Role of Savings and Internal Lending Communities (SILCs) in improving household wealth and financial preparedness for birth in rural Zambia

Savings and Internal Lending Communities (SILCs) are a type of informal microfinance mechanism adapted in many low- and middle-income countries (LMICs) to improve financial resources for poor and rural communities. Although SILCs are often paired with other health and non-health-related interventions, few studies have examined SILCs in the context of maternal health. This study examined the association between SILC participation, household wealth and financial preparedness for birth. The study also examined the association between sex and financial preparedness for birth. A secondary analysis was conducted on individual survey data collected from SILC participants in two rural districts of Zambia between October 2017 and February 2018. A convenience sample of 600 participants (Lundazi: n = 297; Mansa: n = 303) was analysed. Descriptive analyses were run to examine SILC participation and household wealth. Multiple binary logistic regression models were fit to assess the unadjusted and adjusted relationship between (1) SILC participation and household wealth, (2) SILC participation and financial preparedness for birth and (3) sex and financial preparedness for birth. The results show that SILC participation led to an average increase of 7.32 items of the 13 household wealth items. SILC participants who had their most recent childbirth after joining SILCs were more likely to be financially prepared for birth [adjusted odds ratio (AOR): 2.99; 95% confidence interval (95% CI): 1.70-5.26; P < 0.001] than participants who had their most recent childbirth before joining SILCs. Females were more likely to be financially prepared for birth than males if they had their most recent birth before joining an SILC (AOR: 1.79; 95% CI: 1.16-2.66; P < 0.01). SILC participation is shown to increase household wealth and financial preparedness for birth for both men and women. SILCs are a promising intervention that can help poor and rural populations by increasing financial resources and financially preparing parents for birth. A secondary analysis was conducted on cross-sectional SILC impact survey data collected from SILC participants in two Zambian districts: Lundazi and Mansa. At the time of the SILC implementation and impact survey, data were collected from Lundazi and Mansa. Part of Mansa has now been split to make a new district (Chembe), but this change occurred after the implementation of the SILC and does not affect the results. The authors partnered with a local non-governmental organization (NGO) to implement SILCs as part of the Maternity Waiting Home (MWH) project in rural Zambia. The MWH project aimed to understand the impact of MWHs on reproductive health service access and maternal outcomes for women living far (>10 km) from health facilities (Lori et al., 2018; Scott et al., 2018). The local partner implemented the SILCs and collected the survey data. The selection process of 10 communities—5 from Lundazi and 5 from Mansa—where SILCs were implemented is the same as that of the MWH project (Lori et al., 2018; Scott et al., 2018). SILCs were first implemented in January 2016, and data were collected between October 2017 and February 2018 depending on how long the SILCs have been running. A convenience sample of 600 participants was sought from a total pool of 6711 participants from the 10 different communities SILC groups were implemented. Five of the communities resided are located in Lundazi (n = 297) and five in Mansa (n = 303).1 The local NGO’s programme evaluators met the groups on their monthly meeting dates. The description of the study was provided at the end of the regular SILC meetings, and the SILC members were asked to voluntarily participate in the survey. There were volunteers representing each of the 10 different communities. Volunteers for the survey provided verbal consent, and the survey was collected through in-person interviews in either English or the local dialect (e.g. Bemba, Nyanja and Tonga). The process was repeated for each SILC meeting until data reached 300 participants for each district. Inclusion criteria for participants were age 18 years or older and SILC group membership (must have participated for at least one cycle of committed timeline). Ethical approvals for the MWH project were obtained from the authors’ Institutional Review Board (IRB), as well as from the ERES Converge Research IRB, a private local ethics board in Zambia. The purpose of the SILC impact survey was to understand how loan and share-out funds from SILCs were used, how the funds affected the members’ livelihood and how SILC members perceived SILCs. The SILC impact survey included three domains: (1) demographics, (2) economic outcomes and (3) non-economic outcomes and financial preparedness for birth. The demographic domain included information such as participant’s age, sex, district of residence, month and year of the participant’s most recent childbirth (for male