Household gender decision-making dynamics and child reunification outcomes in the Northern region, Ghana

Introduction

The separation of children from family care is a serious child welfare issue worldwide, especially in povertystricken countries such as less developed nations where poverty, migration, trafficking and low social protection make children at high risk of family disruption. Recent estimates suggest that millions of children are still living in institutional care, in foster or kinship care systems or in informal care arrangements (Desmond et al., 2020; UNICEF, 2021) [8,29] . Thus, the goal of reuniting separated children with their natural families or main caregivers is now a focal point of current child protection and deinstitutionalisation policies (O'Kane & Lubis, 2016; Zewude et al., 2023) [21,34] . The ultimate goal of reunification is not necessarily the physical return of children, but also the stable reintegration, protection, wellbeing and long-term developmental outcomes (Frimpong-Manso et al., 2022; Tucker et al., 2025) [10] . Worldwide, the context and functioning of the household and family after the child's return at home is known to play a significant role in successful reunification. The household decision-making process has been a growing focus on the factors influencing reintegration outcomes, as it influences allocation of resources, responsibilities, and caregiving within the household (Acosta et al., 2020) [1] . Household gender decision making structures are defined as how men and women share power in the household and over household resources, which includes income, welfare, education, health, discipline and care, among other factors (Seymour & Peterman, 2018) [22] . Such arrangements for decision-making influence children's access to key resources and services and therefore impact sustainability of child reunification. Gender relations still shape the household governance and child welfare outcomes in Africa. Women are often the main caregivers and are likely to have limited access to household resources as well as strategic household decisions (Doss & Quisumbing, 2020; UN Women, 2020) [9,31] . Evidence from Tanzania, Ethiopia, and other African countries shows that inequitable bargaining power between spouses can impact investments in their children's education, nutrition and healthcare (Anderson et al., 2017; Gebre et al., 2021) [4,13] . Other studies also argue that the welfare outcomes of children in households with inclusive www.extensionjournal.com households or joint decision-making are superior to those of children in households where decision-making is centralized in one gender (Acosta et al., 2020; Voss et al., 2024) [1,32] . Yet, the impact of such decision-making processes on child reunion is not well understood. In sub-Saharan Africa, the separation of children from their families is especially high, resulting from poverty, migration, child labour, trafficking and informal foster arrangements. UNICEF (2022) states that about two thirds of children in the region suffer multidimensional poverty, resulting in limited ability to provide stable care at home. Informal kinship care and fostering are common and often not well monitored or supported, leading to children being at risk of neglect, exploitation and multiple separations (Bentum & Manful, 2022; Mpofu & Machingauta, 2024) [7,20] . This means many children who are reunited are challenged with sustaining stable family life and realizing a positive welfare outcome after reintegration. Child protection issues are related to internal migration, trafficking, child labour and household poverty in Ghana. The challenges are especially highlighted in the northern region of Ghana which has one of the highest poverty rates in the country and high child migration to the southern urban centres (GSS, 2021) [12] . Previous studies have identified that many children from the north are trafficked, engaged in domestic domestic labour, street connected activities, and/or residential care placements, before being reunited with their families (Balfour et al., 2020; Koomson et al., 2025) [6,16] . Family-based care and reunification are important initiatives in child welfare in Ghana; however, evidence shows that reunification outcomes are not uniform because of economic hardship, lack of institutional support, and poor monitoring of reunification post-care (Frimpong-Manso et al., 2017; Teye & Adjetey, 2025) [11,25] . The Northern Region is a region which is particularly relevant to exploring the link between gender decisionmaking and child reunification outcomes. Strong patriarchal norms, extended family norms, high dependency ratios, and livelihood insecurity are key characteristics of the region (GSS, 2021) [12] . Men in most of the houses have the final say on the finances and strategic issues, while women look after the children and do the day to day care taking (Agyei-Sarpong, 2025; Tufuor et al., 2016) [2,28] . These may affect the ability of reunited children to have access to food, education, health care, emotional support, and other resources they need to support successful reintegration. If a woman is less influential in making decisions for the household than she is expected to be in her role as caregiver, the welfare needs of what are often the children of the reunited household may not be given adequate weight. While studies have been conducted on reintegration stability, child protection systems and challenges in child welfare institutions in Ghana (Frimpong-Manso et al., 2017; Frimpong-Manso et al., 2022) [10,11] , there has been little empirical research on the influence of household gender decision-making patterns and reunification outcomes. The current literature has mainly concentrated on poverty, institutional support and readiness of the caregivers with a lesser focus on intra-household power dynamics and decision-making systems that affect children's wellbeing post-reunification. In Northern Ghana where gender norms and cultural expectations are deeply ingrained in the household, this knowledge gap is of special interest. It is therefore essential to understanding the impact of household gender decision making dynamics on child reunification outcome to enhance child welfare governance and improve reintegration programmes in Ghana. This study builds on the current initiatives to advance sustainable family reintegration and support improved wellbeing of children reunited with their families in the Northern Region of Ghana by exploring how gendered authority, care-giving and child welfare outcomes relate to one another.

