Evaluating the Impact of Conditional Cash Transfers on Children, Families and Communities in South A

1
Evaluating the Impact of Conditional Cash
Transfers on Children, Families and Communities
in South Africa

• J. Lawrence Aber
• Professor of Applied Psychology and
• Public Policy
• New York University
• February 13, 2007

2
Current Focus of Empirical and Theoretical Work

• Place-Randomized Trials Experimental Tests of
  Public Policy
• Boruch, R. (Editor). The Annals of the American
  Academy of Political and Social Science, Volume
  599, May 2005.

3
Guiding Assumption

• one must change entire places to enhance the
  well-being of people in those places or to
  decrease the risks that confront them.

4
Place-Based Trials

• Education/School Trials to prevent problem
  behaviors and promote learning.
• Welfare/Community Trials to evaluate antipoverty
  strategies.
• Health/Public Housing Trials to prevent HIV
• Criminology/Hot Spots Policing Experiments to
  prevent crime

5
Current Work

• On
• School-Randomized Trials of
• 4Rs Reading, Writing, Respect and Resolution
  (Social and Character Development)
• First Things First (Whole School Reform)
• Community-Randomized Trials of
• Conditional Cash Transfers
• Other Antipoverty Strategies

• In
• New York City and Nationally
• KZN, South Africa (and New York City?)

6
Going to Scale

• The Cost-Effectiveness of Alternative
  Interventions to Support Vulnerable Children and
  Families in the Context of Poverty and HIV/AIDS.

7
Demographic Context

• HIV/AIDS Prevalence
• Poverty Income Distribution
• Urbanicity
• Race Tribal Language

8
Policy and Services Contexts

• Constitution
• Child Support Grant
• Other Means Tested Supports
• Social, Health and Economic Policies and
  Practices
• - ART, VCT
• - Teacher/Child Ratios
• - Labor

9
Emergent Partners

• In South Africa
• Treasury, Social Development, Presidency
• also, Health, Education
• Human Sciences Research Council
• Internationally
• World Bank, DFID
• also, RBF, Ford, Unicef (NIH? Gates?)
• New York University

10
How Poverty and HIV/AIDS come to affect
Childrens Outcomes?

• THE MICRO-LEVEL MODEL
• Major Household Risk Factors
• Adverse Childhood Experiences

11
Figure 1. Micro-Level Model of Influence of
Parentsa HIV/AIDS Status on Childrens
Psychosocial and Other Outcomes (adapted from
Foster Williamson, 2000)
Death of Childrens Parent(s)

• Moderators
• Community School level
• Family Household Level
• Child Relationship Level

a1
c2
b1
a4
a3
Child Care Status

• Parent(s) with
• Increasingly Serious
• Physical Illnesses /or
• Psychosocial Problems

Parent(s) with HIV/AIDS
Key Child Outcomes
c3
b4
a2
c1
a5
Adverse Childhood Experiences
Households In Economic Distress
Deepening Household Economic Problems
b2
c4
aParents or, in the case of orphans living
outside of their families, parent surrogates
(e.g., grandmother, foster parent).
12
Figure 2. Specific Statuses, Experiences and
Outcomes
Child Care Status
Adverse Childhood Experiences
Key Child Outcomes

• become caregiver
• householding
• fostered out

• withdraw from school
• inadequate food
• problems with shelter/
• material needs
• reduce access to
• health services
• discrimination
• exploitative labor
• sexual exploitation
• life on the streets

• educational
• health
• psychosocial

13
How do Community and School Factors Influence
Households and Children affected by Poverty and
HIV/AIDS?

• THE MACRO-LEVEL MODEL
• Resources and Supports
• Norms
• Collective Efficacy

14
Figure 3. Macro-level Model of the Influence of
Policy- and School-Community Factors on Household
Risk Factors, Adverse Childhood Experiences and
Childrens Outcomes. (Reduced from Model)

• School-Community
• Resources Support
• Norms
• Collective Efficacy

Policy Innovation
f2
d
f1
f3
e1
Major Household Risk Factors
e2
Adverse Childhood Experiences
Key Child Outcomes
15
Possible Policy Strategies

• Wider access to Child Support Grant
• Control Group
• Unconditional Cash Transfer (UCT)
• Intervention Group A
• Conditional Cash Transfer (CCT)
• Intervention Group B
• Home-based Visiting (HBV)
• Intervention Group C
• Others Open

16
should target the two factors that research
shows are most affected by parent illness and
death

• household economic problems
• and/or
• family support

17
Key Question

• If Communities Implement UCT, CCT or HBV, will
• households experience fewer economic problems?
• children attend school more often?
• children experience better educational, health
  and psychosocial outcomes?

18
Key Features of Evaluation

• Design
• cross-sequential
• longitudinal
• Samples
• school communities
• children
• Measures
• implementation
• process
• outcomes

19
(No Transcript)
20
Cross-Sequential Design
Age in Time/Year 1 (2008) 2 (2010) 3 (2012)
7 9 11
9 11 13
Two Age Cohorts Younger Older
Policy Change
21
Samples

• Number of School Communities
• 120
• Number of Children per School Community
• 50 younger 50 older 100
• TOTAL NUMBER OF CHILDREN
• 12,000

22
The Rigorous Test

• New Policy School Communities
• UCT (A) 30
• CCT (B) 30
• HBV (C) 30
• Current Policy School Communities
• Expanded CSG 30

23
Statistical Power in Place-Randomized Trials

• J (Number of places)
• Effect Size
• Intra-Class Correlations
• Measures of Place theory, sensitivity to change

24
Can test whether UCT, CCT or HBV

• are implemented well
• reduce household risk factors
• reduce adverse childhood experiences
• improve childrens psychosocial, educational and
  health outcomes

25
Challenges

• Scientific
• Logistical

• Political
• Economic
• Ethical

26
Benefits

• Immediate and ongoing information for monitoring,
  implementation analysis and cost analysis.
• A clear answer to the main question about whether
  this approach to going to scale works (if
  designed right).
• Valuable information, not only on what works
  but also for whom and where, to guide policy
  improvement.

27
Benefits (Continued)

• Evidence to sustain momentum over transitions in
  governments and funders.
• Becoming the vanguard in the evidence-based
  policy movement in addressing the needs of
  communities, households and children affected by
  poverty and HIV/AIDS.

28
Movement toward Evidence-Based Policy

• International
• The Cochrane Collaboration
• Health
• The Campbell Collaboration
• Social, Psychological, Educational and
  Criminological
• World Bank, OECD, etc.

• United States
• What Works Clearing-House
• http//whatworks.ed.gov
• The Coalition for Evidence-Based Policy
• http//www.excelgov.org

29
(No Transcript)