Tia Palermo's presentation on cash transfers and violence against women and children to UN Women's regional office and Promundo's Learning Dialogue Series in June 2020.
Cash Transfers and Violence Against Women and Children
1. CASH TRANSFERS &
VIOLENCE AGAINST WOMEN & CHILDREN
Tia Palermo, University at Buffalo
on behalf of the Cash Transfer & IPV Research Collaborative
UN Women Webinar: June 24, 2020
2. CASH TRANSFER & IPV RESEARCH COLLABORATIVE
https://www.ifpri.org/project/cash-transfer-and-intimate-partner-violence-research-collaborative
3. PRESENTATION
OVERVIEW
01. Review of the evidence
linking cash transfers &
intimate partner violence
(+ childhood violence)
02. Program design, M&E
insights, COVID-19
considerations
03. Discussion / Q&A
4. Social protection programs
Social services
Insurance schemes
Labor market policies
Social safety nets
Public works,
Subsidies, School feeding,
Economic transfers
Cash
Transfers
~45%
global
coverage
~2.5 billion
in LMICs
~11% coverage
in LMICs Sources: World Social Protection Report 2017-
2019: Universal social protection to achieve the
SDGs, ILO, Geneva, 2017; Ivaschenko et al.
2018. The state of social safety nets 2018
(English). Washington, D.C. : World Bank Group
6. Mixed method review of cash transfers on IPV
1) Review quantitative and qualitative
evidence linking cash transfers (CT) &
IPV, focusing on mechanisms
underlying impacts
2) Build a program theory linking CT and
IPV
3) Propose promising program design
features and research gaps needed to
further understand linkages/leverage
potential of CTs
Buller AM, Peterman A, Ranganathan M, Bleile A, Hidrobo M, Heise L. (2018). "A mixed-method review of cash transfers
and intimate partner violence in low-and middle-income countries." The World Bank Research Observer 33(2): 218-258.
7. Program characteristics (22 studies)
▪ Nearly all programs targeted women
(*Kenya, South Africa)
▪ 10 quantitative & 3 qualitative Govt run
Program type
▪ ‘Plus’: In-kind transfer; Education, health
sector linkages; trainings; behavior change
communication
▪ Multiple in Mexico, Ecuador, Peru, Uganda
0
2
5
7
9
11
UCT UCT plus CCT CCT plus
Quantitative Qualitative
Geographic Region
0
4
8
12
16
LAC SSA MENA Asia
Quantitative Qualitative
8. 0
4
8
12
16
20
Decrease Mixed Increase No
relationship
identified
Quantiative Qualitative
Majority of studies (73%) showed decreases, impacts stronger for physical/sexual
Overall 36% of quant indicators significant & negative; 2% significant & positive
0
10
20
30
40
Controlling/
psycological/
economic
Emotional Physical and/or
sexual
Decrease Increase Not significant
Overall impacts on IPV
[22 studies]
Quantitative indicators
[56 total by IPV type]
High-level results: Impacts
[11 of 14 studies
show any reduction]
10. IPV
PATHWAY 1: Economic security & emotional
wellbeing
“In my household it was like happiness, we all got along, with my
children, with my husband […] in my house we were happy […]
because before we did not have enough money for those things
[food].”
Female from IDI in cash, food & voucher program in Ecuador (Buller et al. 2016)
11. IPV
IPV
“There had been many fights. Because children needed many
things that we could not have afforded. I asked my husband
and he used to say there is no money. Then I used to get upset
and started to yell. We had many fights because of poverty.
Not only for us, for all poor, fights come from suffering”
PATHWAY 2: Intra-household conflict
Female from IDI in CCT plus in-kind transfers in Turkey (Yidrim et al. 2014)
12. IPV
IPV
PATHWAY 3: Women’s empowerment
“Earlier, … my husband would sometimes sell household items without
consulting me. But now that I have my own money, I can have a say
on how to spend income. ..With the money, a woman may buy seedlings
for planting, and hire an ox-plough or tractor or casual labor to dig for her.”
Female from IDI in UCT in Northern Uganda (Nuwakora 2014)
13.
14. cRCT, implemented by WFP,
evaluated by IFPRI
Provided cash or food transfers – with
or without intensive nutrition behavior
change communication (BCC) – from
2012 to 2014
Targeted to mothers of children 0-23
months in poor rural households
26% ↓ in physical IPV from
Transfers+BCC, 6-10 months after
program ended
Example from Asia: Transfer Modality Research
Initiative (TMRI) in Bangladesh
Roy, S., Hidrobo, M., Hoddinott, J., & Ahmed, A. (2019). Transfers, behavior change communication, and intimate partner violence: Postprogram
evidence from rural Bangladesh. Review of Economics and Statistics, 101(5), 865-877.
