There is no “I” in “IE”: A Ganar Impact Evaluation

Karen Towers USAID Education Team Leader LAC Bureau

October 7, 2014

Security Issues/Target Groups

A Ganar Program: $8.9 million, 8 USAID countries

•Employment •Return to School •Start Own Business

• Alternative to Gangs/Drugs

Program Expansion Allowed for IE

Guatemala

• Honduras

Timeline for Evaluation

October 2014 3 year mid-point October 2016

End of evaluation

October 2011 Start of the evaluation

Research Questions

• Question 1 (Proof of Concept): To what extent does participation and completion of the A Ganar program increase the likelihood that youth will obtain and maintain jobs, return to school, start their own business or reduce risky behavior?

• Question 2 (Use of Sports): Does the use of sports in A Ganar increase the retention rate, job insertion rate and effectiveness of the program to teach life skills, language, math, IT and other complimentary activities?

Why Experimental?

• • • • •

Evidence Base Lack of Longitundinal How can program improve? Can it be replicated? Given scarce resources, what are the models that are working? For which groups? Most cost-effective? • Other donors

Key Stakeholders / Coordination

IO

IO

IO

IO

CO / Guatemala

IO

IO

CO / Honduras

Partners of the Americas

FOG Fund. Paiz

USAID

USAID

Guatemala

Headquarters

USAID Honduras

Social Impact

CIE N

ESA

Coordination

Lessons Learned

• • • • • • • •

Build in evaluation from the design Involve IP as early as possible Communication (key people constant; IOs; feedback) Time commitment to manage Be flexible ($ in POA budget) Donor role as referee Ethics of randomization (3 spots for every 25) Working in violent communities

How will results be used?

• Baseline helpful for Targeting • Informing Policy and Programming – Effectiveness of WFD for long-term outcomes – Sports as an effective development vehicle

• National Authority & Country Dialogue

“To Win, To Earn”

“To Win, To Earn”

Modifications •Selection Process

•Hours of Instruction “”

Using the Findings

•Improved Targeting •Program Enhancements •Credibility/Evidence of Success • Workforce Development • Sport “”

Lessons Learned

•Stakeholder Involvement •Importance of Understanding • Evaluators – Program • Stakeholders - Evaluation •Open Dialogue & Willingness to Negotiate “”

A GANAR ALLIANCE IMPACT EVALUATION

Mateusz Pucilowski Deputy Director - Impact Evaluation Social Impact

Part 1: Evaluation Overview

18

Evaluation Design Research Question 1: Proof of Concept “What is the impact of A Ganar participation on…”

Final Outcomes: • Employment • Entrepreneurship • Re-Entry into Formal Education Intermediate Outcomes: • Life Skills • Employment Skills Secondary Outcomes: • Risk Behavior • Gender Norms

Evaluation Design

Research Question 2: Role of Sport

Does inclusion of a sport constitute a ‘value-add’ ?

How can we isolate the effect of sport?

Indice de empleo

Impact of sport Impact of A Ganar

21

Mixed-Methods RCT QUANT APPROACH 2012 2013 2014 2015 2016 • 3,000+ youth surveyed cohorts) 6 7 8 9 10 11 12at 1 2 three 3 4 5 6 7 8 points 9 10 11 12 1 (across 2 3 4 5 6 7 8 99 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 Cohort 1

X

– Baseline Honduras Cohort 2 Cohort 3 – Post-program follow-up Ceiba (Pilot) Child Int 1 – End-line Fundacion Paiz 1 (Pilot) Fundacion Paiz 2 Child Int 2 Fundacion Paiz 3

X X

(t = 0 months) x X (t = 9 months) X X X (t = 27 months)

? X

X

? ?

X • Randomized assignment blocked by implementing organization and sex X ?

Guatemala

X

?

QUAL APPROACH • Case Studies In-depth interviews with 5 individuals over multiple points in time (participant, their life skills and vocational training facilitators, mentor, and employer).

• Exit Interviews Short, open-ended interviews about perceptions of the importance of the sports component in relation to their overall experience in the program

22

Part 2: Evaluation Modifications

23

Proposed Evaluation Design

Question Question 12 –– Proof Role of ofSport Concept 24

Two Country Design

Guatemala City

Tegucigalpa

25

Research Design – HONDURAS

26

Research Design - GUATEMALA

Proposed Sampling Design

Treatment Eligible Applicants Program Applicants

Control

Ineligible Applicants 28

Revised Sampling Design

Treatment

Eligible Applicants

Eligible Applicants

Program Applicants

Control

Ineligible Applicants

Direct Participants

29

Part 3: Lessons Learned

30

Research in high-threat environments • Importance of communication • Unannounced spot-checks • Interview location (household vs. public)

• Tracking youth

31

Keys to successful RCT implementation • Invest in relationships • Understand stakeholder incentives • Build capacity and mutual understanding

• Be flexible • Be inflexible! • Have a strong, communicative arbitrator

32

Questions?