Literature Review 2.1 Household Gender Decision-Making in African Contexts

Household gender decision-making is the allocation of authority and bargaining power between men and women in a household in five domains: income control, child health care, education spending, and care work (Seymour & Peterman, 2018) [22] . The intra-household bargaining theory suggests that the better off a family is, the more evenly distributed the decision making power is within the household (Doss & Quisumbing, 2020) [9] . But in many African contexts, what seemed to be a shared arrangement is actually a negotiating process, meaning women's participation is not substantive (Acosta et al., 2020; Anderson et al., 2017) [1,4] . Traditionally, women have been marginalized from strategic decision-making in Northern Ghana, and men assume the roles of decision-makers and leaders, whereas women participate mainly in day-to-day care and consumption choices (Agyei-Sarpong, 2025; Tufuor et al., 2016) [2,28] . Additional evidence from other sub-Saharan African contexts also shows that poverty strengthens traditional power structures and reduces women's bargaining power and the capacity to invest in child welfare fields (Anderson et al., 2017; Meijer et al., 2015) [4,18] .

Outcomes for child reunification and factors that influence outcomes

Child reunification is now seen as a multi-dimensional, long-term process, and not just one event at the end of the line (Frimpong-Manso et al., 2022; Zewude et al., 2023) [34,10] . Household poverty, caregiver preparedness, length of separation, and post-reunification monitoring are among the factors that have been identified in systematic reviews as important factors in achieving reintegration stability (Zewude et al., 2023; Mpofu & Machingauta, 2024) [34,20] . In Ghana alone, the lack of institutional follow-up, lack of economic stability of receiving households, and the complexity of kinship care are documented as significant barriers (Frimpong-Manso et al., 2017; Bentum & Manful, 2022) [7,11] . Trafficking and internal child labour migration are the dominant routes of child separation in the northern belt, with the possibility of being reunited often complicated by poverty-related household instability and stigma (Balfour et al., 2020) [6] . In many instances, extended family kinship networks act as protective mechanisms, but also as a source of inconsistencies in caregiving authority and in the distribution of resources (Bentum & Manful, 2022; Alhassan, 2017) [7,3] . Age of reunification has also been a recurring factor in predicting the difficulties of www.extensionjournal.com reintegration, with older children having more difficulties (James et al., 2017; Frimpong-Manso, 2022) [15] .

Methodology

The study used a concurrent mixed methods design which combined both quantitative and qualitative methods to give a full picture of the dynamics of household decision making and of the process of child reunification. The respondents for the quantitative data were randomly selected using purposive sampling and systematic random sampling techniques from 270 participants in the three study districts of Tamale Metropolis, Kumbungu District and Mion District in the Northern Region of Ghana. The respondents were primary caregivers of households which had experienced child reunification. The sample was computed based on Krejcie and Morgan's formula with an estimated population of the households that were reunited within the study area. A semi-structured household survey instrument was used to collect the socio-demographic information, household decision-making, and child reunification experiences. The Household Gender Decision Making Index (HGDI) was built as a composite index of household authority patterns, in three categories: male-dominated, female-dominated and joint/shared decision-making. Reunification outcomes were judged stable or unstable based on continuity of residence, access to school, access to health facilities, caregiver reported child wellbeing and re-separation events. Data were analyzed using SPSS, descriptive statistics, Chisquare tests of association and binary logistic regression modelling. The qualitative data was gathered using the method of key informant interview with twelve informants consisting of social welfare officers, NGO practitioners, children home administrators and community leaders, and analysed using thematic content analysis. The University for Development Studies Ethics Committee gave ethical clearance. Everyone consented to participate in the study and confidentiality of the data was maintained throughout the study.

