How IRC Turned Disappointing Results Into Program Innovation

This case study is based on the breakout session presentations by Barri Shorey of the International Rescue Committee (IRC) and Owen Ozier from the World Bank during Day 1 of Making Cents’ 2013 Youth Economic Opportunities Conference in Washington, DC.

Sometimes what looks like “failure” is an opportunity for rich learning. This is what happened during a project aimed at empowering adolescent girls through microfranchising.

The Girls Empowered by Microfranchising (GEM), implemented by the International Rescue Committee (IRC) and its partners, grew out of a small pilot project in Sierra Leone called YouthWORKS. In this project, IRC sought to provide self-employment opportunities for 100 youth (ages 15-24) by facilitating franchise business relationships with existing companies that had products or services that could be distributed and sold independently by the youth. The results of YouthWORKS sparked the interest of the Nike Foundation, who eventually provided funding for what became the GEM project, a similar program that focuses on adolescent girls—one of their prime markets—as part of the Nike The Girl Effect initiative.

Committed to learning from the start

The GEM project team and its donors were committed to pinpointing what aspects of the microfranchising model would work with young girls in Nairobi, Kenya. Researchers from Innovations for Poverty Action (IPA), The World Bank and Population Council all expressed interest in conducting rigorous evaluations alongside the project. Shortly after the beginning of the project, a research agenda was identified and designed. Overall, the GEM project provides girls aged 16-19 with a package of support that includes: basic life skills and business training conducted by local partners; franchise-specific training conducted by the local businesses that agreed to operate, or were already operating, a franchise model; mentoring; and IRC's startup support of business-related assets and initial stock supply. The World Bank, Population Council, and IPA worked together to establish a baseline, and then, using a randomized control trial (RCT) evaluation design, planned to conduct a midline and endline evaluation of Year One. Through a randomized selection of applicants, 244 girls composed the treatment group while 100 girls composed the control group. Ozier stated that the research was meant to answer two questions:  

1. Measuring "process outcomes"—Do participants actually go through the program as the implementers intended?

2. Measuring "economic outcomes"—Are participants working, earning, spending, and saving more, and are they more food secure?

Roadblocks ahead

Not far into Year One, the project was immediately confronted with a number of hurdles. For example, because the impact evaluation was designed at the outset of the program and not hand-in-hand, research partners did not initially establish a relationship with the local implementing partners who were the keepers of the data, so access to the information was initially very challenging. Tracking program applicants in an urban area was really difficult due to high mobility among youth, especially girls within this particularly age group, which was not accounted for. This also led to very low retention. Attrition was close to 50 percent for the first year of the program from girls who started training to girls who ended up starting businesses. To make matters worse, changes in the local market were unpredictable and sometimes detrimental to franchises in the program. In sum, Year One did not allow a large enough sample to be able to tell anything conclusively.  

Despite the inconclusive findings, the project team and the local business partners helping to enfranchise the girls felt that the learning generated was invaluable. On the positive side, two of the local businesses were making profits, along with the young business owners. While one business did drop out after Year One (because it was barely breaking even and the product was not particularly marketable in the new areas it was targeting), based on the benefits they observed from the program, the business did decide to expand the training programs it was providing to include more youth. From this perspective, a number of achievements were made that would not have happened without the intervention.

Learning from “failure”

Instead of trying to sweep their problems under the rug, the project team is energized to start fresh in Year Two. In addition to integrating cash transfers into its design, the team would like to test whether "flooding the market" with new franchises has effects, either positive or negative, on local business community. The data collection methods are now more streamlined (including application roll-out and not all at once), and while one business had to drop out of the program, it plans to expand its training program for adolescent girls in rural areas as a result of their experience. The other two businesses are making a profit and are enthusiastic about the project’s continuation. The GEM team is looking forward to sharing impact-level data from the project by the end of 2014 and expanding its reach of microfranchising for girls and boys in Kenya, engaging new businesses and new beneficiaries. 

CLA in Action articles are intended to paint a more detailed picture of what collaborating, learning, and adapting (CLA) looks like in practice. Unlike other disciplines, CLA is not a technical "fix;" it looks different in different contexts. This series will showcase examples of intentional collaboration, systematic learning, and resourced adaptation, some of which you may find applicable to your own work. The case studies, blogs, and resources represented in this series document the real-world experiences of development practitioners experimenting with these approaches for the benefit of sharing what's possible.