15. Why? Unintended benefit of BCC was sustained
increases in women’s social interaction and status
3 components of BCC
Weekly group meetings of women
with a trained community nutrition
worker
Bi-monthly home visits by
community nutrition workers
Monthly meetings with influential
community leaders
Material focused on infant and
young child feeding
Format led to ↑ in women’s
social interaction & status –led
to sustained ↓ in IPV
16. • Pantawid Pamilyang Pilipino Program (4P), a CCT Government program
operating in the Philippines since 2008 [targeted to poor households with
children under the age of 18 on the condition that they invest in children’s
health and education and participate in maternal health services] Evaluation
by GIL World Bank
• Laos Road Maintenance Program (cash-for-work) [targeted to women in poor
households, wages given monthly] Evaluation by GIL World Bank
• Maternal cash transfers & SBCC in Myanmar [UCT targeted to mothers in the
first 1000 days, SBCC includes information on nutrition & determinants]
Evaluation by Innovations for Poverty Action (Duke & University of Michigan)
• Bangladesh Maternity Allowance Program, a Government UCT + supplements
+ BCC [targeted to mothers in the first 1000 days] Evaluation by IFPRI (CT
& IPV collaborative)
Other ongoing studies in Asia
17. CTs & IPV: conclusions & research gaps
Cash transfers
can improve wellbeing beyond primary program objectives
are proven ‘structural’ prevention complements to dedicated GBV
prevention programming
Pathways include increased resources, emotional wellbeing, women’s
empowerment, reduced stress, social support
Evidence gaps:
Geographic & humanitarian settings
Sustainability of impacts
Program design: transfer recipient, plus components
Understanding mechanisms/pathways
Cost-effectiveness analysis to compare to dedicated GBV programming
18. What about violence against children?
Measurement more complex:
o Appropriate & specific violence measures vary
across age ranges
o Ethical issues more acute
Mechanisms more complex:
o Violence in different spaces (home, school, &
labor settings)
o Violence from different perpetrators (household
& strangers)
o More indirect pathways
Peterman A, Neijhoft N, Cook S, Palermo T. (2017). "Understanding the linkages between social safety nets and childhood violence: a
review of the evidence from low-and middle-income countries." Health policy and planning 32(7): 1049-1071.
19. Overall results: 57 indicators (11 studies)
In total, 19% represent
protective effects of
SSNs (no adverse
effects)
Regionally grouped –
sexual violence from
adolescent studies in
SSA, violent discipline
from LAC
Only one mixed-methods
study (Palestine)
[CELLRANGE]
[CELLRANGE]
[CELLRANGE]
[CELLRANGE]
[CELLRANGE]
[CELLRANGE]
[CELLRANGE]
Homicide
Dating violence
Physical against
minors
Violent discipline
Peer bullying
Sexual exploitation
Sexual abuse
20. Summary: SSN for childhood violence reduction
• Results are promising for protecting adolescent girls against sexual
exploitation & abuse, less so for other types of violence
• Many evidence gaps: regional, program type—lack of comprehensive
studies showing how SSNs can affect multiple types of risk for children
(including in different settings, witnessing IPV, frequency or severity of
violence)
• Mechanisms are diverse: schooling, caregiver stress, adolescent risk
behaviors, exposure to high-risk environments, girls’ empowerment.
• Lack of studies able to test plus components, including links to integrating
child and social protection systems.
22. What matters for program design:
Complementary activities
Complementary activities may be important for sustaining impacts –
however what form this takes will vary by context:
Edutainment or aspirational / empowerment messaging
Mental health interventions (i.e., Cognitive behavioral therapy)
Group or couples' trainings on gender norms / roles
Mentoring for girls / guardians
Case management for child/family protection
** Depends what drivers of GBV are in a setting / target group =
suggests pre-program gender risk assessment crucial
23. Intra-household relationships key: design features to allow women to
retain control (messaging, frequency, size of transfer) micro-changes
to gender roles in the household with “buy in” of men
Woman as transfer recipient appears important, but few tests of this
theory
Public works (cash for work) less effective in terms of overall
economic impacts & rarely increase gender empowerment, but are
politically attractive
Gender-sensitive approaches (assets developed, women’s participation in
decision-making, gender parity in wages, life-course exemptions)
What matters for program design:
Gender targeting & cash for work?