Results

Socio-Demographic Characteristics of Respondents

Table 1 shows the socio-demographic characteristics of the 270 individuals who were interviewed from the households that had been reunited with their children. A slight majority of the females (53.7%) were compared with the males who made up 46.3%. The age distribution of the respondents was fairly even with 22.2% in the age group 40-49 years. The majority (26.7%) were married, followed by widowed (25.6%) and divorced/ separated (24.4%). When it comes to educational attainment, a large percentage of the respondents (40.7%) had no formal education and only 13.0% had attained the primary education level. Those who have Junior High School education were 18.5%, Senior High School education were 16.7%, and tertiary education were 11.1%. This finding indicates that child reunification households tend to have relatively low educational level. The major religions were Islam (61.1%), Christianity (33.3%) and traditional religion (5.6%). In terms of household structure, the most common type was extended family households (40.7%), followed by polygamous (31.5%), nuclear (16.7%) and single-parent households (11.1%). In terms of poverty status, most of the respondents (57.4%) were considered as poor, with 42.6% considered as non-poor, reflecting the economic vulnerability of reunified households.

Patterns of Household Gender Decision-Making 4.2.1 Distribution of Household Decision-Making Authority

Table 2 present the distribution of household decisionmaking authority, with education-related decisions being used as a proxy that indicates decision-making authority. Joint decision making with the husband and the wife was the highest proportion (35.2%) and the household heads were the next largest (25.9%). The 14.8% of decisions were influenced by NGO, and the 16.7% of decisions by extended family members. In 7.4% cases, female caregivers had independent authority. The relatively low percentage of independent female decision making is related to the fact that gender differences are still prevalent in households. Key informants explained that even when the household explained that they made decisions together, the men generally had the final say on large-scale financial and education investments. Women are generally not included when decisions are made about finances in households; the man makes most of the financial decisions, yet women are nearer to the children, as one Social Welfare Officer said: "Women are normally not consulted when it comes to financial aspects of decision making in the household, it's the man who makes most of the financial decisions" (Social Welfare Officer, Interview 1). www.extensionjournal.com

Household Gender Decision-Making Index (HGDI)

The HGDI classification is shown in Table 3. The most common type of decision-making was Joint/shared decisionmaking (households 60.0%). The proportion of femaledominated arrangements is slightly larger than the number of male-dominated arrangements (21.9% compared to 18.1%). The increasing number of joint decision making could indicate an increased awareness of cooperative household governance, yet qualitative evidence suggests that participation does not equalise.

Association Between Household Characteristics and HGDI

Selected household characteristics and decision making patterns were related using a chi-square analysis. The household structure and the household size did not significantly affect HGDI (Household Structure: χ² = 4.87, p = 0.561 and Household size: χ² = 5.92, p = 0.432), suggesting that structural and demographic variables do not affect the authority arrangement. Likewise, there was no significant relationship between the degree of HGDI and caregiver education (χ² = 7.18, p = 0.517), meaning that education does not directly lead to a change in gender authority in this context. Statistically significant associations were however found between poverty status and sex of the household head. The results in Table 4 indicate that the non-poor households were significantly higher than the poor households in terms of joint decision making (95 households versus 67 households) with a high Chi-square value (χ² = 6.84, p = 0.033). This fits with the intra-household bargaining theory, which associates economic resources with bargaining power. The relationship between sex of the household head and HGDI is given in Table 5. A higher proportion of femaledominated decision making was observed in female headed households (34) than in male headed households (25), and a higher proportion of male dominated decision making was observed in male headed households (38) than in female headed households (11). The association was significant (χ² = 8.26, p = 0.016) and therefore the distribution of decision making power is influenced by the gender of household headship.

