Can social protection simultaneously reduce violence against children and violence against women?
Despite the identification of cash transfers as a promising evidence-informed strategy to address violence against children (VAC) and violence against women (VAW) – until recently, there was little evidence from low- and middle-income country settings (LMICs), that assessed the effects of such interventions on both in the same study. Yet, cash transfers and broader forms of social protection have potential to address both forms of violence through shared risk factors, including reductions in poverty and economic stress, or through complementary ‘plus’ programming targeting risk factors related to conflict in the home or violence norms (see reviews on VAC and IPV).
In this blog, we summarize five studies completed in the last two years that examine impacts of cash, cash-for-work and cash plus programmes on both violent discipline of children and male intimate partner violence against women (IPV) from diverse LMICs – Bangladesh, Colombia, Mali, the Philippines and Rwanda. We include both studies evaluating the effects of cash alone, cash plus programming, as well as those that evaluate just the impacts of the ‘plus’. We focus on studies that were rigorously designed and measured violent discipline and IPV in the same household (rather than those focused on violence against adolescent girls, which might fall in the nexus of both categories of violence). To our knowledge, with the exception of a previous study in Mexico (from 2013), these 5 studies are the only available publications from LMICs examining both types of violence in the same evaluation. We also offer key take away messages and suggest areas for future research.
- Bangladesh: An experimental study examined post-intervention effects of the Transfer Modality Research Initiative pilot, implemented in rural areas over 24 months by the World Food Programme. The intervention provided both transfers (cash and food), as well as a group-based nutrition behavior change (BCC) intervention to women with young children living in poor households. The evaluation found reductions in physical IPV of 26% among women in the cash plus BCC arm, however no impacts in the transfer only arm (and no impacts on emotional IPV). In addition, the authors examine two indicators of physical violent discipline from parents taken from the HOME inventory (whether mothers had hit the child during the week prior to the study and if parents react with physical discipline if they are hit by the child). The study found reductions of 25% to 38% across violent discipline indicators, again in the cash plus BCC arm (8 to 12 percentage points [pp]- reported in the online appendix). The author’s examination of mechanisms for IPV suggests that reductions in poverty-related stress, and increases in household economic status, which were larger in the BCC arm, may be a possible joint pathway for reductions in both violence measures (Roy et al. 2019 in the Review of Economics and Statistics).
- Colombia: A quasi-experimental study of the government’s conditional cash transfer program targeted to poor households with school-aged children, Familias en Acción, used variation in the timing of bi-monthly payments at the municipality level paired with municipality-level administrative data on reported levels of violence from health and legal services. The authors show that rates of overall domestic violence, as well as rates of IPV from administrative data reported to health and justice systems, decrease by 6% in payment months. In contrast, there are no changes in reported domestic violence specifically against minors. Authors also show that household spending is higher in payment months—suggesting a poverty and stress reduction mechanism achieved via higher consumption expenditures (Camacho & Rodriguez 2020 in the CEDE Working Paper series)
- Mali: An experimental study of the government’s Jigisémèjiri program, an unconditional quarterly cash transfer given primarily to male heads of household found decreases in IPV after 24-months. These decreases were concentrated in polygamous households (making up 40% of the sample), where reductions were found for controlling behaviors (23% or 16 pp), emotional IPV (37% or 13 pp) and physical IPV (40% or 7 pp). The study also reported on VAC among a target child aged 2 to 4 years old using the UNICEF Multiple Indicator Cluster Survey measures – showing similar trends as for IPV. Measures of psychological aggression against children were insignificant in the full sample, however physical punishment and number of total acts showed reductions—which increased in magnitude and significance among the polygamous households (showing decreases in psychological aggression of 16% or 11 pps and in physical punishment of 20% or 17 pps). Key mechanisms underlying impacts were reductions in men’s stress and anxiety, as well as in reported household disputes (Heath et al. 2020 in the Journal of Development Economics).