24. No studies to date that compare conditional vs. unconditional
programs – or removal of conditions – with respect to IPV [or
GBV more broadly]
Protective effects on IPV appear to be more common for CCTs
– however this may be in part due to regional differences & very
challenging to tell benefits of “conditions” vs. “complementary
programming”
More broadly, conditions are not needed to have beneficial
“family friendly” impacts
Some evidence conditions excludes most vulnerable & may
have detrimental effects
What matters for program design:
Conditionalities
25. Implementation matters
• Clear communication of program objectives & messaging
• Extra support & effort needed to ensure gender-sensitive
components are implemented as designed
• Need for clear communication on intention & motivation of
gender-sensitive design components from national to sub-
national level
• Existing research largely evaluates programs that are
gender blind/neutral; a lot of momentum now to understand
programs designed specifically for gender-transformative
effects
26. Considerations for design & implementation
during COVID-19
Brief & blog available on the
IFPRI website
• As of June 12, 195 countries have adapted or
introduced 1000+ social protection measures to
mitigate against the COVID-19
• Gender considerations are likely to be overlooked
in an urgent effort to save lives
Brief & blog summarized evidence on considerations
for gender-sensitivity in across five areas:
1. Adapting existing schemes & SP modality choice,
2. Targeting,
3. Benefit level and frequency,
4. Delivery mechanisms and operational features,
5. Complementary programming.
27. M&E considerations
• Why M&E?
• Assess to what extent meeting program objectives
• Understand participation among women
• Learn and modify programs
• M&E vs. impact evaluation (IE)
• IEs usually done by external experts; more complex evaluation methods
• M&E can not tell you what might happen in absence of program
• M&E may overestimate/underestimate program impacts due to time trends
(e.g., economic development, inflation, droughts, floods, COVID-19, etc.)
• IEs give causal estimates of impact (i.e., what would have happened had
program not been implemented)
• Impact evaluation more costly, takes more time
28. FAO Technical Guide: A toolkit on gender-
sensitive social protection
• Part of a 3-part guide, focused on
step-by-step development of an
M&E framework
• Define change you want to
measure
• Develop indicators to track
changes (samples in guide)
• Identify source of data &
frequency of data collection
• Describe how data will be
reported
http://www.fao.org/social-protection/resources/resources-
detail/en/c/1170231/
29. Violence
• Intimate partner violence
• Non-partner violence
• Help seeking behaviors
• Violence against children
• Risk factors (pathways)
• INSPIRE Indicator
framework
• KNOw VAWdata Asia-
Pacific
* Ethical considerations
must be accommodated!
M&E indicators
Intra-household
decision-making
• Decision-making power
• Agency & autonomy
• Locus of control
• Self-efficacy
• Voice (individual &
collective) & participation
• EMERGE (measures of
empowerment for gender
research)
• Measuring women’s
agency (Donald et al.
2020)
Division of resources
• Labor force participation &
earnings
• Savings & financial
inclusion
• Asset ownership / bundle of
rights
• Consumption &
expenditure (individual)
• Food security & dietary
diversity (individual)
• Time use (incl. care work)
• Women’s economic
collectives measurement
guide
• J-PAL measurement guides
30. Cash Transfer & IPV Research Collaborative (2019-2021)
Goal 1: Expand evidence base, focusing on geographic diversity, mixed method
work (short-term)
Bangladesh: Maternity Allowance Program RCT
Ethiopia: PSNP plus qualitative
Ghana: LEAP 1000 qualitative follow-up
Goal 2: Communication, dissemination & research coordination
Policy & communication products
Presentations to diverse audiences
Technical assistance & coordination
Goal 3: Develop longer-term studies exploring research gaps (starting 2021)
Currently in development phase
31. THANK YOU!
Email: tiapaler@buffalo.edu
@tiapalermo | @TransferProjct
transfer.cpc.unc.edu
Cash Transfer and IPV Research Collaborative
https://www.ifpri.org/project/cash-transfer-and-
intimate-partner-violence-research-collaborative
Can CT programs be a globally relevant and scalable platform for sustainably reducing IPV?
Cash transfers including social pensions, family and child benefits, universal basic income, can be mobile or manual payments – these programs are a growing share of SSNs globally (particularly in SSA and LMIC regions) and they are pro-poor, so they reach a larger share of the poorest populations in countries. Rise of cash based on wide spread impacts, cost effectiveness and ability to give dignity/choice to participants -- Increasingly used in humanitarian and fragile settings.
Inclusion criteria (14 quantitative & 8 qualitative):
Published? Quant (7 published, 8 technical reports/working papers), Qual (2 published, 7 working papers/technical reports)
2/3 of studies were CCTs & CCT plus
A little over half were govt
2/3 in LAC
If not a lump sum transfer, most programs delivered benefits on a monthly basis, most were means targeted (poverty or demographic) and ranged from 6-50% of pre-program consumption exp.