Prevalence and Patterns of Child Reunification 4.3.1 Reunification Stability Outcomes

Overall reunification outcomes are shown in Table 6. Of the 270 children whose reunification was examined, 162 (60.0%) had a stable reunification, and 108 (40.0%) had unstable outcomes for their reunification. As discussed in the regression analysis below, most stable reunifications happened in homes where parents shared decision-making, which was also reported as a greater role in the majority of homes in the sample.In homes, the majority of which reported joint decision-making, the majority of stable reunitions took place, as discussed in the regression analysis www.extensionjournal.com below. 40.0% of the outcomes were unstable, highlighting the instability of the reunification process if the families do not receive on-going support in the home and/or the institution.

Pathways of Child Separation and Reunification

The key pathways in which children were separated and then brought back together are shown in Table 7. The most common pathway (18.1%) was trafficking, the victimisation of children, particularly in the impoverished communities of Northern Ghana, being exploited by trafficking networks. Children's homes accounted for 16.3% (as part of the continuing process of deinstitutionalisation) and foster/kinship care arrangements 15.6% of cases. There was also a high level of street involvement (13.3%), child labour migration (11.1%) and living with relatives (11.9%). The variable nature of the pathways illustrates the multi-causal aspects of child separation in this area.

Duration of Child Separation Before Reunification

The distribution of separation duration is given in Table 8. More than 60% of the children had been separated for over one year, while the largest group (23.7%) had been separated for less than six months. The vast majority of children in the study area experienced chronic family disruption before they were reunited, with 19.3% of children having been separated for more than 5 years, and 21.1% for 3-5 years. Larger separation periods can be linked to higher levels of emotional disconnection and increased reintegration challenges (Zewude et al., 2023) [34] .

Duration of Separation and Reunification Stability

Table 9 cross tabs duration of separation and reunification stability. In children separated at a setting that is not longer than six months, stable outcomes were similar to unstable outcomes (34 compared to 30). As the length of separation increased, however, the rates of instability grew quickly: 10 children were reunited in a stable way with their parents after 3-5 years of separation, but 47 became unstable; only 5 were reunited in a stable way after more than 5 years of separation, while 47 were unstable. The Chi-square test showed a statistically significant association (χ² = 39.84, df = 4, and p< 0.001) between prolonged separation and unstable reunition, which made it a significant predictor of reunition instability.

Marital Status and Reunification Stability

The relationship between marital status of the caregiver and reunification stability is presented in Table 10. The married caregivers had the most positive pattern (52 stable to 20 unstable), while the single caregivers had a fairly close split (31 stable to 32 unstable). The association was statistically significant (χ² = 12.48, df = 3, p = 0.006). This stability may be attributed to the fact that married households are more likely to have two caregivers who could be helpful to children when they are reunited in the household and to share resources within the household that benefit children when reuniting takes place. Relatively positive results also were obtained for widowed respondents (42 stable vs. 27 unstable), possibly because of support by the kinship network when one parent is not present.

Binary Logistic Regression: Predictors of Reunification Stability

The binary logistic regression analysis for the predictors of the stable reunification is presented in Table 11. The model was significant (LR χ², p< 0.001) and had a Pseudo R² of 0.248, which shows that the model has some level of explanatory power. The strongest negative predictor was duration of separation (β = -0.436, OR = 0.647, p< 0.001): For every 1-day increase in separation duration, the chance of stable reunification decreased by approximately 35.3%. Household poverty status also had a significant impact on reducing stability odds (β = -0.582; OR = 0.559; p< 0.001), which lowered the odds by 44.1%. Household structure was associated with significantly lower reunification stability for extended (OR = 0.481, β = -0.731, p = 0.002) and polygamous (OR = 0.597, β = -0.516, p = 0.022) households suggesting problems in distribution of resources and accountability for care in complex household structures. The odds of being reunited at a stable outcome were inversely associated with child age (stable reunification -17.1% per additional year, β = -0.187, OR = 0.829, p< 0.001). Child education, on the other hand, was a significant positive predictor (β = 0.281, OR = 1.324, p = 0.009) to improve the odds for stable outcomes by 32.4%. Single parenthood was not statistically significant when other factors were taken into account (p = 0.123).