participants we asked for their wife’s/partner’s most recent childbirth), and the month and year when they joined the SILC. The economic domain included information on the amount of the first loan, usage of the loan and share-out funds, and engagement in agriculture, business and/or animal husbandry. Furthermore, data about the specific amounts of investments and gain from agricultural, business and animal husbandry before and after joining the SILC were gathered. The survey information regarding what materials comprised house and roofing structures before and after joining the SILC were also included. These questions were included in the economic domain because they are used to create a wealth index by many low-income countries’ Demographic and Health Surveys (Kolenikov and Angeles, 2009). The non-economic domain included variables such as the ability to pay for child school fees, uniforms and shoes; food security and the ability to purchase all the required supplies for the most recent delivery. The survey ends with open-ended questions asking for examples of how membership in the SILC has helped the participant or their family, whether they would recommend SILC to their family and why they would or would not recommend the SILC membership. Last, the financial preparedness for birth was assessed by asking the participants whether he/she was able to purchase all the required supplies—plastic sheet, gloves, baby hat, baby clothes, wrap and so on—for the most recent delivery. The participants who answered ‘yes’ to the question were categorized as financially prepared birth and those who answered ‘no’ were categorized as not financially prepared for birth. Because many people in low-income countries like Zambia often lack regular income, household wealth is frequently assessed by counting assets and assessing the quality of housing, sanitation facility and/or water supply (Kolenikov and Angeles, 2009). Similarly, to capture the impact of SILCs on household wealth, the ‘increase of wealth index’ variable was created using both the economic and non-economic variables. Using these variables from the economic and non-economic domains, a total of 13 new discrete indicators were created. Each indicator was compared across two time points—before and after joining SILCs. Post-SILC participation improvements were coded as ‘1’. No change or post-SILC participation decline/decrease were coded as ‘0’. The ‘increase of wealth index’ was then created by summing the 13 new indicators. According to the United States Agency for International Development’s guideline for housing conditions (2016), brick and cement were considered improved housing materials. Metal and cement were considered improved roof materials. If participants reported having these improved materials for housing and/or roofing after joining the SILCs, the two variables were coded as ‘1’. The reliability coefficient for the increase of wealth index was 0.86 (0.8 > α ≥ 0.7 = acceptable; 0.9 > α ≥ 0.8 = good and α ≥ 0.9 = excellent). To understand the impact of SILC participation on financial preparedness for birth, all SILC participants were divided into two groups: those who had (or their wife/partner had) their most recent childbirth before joining an SILC and those who had (or their wife/partner had) their most recent childbirth after joining an SILC. The sample was dichotomized by the most recent childbirth date and SILC initial join date to assess how income earned through SILCs influence financial preparedness for birth. The aim of this analysis was to describe SILC participation and household wealth for birth and to examine the association between (1) increase of wealth and financial preparedness for birth and (2) sex of the participants and financial preparedness for birth. Descriptive statistics were analysed with means and standard deviations (SD) provided for the overall sample as well as the stratified sample between those who were financially prepared for birth and those who were not. A set of chi-square tests of independence and two sample t-tests were conducted to examine the differences between participants who were financially prepared and participants who were not for the overall and stratified samples. The financially prepared sample was further stratified by sex. Means and SD were calculated for the overall and stratified samples from Lundazi and Mansa. Several binary logistic regression models were fit to assess the unadjusted and adjusted relationship between increased wealth index and financial preparedness for birth. Adjusted logistic regression models included age, sex, district of residence and the period of the most recent childbirth as covariates. All logistic regression models provided adjusted odds ratios (AORs) and 95% confidence intervals (95% CIs). To understand the relationship between sex and financial preparedness for birth, logistic regression models were fit between those who had their most recent childbirth before and after joining SILCs. The data were analysed using Stata 15.0 (StataCorp, College Station, TX, USA).