4 qualitative & 5 qualitative studies supported this pathway
Linear linkages well supported by rigorous literature
HH level mechanism evolving from a pure income effect to emotional wellbeing
Talking points:
This pathway operates at household level, regardless of who receives the money
CT and decreased poverty:
Large and robust literature across regions and programme typologies showing that CT’s have significant positive impacts on a range of hh level economic security outcomes
Also, growing literature on documenting positive local economy impacts of CT’s (implying spill-overs
Program design and implementation characteristics such as relative size of transfer, regularity and duration of benefits is important
Economic security and emotional wellbeing
Increasing evidence that poverty and poor mental health are linked in a two-way
Emotional wellbeing and IPV
Evidence suggests that the relationship between poor mental health and IPV victimisation is bidirectional
Situational stress and IPV have also been documented
Economic security, alcohol abuse and IPV
Robust body of evidence from LMIC’s show a strong and consistent association between men’s use of alcohol and women’s risk of IPV
0 quantitative and 4 qualitative supported pathway
Linear linkages supported by reviews and select studies
Effect of cash on marital dynamics & conflicts
Less evidence for impacts on temptation goods
Talking points
This pathway seems to be particularly relevant in poorest households
If money used for temptation goods then more conflict and IPV could result, however there is very little evidence for the use of CT to buy temptation goods
More cash can reduce marital conflict over money or it can increase conflict if the money is diverted to temptation goods or partners disagree on how the money is spent.
CT helped to pay for a number of items such as school fees, medical bills or immediate needs
LESS evidence for people using the money in temptation goods
11 quantitative and 4 qualitative supported pathway
Large body of literature with mixed/inconclusive findings
Operates at the individual level
Talking points
Also mixed results
TMRI was a 2-year pilot safety net program in rural Bangladesh implemented by WFP. It was designed as a cluster randomized controlled trial that tested the impacts of food or cash transfers provided to poor rural women, with or without nutrition behavior change communication (BCC). It is worth noting that the study had no explicit gender or violence-prevention objectives. The focus of the programming was to improve household food security and child nutrition.
Why then did the combination of transfers + nutrition BCC lead to sustained reductions in IPV, when transfers alone did not? Evidence suggests that, although the BCC focused on changing women’s behavior related to infant and young child feeding practices, it ended up more broadly transformative for women. The BCC had three components – 1) weekly group meetings of 9-15 women with a trained nutrition worker in their village, 2) twice-a-month home visits from the nutrition worker, 3) monthly meetings of program staff with community leaders. Given female seclusion norms in this context, prior to the intervention, beneficiary women were quite isolated and had limited social ties; they were also viewed as beggars and pariahs, due to their poverty. The regular meetings with other women and the increase in their knowledge that was valued in the community led to unintended increases in women’s social interactions and status in their communities – which were likely to have been sustained even after the transfers and BCC programming ended. The increases in social interaction and status in turn increased women’s bargaining power within their relationships – so they could assert not accepting violence – and also increased their visibility in the community – so their husbands faced more disapproval for inflicting violence.
GIL=Gender innovation lab
“control” of: (messaging, frequency, size of transfer)
Percentages represent the % of indicators which are significant
two new papers (Mali and Bangladesh that show in both cases where there are IPV effects, they also resulted in VAC decreases
All of these may also have Indirect increases in social network
Gender sensitivity in payments: simplified, low cost accounts; flexibility in documents required to open; services close to home; adapt to low literacy levels
General recommendations to make PWPs more gender-sensitive include the following(20):
institutionalization of quotas for women’s participation;
recognition of life-cycle needs including no work requirement for women 6 or more months pregnant and for women nursing an infant younger than 10 months;
offering women flexible work hours;
targeting female-headed households;
guaranteeing equal wages for men and women; opening supervisory positions for women;
involving women in program planning;
supporting women’s access to savings and credit;
use PWP to create infrastructure that eases women’s time burden
link participants to complementary services: eg, literacy training for adult women, services for domestic abuse, etc.
More information here: Holmes, R., & Jones, N. (2013). Gender and social protection in the developing world: beyond mothers and safety nets. Zed Books Ltd.
Often, implementation fails when sub-national implementers don’t understand motivation for gender-sensitive components, such as household-level exemption of work in PWP when woman pregnant, etc.
Note that M&E indicators cannot give you the same information as an impact evaluation. The latter examines trends in both groups receiving a program and those not receiving a program and provides causal estimates of the “impact” of a program. However, M&E data are usually only collected among those participating in a program (and thus can’t tell you what would have happened in the absence of the program – i.e., the “counterfactual”). For examples, some limitations to the use of M&E data among program participants only are that these data don’t control for general trends where the programme is being implemented. For example, if a program is implemented at the same time as a country is experiencing rapid economic development, you may overestimate program impacts and attribute improvements due to general trends in economic development to your program. Similarly, if there are negative shocks, such as COVID-19, floods, droughts, rapid inflation, etc., you may underestimate program impacts; participants in the program may appear worse off than pre-program, but compared to a group not receiving the program, they may still be better off, and thus your programme had protective effects. Nevertheless, M&E data can provide important and useful information, often in more “real-time” than an impact evaluation.
Bundle of rights refers to access, use, ability to sell, bequeath, decisionmaking, management, ownership – the bundle of different rights over assets