Discussion

Overall, joint/shared decision-making is the predominate form of decision-making among households in the Northern Region, broadly mirroring the pattern observed in other African settings where there has been an increase in forms of formalized collaborative governance, whereas informal power dynamics remain (Acosta et al., 2020; Anderson et al., 2017) [1,4] . The qualitative findings in this study however, reinforce that there is no automatic link between joint decision making and equal bargaining power, notwithstanding that men have strong influence over strategic financial and education-related decisions in line with Doss and Quisumbing (2020). The significant association between poverty status and joint decision-making (χ² = 6.84, p = 0.033) is consistent with intra-household bargaining theory which suggests that where a household is poor, it tends to have a more patriarchal organization of decision making, because women have less bargaining power if the household is resource constrained. In the same way, the relationship between the sex of the head of the household and HGDI (χ² = 8.26, p = 0.016) is in line with Seymour and Peterman (2018) [22] and Voss et al. (2024) [32] and indicates that the sex of the primary authority figure has an effect on the governance structure of the household. The 60.0 per cent stable reunification rate is encouraging, but this must be set against the 40.0 per cent who had unstable reunification outcomes which are significant and indicate structural vulnerabilities in reunification systems that are ongoing. The continued high figures for reunification through trafficking (18.1%) are a reflection of the ongoing socioeconomic factors behind child separation in the region, as observed by Balfour et al. (2020) [6] . Household poverty and household separation for a long period were the two most significant findings of the binary logistic regression analysis for unstable reunification. The findings of the negative relationship between extended and polygamous household configurations and reunification stability corroborate the claims of Bentum and Manful (2022) [7] and Alhassan (2017) [3] that the nature of kinship support systems is ambivalent, as it can spread out responsibility for caring for children and leave them inadequately supported. The positive association between child education and stable reunification (OR = 1.324) further supports the tendency of continued schooling to support sustainable family reintegration, and therefore has policy implications for incorporating education services into post-reunification monitoring programs.

Conclusion and Recommendations

This study offers empirical data regarding two important aspects of child welfare governance in Northern Ghana, namely, household gender decision making and child reunification outcome. In the study area, joint decision making predominates, but is influenced by gender inequalities that are based in patriarchal norms and economic deprivation. Household structure, size and caregiver education were not seen to be independently significant in this context, whereas poverty and sex of the household head were key structural factors significantly associated with decision making patterns. Child reunification is accompanied by a high degree of instability and 40.0% of reunified children suffer adverse outcomes. Older child age, prolonged separation, household poverty and complex family structure are protective factors for reunification stability, whereas child education and dualcaregiver (married) households are risk factors. The results of this study have implications for child welfare policy and practice in Ghana. Participating in the study, the researchers recommend that:

(1) gender-sensitive household assessment tools be incorporated into pre-reunification screening activities to identify households vulnerable to decisions being marginalized by their household members; (2) household economic empowerment programmes be strengthened, especially for poor households and female-headed households; (3) early reunification be the priority, whenever possible, to minimize the negative impact of prolonged separation; (4) support for school enrolment be included as part of post-reunification packages; and (5) institutional monitoring capacity be strengthened to accommodate the complexity of household structures in extended and polygamous household settings.

