Financial education for HIV-vulnerable youth, orphans and vulnerable children: a systematic review of outcome evidence

Figures & data

Figure 1. Review process.

Figure 2. Evidence map.

Table

Parameters of Quality	Associated Questions	Weak (1)/ moderate (2)/ strong (3)/ not enough info (NA)	Notes for Narrative Write- up
Conceptual Rigour	1.Is the financial education built on a clear and coherent theory of change and financial goals (cogency- clear thread linking claims with conclusions in impact studies)?
	2.Either in program monitoring or study design is there a clear attempt to measure changes in outcomes?
	3.Is there evidence of behavioural impact or outcomes including how it was measured?
Methodological Rigour	4.Is the chosen methodology for the study well suited to the research question?
	5.Is there a clear experimental or quasi-experimental design? Is there internal validity? (randomization; minimization of risk errors; minimization of selection bias and other risks including incomplete data, sampling and controls) Can the results be applied to another setting? (external validity)
	6.Is the impact methodology robust? Is outcome assessment blinded? Have authors considered the study’s limitations and alternative interpretations of the analysis, consistent results across contexts? (cogency) Have authors ensured that results would be consistent across contexts and tools? (reliability)
Relevance	7.Does the financial education program demonstrate a tailoring or adaptation to the target audience? (form; language; background of trainers; other)
	8.Does the financial education program demonstrate design based on local market research?
	9.Is there a clear gender awareness built into design?
Pedagogical Rigour	10.Is there an appropriate scope and sequence for the curriculum?
	11.Does the curriculum build on the adult learning cycle and principles?
	12.Are the training delivery/outreach mechanisms (forum; venue; site) appropriate and accessible?
Total Score	Total possible score is 36 points for a rating out of 100%.

Financial education intervention impact studies and outcome effects on vulnerable youth and OVC.