- The Philippines: An experimental study of a locally-adapted 12-session group-based parenting program (Masayang Pamilya Para Sa Batang Pilipino – or MaPa) layered on the government flagship conditional cash transfer (Pantawid Pamilya Pilipino Programme) reported impacts at program end and 12-months post program. The evaluated intervention reached female caregivers of children aged 2 to 6 years in low-income families in urban Manila. The evaluation found reductions in incidence and frequency of child maltreatment at both follow-up waves, measuring using the ISPCAN Child Abuse Screening tool (e.g. a 49% reduced risk of physical abuse at post-intervention and a 48% reduced risk of neglect). For IPV, risk reductions at program end were 63% and at 12 months post-intervention were 49%. Possible common mechanisms of impact were those reducing overall incidence of family conflict and stress, increased caregiver efficacy and confidence when dealing with male spouses, fewer daily child behavior problems, and lower parenting dysfunction, among others (Lachman et al. 2021 in The Lancet Regional Health – Western Pacific).
- Rwanda: An experimental study examined the impacts of the Sugira Muryango program – a home-visiting-based parenting intervention – layered on a government flagship social protection program ‘Vision 2020 Umurenge Programme’ targeting poor households with direct cash support and public works. Sugira Muryango included 12 sessions delivered over 3 months by community-based coaches promoting early childhood development (ECD) and preventing family violence. The study examined outcomes at 12-months post-intervention, showing reductions in female caregiver reports of IPV experience (IRR=0.616, 95% CI 0.458 to 0.828) as well as VAC as measured by harsh parenting using the UNICEF Multiple Indicator Cluster Survey measures (IRR=0.741, 95% CI 0.657 to 0.835). However, no significant impact was shown on male caregiver report of IPV perpetration (among an alternate sample of households where male caregivers were surveyed). The program also showed significant increases in engaging fathers in childcare and select child development outcomes (Jensen et al. 2021 in BMJ Global Health).
Take away messages and future studies
Taken together, these studies suggest a similar pattern of impacts. Across all but one study (Colombia), where impacts are found on IPV—the similar direction of impacts are found for VAC. In Bangladesh, impacts are only observed with the addition of plus components. However, in Rwanda and the Philippines—the evaluation focuses exclusively on the plus intervention—thus we are unable to fully disentangle if there are synergistic or countervailing impacts of the economic component alone. In three cases (Bangladesh, the Philippines and Rwanda) the evaluation includes post-intervention effects, showing that reductions in both IPV and VAC are sustained even after the program ends. The lack of impacts for VAC in Colombia may be due to the use of administrative data on cases of violence reported to health and justice systems (which capture only a fraction of violence prevalence), or the identification strategy (relying on the timing of payments) may not be meaningful enough variation for impacts. In spite of differences in social protection strategies and methodologies used to measure impact, overall, these results show that social protection is a promising intervention and platform to reduce both violence against children and violence against women.
Moving forward, more studies are needed that explore the effectiveness of social protection on multiple dimensions of violence. In doing so, evaluations will need to take a more holistic approach to map out pathways of impact and measure violence. For example, to affect IPV, social protection evaluations often target and focus on women alone and seek to empower her and change her circumstances – however to fundamentally change violence inside the home and parenting practices tied to violent discipline, it is essential to involve and collect data from men as well. In addition, more evidence is needed on possible intergenerational effects – another key point of intersection between VAC and VAW — for example, linking benefits realized by adolescent girls in households receiving social protection benefits to stability and freedom from violence in future intimate relationships. From a methodological standpoint, this research agenda is ripe for inter-disciplinary collaboration between development economists who typically evaluate social protection programming, and public health experts on VAC and VAW.
Stay tuned for more work from UNICEF Innocenti and partners on the intersection of VAC and VAW, including systematic reviews (on effective interventions and shared risk factors) and results from Mozambique’s Child Grant evaluation measuring impacts on violent discipline and IPV.
Authors: Amber Peterman is a Research Associate Professor at UNC where she co-leads the Cash Transfer and Intimate Partner Violence Research Collaborative and consultant to UNICEF Innocenti, Alessandra Guedes is the Gender & Development Research Manager at UNICEF Innocenti.
The authors would like to thank Elena Camilletti and Zahrah Nesbitt-Ahmed for helpful comments.