References

1. Acosta M, Lorimer S, Chirwa E. Gender and decisionmaking in African households: A review. Journal of Development Studies. 2020;56(4):789-805.
2. Agyei-Sarpong K. Gender relations and household www.extensionjournal.com decision-making in rural Ghana: Evidence from Chereponi District. African Journal of Social Work. 2025;15(1):45-62.
3. Alhassan A. Child care institutions in Northern Ghana: Between welfare and exploitation. Child Welfare Review. 2017;4(2):33-51.
4. Anderson S, Kebede B, Voss A. Spousal disagreements on household decision-making in Tanzania. World Development. 2017;90:136-148.
5. Atanga E. Post-reunification child adjustment in Ghanaian households: Evidence from Accra. Ghana Journal of Child Welfare. 2023;9(1):14-29.
6. Balfour K, Osei-Hwedieh B, Teye JK. Trafficking and reintegration vulnerability in Northern Ghana. African Human Rights Law Journal. 2020;20(2):312-335.
7. Bentum I, Manful E. Informal foster care and child reunification in Ghana. Social Work and Society International Journal. 2022;20(1):1-18.
8. Desmond C, Sherr L, Tomlinson M, Skeen S. Global estimates of children living in institutional care. Child & Youth Care Forum. 2020;49(2):193-204.
9. Doss C, Quisumbing A. Towards a more complete measure of women's empowerment: Evidence from Ghana. World Development. 2020;135:e105053.
10. Frimpong-Manso K, Akabzaa T, Birikorang A. Determinants of reintegration stability for children leaving residential care in Ghana. Children and Youth Services Review. 2022;136:e106440.
11. Frimpong-Manso K, Korang-Okrah R, Manful E. Social work practice and child welfare in Ghana: Challenges and prospects. International Social Work. 2017;60(3):716-728.
12. Ghana Statistical Service (GSS). Ghana 2021 Population and Housing Census: General Report. Accra: GSS; 2021.
13. Gebre M, Muluneh M, Taddele M. Women's education and household decision-making in Ethiopia. African Journal of Reproductive Health. 2021;25(1):130-141.
14. ILO. Child Labour: Global Estimates 2020, Trends and the Road Forward. Geneva: International Labour Organization; 2021.
15. James S, Monn A, Palinkas LS, Leslie LK. Reunification from foster care: Children, family, and service factors. Children and Youth Services Review. 2017;74:137-147.
16. Koomson D, Quampah T, Antwi J. Trafficking survivors' reintegration challenges in Northern Ghana. Journal of Trafficking and Human Exploitation. 2025;9(1):77-93.
17. Korang-Okrah R, Frimpong-Manso K, Asante K. Child protection and social work practice in Ghana. Ghana Social Work Journal. 2016;2(1):1-22.
18. Meijer J, Nweke E, Khadiagala G. Kinship and gender norms in child welfare in Malawi. International Journal of Social Welfare. 2015;24(3):240-252.
19. Milligan I, Withington R, Connelly G. The role of residential care and deinstitutionalisation in child protection globally. Edinburgh: Centre for Excellence for Looked After Children in Scotland; 2016.
20. O'Kane C, Lubis P. UN Guidelines for the Alternative Care of Children: Implementation guide. New York: UNICEF; 2016.
21. Seymour G, Peterman A. Understanding the relationship between time and women's empowerment: Evidence from three countries. World Development. 2018;109:204-216.
22. Shawar YR, Shiffman J. Governing child welfare globally: Agenda setting and accountability failures. Global Policy. 2023;14(1):4-16.
23. Sitjes-Figueras A, Mas-Expósito L, Subirats-Dolcet N. Outcomes of child reintegration post-residential care: A systematic review. Child and Adolescent Social Work Journal. 2025;42(1):25-47.
24. Teye JK, Adjetey A. Child welfare governance in Ghana: Between policy aspiration and practice realities. Journal of Social Policy. 2025;54(2):310-328.
25. Teixeira P, Carvalho J, Delgado P. Family reunification from foster care in Europe: Predictors, process and outcome. Children and Youth Services Review. 2022;141:e106593.
26. Tucker M, Bhabha J, Murnan E. Child reintegration outcomes: A global review. Child Welfare. 2025;102(3):5-30.
27. Tufuor TK, Asante KO, Barimah KB. Gender dynamics and household reintegration among returned migrant women in Northern Ghana. Gender, Place & Culture. 2016;23(9):1324-1338.
28. UNICEF. Orphans. New York: UNICEF; 2021. https://www.unicef.org/social-policy/orphans
29. UNICEF. Child Poverty in Africa: A MultiDimensional Approach. New York: UNICEF; 2022.
30. UN Women. Unpaid Care Work: The Missing Link in the Analysis of Gender Gaps in Labour Outcomes. New York: UN Women; 2020.
31. Voss A, Kwabia R, Atta N. Joint decision-making and symbolic participation: Evidence from rural Ghana. Gender and Development. 2024;32(1):55-72.
32. World Bank. Women, Business and the Law 2022. Washington, DC: World Bank; 2022.
33. Zewude G, Mekonnen A, Hirpa T. Reintegration of children from institutional care in sub-Saharan Africa: A systematic scoping review. International Social Work. 2023;66(4):1125-1141.