	Context		Intervention		Outcomes
Intervention and impact evaluations	Target population	Geographical Context Outreach	Treatment arms and Controls	Length of impact follow-up following intervention	Outcome Effects + positive --negative = no effect	Odd ratio and Confidence Intervals (CI)/ F, Z outputs (reporting data of the results)
IPA, Finca Uganda (Jamison et al., 2014)	2,680 female youth	Mbarara district Uganda School, 15-hour course was delivered over ten weekly meetings	T1: Financial education T2: Access to group accounts (through FINCA) T3: Both C: No intervention	9 months	+ Financial education increased financial literacy -- Account-only treatment does not increase financial literacy -- Little evidence that account access and financial education are strong complements. Some evidence that they are substitutes.	Financial literacy index that combines the knowledge, awareness, and numeracy indices. Was found an increase of 0.039 standard deviations in the education-only treatment group (p-value = 0.078), an increase of 0.056 standard deviations in the education + account treatment group (p-value = 0.001), and no significant effect in the account-only treatment.
IPA, Aflatoun, Ghana (Berry et al., 2015)	5000 primary and junior high school students 6-14 years	Southern Ghana districts: 36 in Nkwanta, 30 in Greater Accra East, and 69 in Sekondi Takoradi Metropolitan Area. 135 primary and junior high schools	T1: Financial education and social education (personal exploration, child rights) plus locked money box with accompanied training T2: Only financial education plus locked money box with accompanied training C: No intervention	9 months	+ Both treatments had a positive effect on savings behavior, mostly because students shifted their savings from home to school in a voluntary after-school savings club with a locked money box --No changes in savings attitudes and financial literacy compared to control group + Few significant effects on risk preferences and spending --Financial education alone led children to work more, possibly shift from leisure or home production. However, school attendance did not change. No effect for integrated curriculum with money box. Also, differences are not statistically significant. --No treatment effects on child labor	HMB leading to an increase of 0.16 standard deviations (s.e. = 0.058), and Aflatoun producing a 0.12 standard deviation (s.e. = 0.053) increase. The difference between the Aflatoun and HMB program is not statistically significant (p = 0.48).
Suubi Project, Uganda (Ssewamala et al., 2009a)	131 HIV-orphaned adolescents The average age was 13.7 years at baseline (with a range from 11 to 17). Children were in their final year of primary school (which in Uganda goes up to grade 7), 57% girls	Rural Uganda Rakai District 15 primary schools	T: Family-focused financial ed.; peer mentorship on life options; targeted child savings account plus counseling on life skills, health education and school supplies C: Typical orphan care (counseling on life skills, school supplies)	10- month follow-up	+ adolescents’ self-rated health, sexual risk-taking behavior and mental health functioning. Health and mental health functioning were found to be positively associated with each other.	2.15 (1.05, 4.42) The treatment group has over twice the odds of rating their health as good or excellent than the control group (OR = 2.15). Limitations: savings data only collected from treatment group.
Suubi Project, Uganda (Ssewamala & Ismayilova, 2009b)	135 HIV-orphaned youths (ages 11-17), the average age was 13.7 years, 57% girls	Rakai district of Southern Uganda 15 rural primary schools	T: Family-focused financial ed.; peer mentorship on life options; targeted child savings account plus counseling on life skills, health education and school supplies C: Typical orphan care (counseling on life skills, school supplies)	10- month follow-up	+ Attitudes toward saving, + Attitudes toward academic performance +Attitudes toward educational aspirations +sexual risk-taking behavior and attitudes.	1) F(1, 257)=27.91, p<0.001, partial eta-squared (δη^2)=0.1 2) F(1, 188)=7.57, p<0.01, partial eta-squared (δη^2)=0.039 3) F(1, 225)=4.52, p<0.05, partial eta-squared (δη^2)=0.025 4) F(1, 254)=28.66, p<0.001, partial eta-squared (δη^2)=0.101
Suubi Project, Uganda (Ssewamala et al., 2010)	135 HIV-orphaned adolescents The average age was 13.7 years, 60% girls in the treatment	Rakai district of Southern Uganda 15 rural primary schools	Same as above	10- month follow-up	+ attitudes toward sexual risk for males = no change for females = females and males saved comparable amounts	F(1,266)=16.1, p<0.0.001, partial eta-squared (δη2)=0.6, 95% CI
Suubi Project, Uganda (Ssewamala et al., 2012)	138 HIV-orphaned adolescents, specifically in primary six or primary seven (7^th and 8^th grade in the US education system), 57% girls	Rakai district of Southern Uganda 15 primary schools	Same as above	10 and 20 month	+ significant decrease in depression compared to control group (no change)	Depression among HIV-orphaned youth intervention group had a statistically significant reduction over time (B=-.34, t(256)=-2.41, p = .02; 95% CI=(-.61,-.06)
