how children experience and express the psychological distress caused by disasters (Fisher, 2009; Garfin et al., 2014; Seddighi et al., 2021). Also included in our models is the child’s COVID-19 awareness. The respondents were asked how aware they were of COVID-19 on a scale from 1 (not at all) to 5 (entirely). Finally, we controlled for school attendance by recording whether the children were going to school over the previous 2 weeks. The possible answers were 0 (no), 1 (yes, often), and 2 (yes, regularly).
As discussed, children’s mental health is also influenced by family factors. We thus controlled for a series of household-level factors obtained from our household survey, such as household income and parents’ education (Lowe et al., 2013). We also controlled for the parents’ perception of COVID-19′s impact on their household and for their depression level. At the macro level, we control for country (Zambia) and urban-rural residence (Rural). Summary statistics are available upon request.
Findings
Our findings from ordinary least squares (OLS) regression analysis are presented in Table 1. As shown in Model 1, both group indicators (CDP:NonSpon and CDP:Spon) were positive and significant, indicating that the hope scores of these two groups were higher than that of the reference category (NonCDP). This held after controlling for child-level confounders (Model 2), household-level factors (Model 3), and macro-context variables (Model 4). In Models 5 and 6, we explore the association between our group indicators and the Hope Score in each country, and our findings hold.
Hope Scores for Children.
Note. Coefficient estimates with standard errors in parenthesis.
p < .1. **p < .05. ***p < .01.
Using the estimates from Models 5 and 6, Figure 1 visualizes the predicted Hope Score by country and group when all other variables are set to their respective means. The vertical lines indicate 90% confidence intervals. Note that, on average, children in Zambia exhibit lower levels of hope than those in Kenya. This difference is also captured by the negative and significant coefficient estimate for the Zambia dummy in Model 4. We attribute this country-level difference in the baseline of the hope score in part to the differences in the socioeconomic condition between the two countries. The overall income level of Zambia is lower than that of Kenya, and the lack of resources that can enable success can undermine the development of hope (Frankham et al., 2020; Kenny et al., 2010; Petterson & Friel, 2001).
Predicted Hope Score by country and good neighbors program participation.
More importantly for the purposes of this study, in both countries, the level of hope of children in the NonCDP group was clearly lower than that of the other two groups (CDP:NonSpon and CDP:Spon). These findings lend support to Hypothesis 1. Interestingly, in Kenya, sponsored children exhibited only a moderately higher level of hope than non-sponsored children within the CDP area. In Zambia, there was no meaningful difference in the hope scores between sponsored and non-sponsored children within a CDP area. The findings thus provide little support for Hypothesis 2. This may be because there are little additional hope-raising effects tied to the instrumental benefits gained from one-to-one sponsorship. It is possible that community-level intervention, which is given primarily in the form of informational and emotional support, is the dominant mechanism for fostering and sustaining hope.
To further investigate the underlying mechanisms, we conducted an additional analysis that included instrumental support as a four-category ordered measure in the model specifications. This measure is based on questions in our household survey asking whether the household received (1) COVID-19 related food aid, (2) hygiene and protective goods, or (3) educational equipment or materials from Good Neighbors during the pandemic. Table 2 reports the distribution of responses for each group. These data confirm that instrumental support was concentrated in families with a sponsored child. This, combined with our finding of a significant positive effect for CDP:NonSpon, hints that instrumental support is unlikely to be the main mechanism that explains the higher level of hope among CDP children.
Instrumental Support From Good Neighbors.
The regression results, including the Instrumental Support variable, are presented in Table 3. While Instrumental Support was significant and positive in Model 7, it was no longer significant when our group indicators were included in Model 8. Similar findings were observed for Models 9 and 10, which examined CDP areas only. Within CDP areas, Instrumental Support was not associated with Hope Score (Model 9). If the instrumental support provided by Good Neighbors was the key mechanism explaining stronger hope among sponsored children, CDP:Spon should have had a stronger positive association with Hope Score where the sponsored children received more instrumental support. However, we did not find any evidence for this interactive effect (Model 10). The results held when we analyzed the two countries separately (Models 11 and 12).
Hope Scores for Children with Instrumental Support.
Note. Coefficient estimates with standard errors in parenthesis.
p < .1. **p < .05. ***p < .01.