Suubi Project, Uganda (Ssewamala et al., 2016)	AIDS-orphaned adolescents aged 12-16 years in the last 2 years of primary school. The average age was 13.4 years, 65% girls	Rakai and Masaka district of Uganda National Institution of Mental Health	T: Targeted child savings account plus counseling on life skills. Family level economic strengthening intervention in the form of Child Savings Account. Savings and educational workshops implemented in 10 rural public primary schools. C: Typical orphan care (counseling on school uniforms, school lunch, notebooks, and textbooks, bolstered with mentorship from a near-peer and life skills).	24 month follow-up	+ Better educational outcomes (children had much higher confidence in their educational plan) + Children reported having better physical health + More confident in achieving their educational plans	Children in the Suubi-Maka intervention group were 7.16 times more likely to take the Primary Leaving Examination (PLE) than children in the control group (OR =7.16, 95%CI: 2.03, 25.29) children in the treatment group had much higher confidence in their educational plans than did their counterparts in the control group (OR=6.06; 95 % CI=1.23, 29.91; p<.05). Also, children who, at baseline, reported having excellent physical health (OR=4.35; 95 % CI=1.54, 12.26; p<.01) or good physical health (OR=2.35; 95%CI=1.16, 4.8; p<.05) had higher odds of being more confident in achieving their educational plans than children who reported having poor physical health.
Suubi-Maki Project, Uganda (Karimli, 2013)	179 AIDS-orphaned adolescents (10-15 years old) and their parents	SouthwestUganda’s Rakai and Masaka Districts 10 primary schools	T: Family-focused financial ed.; peer mentorship on life options; targeted child savings account plus counseling on life skills, health education and school supplies C: Typical orphan care (counseling on life skills, school supplies)	12 and 24 months	+ Adolescent’s self-reported savings significantly affected (effect weakened by number of children in household) = savings attitudes not affected However, savings attitudes found to be significantly affected by family relations, family finance, caregivers’ gender, adolescent’s gender and educational aspirations.	Older adolescents were more likely to report having saved money (odds ratio=1.35; 95% CI = 1.19, 1.53), and they reported saving more (B=0.83; 95% CI=0.34, 1.31) than their younger counterparts. Female adolescents were less likely to report having saved money (odds ratio=0.48; 95% CI = 0.33, 0.7), and they reported saving less (B=-1.84; 95% CI=-3.01, -0.67) than their male counterparts. Similarly, adolescents with higher scores on the Beck Hopelessness Scale were less likely to report having saved money (odds ratio=0.93; 95% CI = 0.88, 0.99), and they reported saving less (B=-0.18; 95% CI=-0.3, -0.06) than adolescents with lower scores on the Beck Hopelessness Scale. Also, participants who planned to start secondary school saved less (B=-1.03; 95% CI=-1.6, -0.46) than adolescents who did not plan to start secondary school.
Suubi-Maka Project, Uganda (Karimli et al., 2014)	179 HIV-orphaned adolescents (10-15 years old) and their paerents. The average age of study participants was 13 years, 65% girls.	SouthwestUganda’s Rakai and Masaka Districts 10 primary schools	T: Family-focused financial ed.; peer mentorship on life options; targeted child savings account plus counseling on life skills, health education and school supplies C: Typical orphan care (counseling on life skills, school supplies)	12 and 24 months	+ Increased likelihood of family reporting having money saved = No significant effect on the reported amounts saved = No significant effect of the use of formal financial institutions	Children in the treatment group were more likely to report having money saved both at wave 2 (odds ratio = 2.72; 95 percent Confidence Interval [CI] = 1.4, 5.29, p < 0.01) and at wave 3 (odds ratio = 2.26; 95 percent CI = 1.12, 4.57, p < 0.05).
Suubi-Maka Project, Uganda Jennings et al., (2016)	346 orphaned adolescents, aged 10-17 years, on the last two years of primary school (equivalent to US 6th and 7th grades)	Rakai and Masaka districts	T: Family-focused financial ed.; peer mentorship on life options; targeted child savings account plus counseling on life skills, health education and school supplies C: Typical orphan care (counseling on life skills, school supplies)	12-24 moths	+ Significant relative increase over time in HIV-preventive attitudinal scores in youth Had effect. + Significant increase on the savings behavior of vulnerable youth (cash savings over time)	Significant relative increase over time in HIV-preventive attitudinal scores in youth (β= +0.19, ±0.09, p < .05), intervention had 2.017 significantly greater odds of a maximum HIV-prevention score (OR = 2.017, 95% CI= 1.43,2.84), most commonly toward perceived risk of HIV (95.8%, n = 159), sexual abstinence or postponement (91.6%, n = 152) and consistent condom use (93.4%, n = 144).   Had effect on the savings behavior of vulnerable youth; adolescents had an average significant increase of $US12.32 in CSA deposits per time period (β= 12.317, ±1.122, p < .001) and an average significant increase of $US15.00 in combined savings per time period (β = 14.996, ±1.657, p < .001), adjusting for changes over time in the control group.
Undaraga, Mongolia (Witte et al., 2015)	107 Female street-based sex workers at least 18 years. Average age 36 years. Most high school graduates, divorced, widowed, or separated.	Ulaanbaatar, Mongolia, In select districts	T: HIV sexual risk reduction training; matched savings; financial education (including negotiating); business training; mentoring C: HIV sexual risk reduction training Combined 4-session	At 3- and 6-month follow-up assessments	+ 31% decrease number paying sexual partners + 39% decrease unprotected sexual contacts with paying partners	Participants exhibited a 31% decrease in the number of paying sexual partners at each time point, participants in the Undarga Program group exhibited a 22% greater decrease than those in HIVSRR-alone group for each time point, participants in the Undarga Program group reported 50% fewer paying sexual partners at the 6-month time point, participants exhibited a 39% decrease in the number of unprotected vaginal sex acts with paying partners at each time point.