The findings together indicate that in CDP areas, children’s hope is cultivated not with targeted instrumental support but with more comprehensive social support. We by no means intend to downplay the importance of instrumental support, which can indirectly affect children’s hope by, for instance, reducing pressure on household incomes and improving social support at the family level. Our findings implicate that informational and emotional support such as campaigns and counseling can be a crucial yet overlooked means by which international NGOs can make a difference in cultivating children’s hope during a disaster.
This implication is in line with the findings of our FGI with sponsored children in a Kenyan CDP (Mukuru). Multiple children mentioned Safe Club activities as their favorite service among those offered by Good Neighbors. Safe Club is a club activity open to any child in the target village regardless of sponsorship. It is intended to encourage active participation in community affairs. Before COVID-19, Safe Club was mainly concerned with educating children about their rights. During the pandemic, Safe Club children actively participated in a campaign to improve COVID-19 awareness among village residents. In the FGI, one of the participating children said that: “It was great to participate in Safe Club. We can communicate with other children through Safe Club. Through these activities, I feel engaged.”
The activities of the Safe Club have little to do with the direct provision of material resources, but we believe the club activities aid children in developing a sense of belonging to the community while strengthening their hope. As Larson (2000) explained, structured youth activities can provide “an environment of possibilities for playful action, for initiative,” and “a context that demanded self-directed constructive attention over time” (Larson, 2000, p. 177): children “come alive in these activities,” which facilitates the development of an array of other positive qualities (Larson, 2000, p. 178).
In a follow-up survey question, we asked the children to imagine creating an organization to help children like themselves. Among the services they wished to provide, the participating children mentioned non-instrumental types of support as often as instrumental types. They believed that emotional and information support such as the “use of television and media” and “counseling and guidance” could help children cope better with COVID-19. They considered it important to provide children and people with the “right information about COVID-19, not rumors.” Overall, our findings indicate that non-instrumental support from NGOs can promote and sustain the psychological health of children during a disaster.
Conclusion
Children of economically disadvantaged families in developing countries are among those who have been most at risk of psychological harm during the COVID-19 crisis. Analyzing original survey data from Kenya and Zambia, we found that children in a CDP area tend to exhibit higher hope scores than children in a non-CDP area. In areas where CDPs are implemented, even those children who do not benefit from sponsorship programs exhibit a level of hope that is as high as that of sponsored children.
This study is not without limitations. Due to a lack of data, we were not able to utilize the difference-in-differences method where the control and treatment groups’ post-treatment (i.e., post-Good Neighbors program participation) differences are compared to the pre-treatment differences. We strove to address this limitation through adequate modeling of our cross-sectional data and refraining from interpreting the findings as an indication of clear causality. Future research will be able to use our dataset for more rigorous causal inference. Good Neighbors is indeed expanding its beneficiaries in Kenya and Zambia, and the non-CDP villages included in our sample are likely to be included in the CDP in near future. We expect that a follow-up study will be designed to incorporate the temporal differences in these villages. Later researchers should also test our hypotheses in other developing countries.
Despite its limitations, we believe that our study draws attention to the importance of emotional and informational support for fostering and sustaining hope among children. International NGOs that specialize in assisting children tend to focus on the delivery of instrumental support. We find that this instrumental support may not be the primary mechanism for helping children remain hopeful during a disaster. By expanding programs that provide informational and emotional support, NGOs can better fill the gap in social support experienced by vulnerable children.
Footnotes
Acknowledgements
We want to express our gratitude to Ilha Yi (Chairperson) and Minho Choi (Secretary-General) at GN GPC. We also thank Namun Heo (Country Representative) and Sunyul Jeun (Project Manager) from Good Neighbors Kenya, and Jaewoong Lee (Country Representative) from Good Neighbors Zambia. We appreciate the valuable comments from Hyunjin Choi, Hyo Won Lee, two anonymous reviewers, and the editor. While the paper shares some content with a project report (in Korean) prepared for the GN GPC that the authors and Hyeran Hong, Hyojung Kim, and Chaewon Sohn contributed to, it is entirely rewritten in an academic style.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by Good Neighbors Global Partnership Center (GN GPC) under the project title “The Psychological and Socio-economic Impact of COVID-19 on Children in Kenya and Zambia” in 2021.