SHAZ!, Zimbabwe (Dunbar et al., 2014)	315 Adolescent female orphans, aged 16 to 19 was recruited through community out of school, not currently pregnant, HIV-uninfected	Chitungwiza, a high density, urban area outside of Harare, Zimbabwe In Community	T: Mixed life skills, health education, financial education, vocational training (to start income generating activity). Following intervention, micro-grants for business and social supports C: Mixed life skills and health education	2 year follow-up every six months after for 24 months	+ Reduced food insecurity, improved economic situation + Lower risk of transactional sex + Higher likelihood of using condoms + Fewer unintended pregnancies	Reduced food insecurity [IOR=0.83 vs. COR=0.68, p-0.02], and having their own income [IOR=2.05 vs. COR=1.67, p50.02], there was a lower risk of transactional sex [IOR=0.64, 95% CI (0.50, 0.83)], and a higher likelihood of using a condom with their current partner [IOR=1.79, 95% CI (1.23, 2.62)], there was also evidence of fewer unintended pregnancies among intervention participants [HR=0.61, 95% CI (0.37, 1.01)], over time compared to baseline
Biruh Tesfa, Ethiopia (Erulkar et al., 2013)	640 Girls in slum out-of-school, and aged 12–19.	Addis Ababa, Bahir Dar, and Gondar, Ethiopia Kebele (sub-district) community halls, mainly government-run facilities	T: Mixed health education (including HIV awareness training, counseling and testing), self- esteem, gender and power, basic literacy and financial education C: No intervention	30 months of implementation (2008-mid-2011)	+ More than twice likely to report social support + Twice as likely to score highly on HIV knowledge questions + know where to obtain voluntary counseling and testing + want to be tested	Girls in the intervention sites were more than twice as likely to report social support (OR) = 2.0) compared to girls in the control site. They were also twice as likely (OR = 1.9) to score highly on HIV knowledge questions, to know where to obtain voluntary counseling and testing (OR = 2.0) and to want to be tested (OR = 1.9)
Kishoree Kontha Project, Bangladesh (Scales et al., 2013)	200 (2009) and 534 (2010) adolescent girls aged 10–19 years	Rural Bangladesh villages Babuganj, Muladi, Patuakhali Sadar, Bauphal, and Bhola Sadar), in three coastal districts (Barisal, Patuakhali, and Bhola) of the South Central region Spare household spaces that served as learning centers for the project girls	T1: Social skills, literacy and school learning, health education. T2: Above plus financial education T3: Above plus nutrition incentive when marriage is delayed C: No intervention	6 and 9 month	+ Significant increase in external assets (support, empowerment, boundaries and expectations, constructive use of time) + Significant increase in internal assets (commitment to learning, positive values, social competencies, positive identity)	Combining the cohort data, these results suggest a mean improvement of 22% in the amount of developmental assets experienced, net of contamination and control group effects, and a mean net effect size of .80, conventionally considered a large effect
Reach, India (Spielberg et al., 2013)	2,227 Women members of women’s self-help groups (SHGs), adolescent daughters or daughters-in-law (10-19 years old) 547 (after 12 months)	Nadia West Bengal India 55 villages	T: Learning games with mixed health education (including sanitation, body, HIV protection) and financial education (including bargaining power) C: No intervention	6 and 12 month	+ Significant gains in HIV knowledge, awareness that condoms prevent HIV + Self-efficacy for HIV prevention + Confirmed use of clean needles -- Savings education showed no impact on financial knowledge or behavior change Limitation: Outcomes are not age or youth disaggregated	Girls had significantly higher motivation to save money (adjusted odds ratio 2.1, p = 0.002) and were more likely to have a plan to save money [adjusted odds ratio (OR) 2.2, p = 0.04]. Girls who received HIV education had significant increases in HIV knowledge: heard of HIV (OR = 3.6, p < 0.05), heard of sexually transmitted infections (STIs) (OR = 3.6, p < 0.05), and aware that condoms can prevent HIV (OR = 2.9, p < 0.05).
BRAC, Uganda (Bandiera et al., 2012, 2015)	4,800 Girls 14-20 years. Average age is 16 years. 70% enrolled in school.	Kampala, Mukono, Iganga and Jinja in Uganda Community clubs	T: Mixed health education (life skills to build knowledge and reduce risky behaviors), financial education (including negotiation skills and how to make use of financial services), vocational training (to start income generating activity) C: No intervention	2 and 4 years	+ Significantly improves HIV and pregnancy related knowledge + Significantly improves HIV and pregnancy related behaviors (sex against their will dropped 26%; early marriage and cohabitation fell by 58%; teen pregnancy fell by 26%) + Raised the likelihood of girls being engaged in income generating activities mainly driven by increased participation in self-employment + Raised likelihood of engaging in income-generation activity (2015) = No raised likelihood of engaging in income-generating activity (2012)	ITT estimates: girls in treated communities are 6.8pp more likely to engage in income-generating activities relative to girls in control communities, a 72% increase that is driven by additional engagement in self-employment. 90% more likely to be self-employed. Accompanied by monthly increases in consumption expenditures (38% from baseline) and significant reductions in job related anxieties. 7.2pp higher for treatment. Share of girls reporting sex against their will, early marriage and early childbearing drops