The Agency Fund

Every year millions of students with the skills and interest for advanced careers in STEM fall through the cracks of the education system because they don't see themselves as "the type of person" who can achieve in these domains. But there's promising evidence that role model narratives, communicated through film, can break through such self-limiting stereotypes and catalyze young people's potential. That’s why we're thrilled to have secured funding from Development Innovation Ventures at USAID to test a movie-based behavioral intervention in Brazil. With DIV's support, we'll evaluate the impact of a 30-minute inspirational video, in partnership with Brazil's Math Olympiad, an annual competition that reaches 18 million public school students. The intervention will target the top 5% of students in the competition. It aims to expose them to new possibilities, build their confidence in their own abilities, and help them sustain grit through the contest. If the video intervention proves effective, it could be scaled across Brazil at a minimal cost through the Math Olympiad organization. Excited to collaborate with DIV, IMPA Novo Film, Diana Moreira and Fernanda Estevan 🔗 Read more about the project here: https://lnkd.in/g77dvUbk Patricia Andrews Fearon Richard Sedlmayr Temina Madon

Every year millions of students with the skills and interest for advanced careers in STEM fall through the cracks of the education system because they don't see themselves as "the type of person" who can achieve in these domains. But there's promising evidence that role model narratives, communicated through film, can break through such self-limiting stereotypes and catalyze young people's potential. That’s why we're thrilled to have secured funding from Development Innovation Ventures at USAID to test a movie-based behavioral intervention in Brazil. With DIV's support, we'll evaluate the impact of a 30-minute inspirational video, in partnership with Brazil's Math Olympiad, an annual competition that reaches 18 million public school students. The intervention will target the top 5% of students in the competition. It aims to expose them to new possibilities, build their confidence in their own abilities, and help them sustain grit through the contest. If the video intervention proves effective, it could be scaled across Brazil at a minimal cost through the Math Olympiad organization. Excited to collaborate with DIV, IMPA Novo Film, Diana Moreira and Fernanda Estevan 🔗 Read more about the project here: https://lnkd.in/g77dvUbk Patricia Andrews Fearon Richard Sedlmayr Temina Madon

💡 Results from a new RCT on mentorship and goal-setting as a route out of poverty. This new working paper studies the effects of the Bridges to Success program – an intensive, wraparound anti-poverty program in Rochester, New York between 2017 and 2020. Study participants, all very poor, faced persistent barriers to employment and had recently experienced financial shocks at time of enrollment. Between 2017 and 2020, 430 participants across two cohorts worked closely with a mentor to set their own goals and create personalized action plans. They met at least monthly for up to two years, to set and achieve customized goals in nine domains, including family, health finances and employment. Navigators also supported participants with accessing community resources and government programs, and the program includes cash incentives for completing planned steps. Rather than aligning services with specific goals, the program focused on empowering participants to pursue their own chosen paths with guidance. A year on, participants’ employment rate was 10 percentage points higher than that of the control group and participants showed increased hopefulness, evidenced by the steps they were taking toward their goals. The program will break even if employment effects last around seven years, with after-tax earnings gains offsetting its costs. 🔗 Read the full paper: https://lnkd.in/eHV8EJsQ Study conducted by: Javier Espinosa, William N. Evans, David C. Phillips, Tim Spilde

Much of modern development economics is based on a model in which the intervention stays fixed. You build an intervention, tinker with it a bit, rigorously test to see if it works, then try to scale it. This can be thought of as a “nail it, then scale it” approach. As a model it has well understood virtues: It generates solid evidence of impact. But even when evidence-based interventions effectively traverse the challenges of scale, they can create brittle and inflexible solutions. At the Agency Fund, we invest in organizations that provide inputs like coaching support, where flexibility is key. In our view of development, the best organizations are constantly learning and adapting. They pilot new ideas, they sandbox, they A/B test, which means that their interventions are constantly evolving and improving. We find this approach incredibly exciting. We think it’s time to stop fixating on fidelity, and for organizations to build interventions that shed their skin. Read more from Richard Sedlmayr and Paul Wang on our blog: https://lnkd.in/dThq3V5i

⏰ Last day to register for this webinar on Frameworks for Creating User Funnels, hosted by our friends at Glific and facilitated by Robert On from The Agency Fund.

How can we rigorously measure mindsets at scale in settings with limited digital access? Most research showing the benefits of a growth mindset—the belief that intelligence is not fixed and can be developed—comes from high-resource settings where widespread internet access makes interventions easier to deliver and evaluate. But in resource-constrained contexts, different approaches may be needed to assess whether mindset interventions are achieving their intended outcomes. In early 2021, amid pandemic restrictions that ruled out face-to-face surveys, a team of researchers took on this challenge in São Paulo, Brazil. They developed a toolkit that uses SMS-based surveys and AI-powered processing to measure and classify student mindsets, and rolled it out among 92,000 public school students. The students could respond to survey questions directly on their phones, bypassing the need for internet connectivity. Then, to analyze these open-ended responses, the team developed and deployed a Natural Language Processing (NLP) model to classify answers by mindset type. They validated the method by comparing it with data gathered through face-to-face surveys in Chile. The evidence gathered in São Paulo shows similar patterns to findings from studies with more established research infrastructure, specifically in how student mindsets relate to socioeconomic factors. Students with a growth mindset exhibited weaker correlations between family income and academic performance, a trend also seen in the face-to-face mindset data gathered in Chile. This alignment suggests that, in this context, the SMS-based approach effectively captured insights into the relationship between growth mindset and educational outcomes. The team have now made all their datasets, dictionaries, and codes open access. Could this toolkit help strengthen the evidence base for your work? Read the full paper and share your thoughts in the comments: https://lnkd.in/gjxiWKsj Access the free version here: https://lnkd.in/g5k5RdzZ Study conducted by: Guilherme Lichand, Elliott Ash, Benjamin W. Arold Jairo Gudino, Carlos Alberto Doria, Ana Trindade, Eric Bettinger and David Yeager

🤝 How do we foster open source mindsets in social sector tech? At The Agency Fund, our strategy for building the development tech stack is rooted in openness and collaboration. When we accelerate promising deployments of LLMs, user engagement solutions or back-end automations in individual organizations, our ultimate goal is to share those innovations and catalyze progress across the entire sector. Sprints are an important part of this mission, an opportunity to bring developers together from different NGOs, to surface common challenges and co-create solutions. Hear more from Project Tech4Dev on how sprints and community meet ups inspire change⬇ Donald Lobo Vinod Rajasekaran

We often see agency framed as freedom—freedom from constraints. But this view misses how agency actually develops. For socially privileged people, it's easy to take a sense of agency for granted. If you've always had access to role models and opportunities, if people have always been willing to give you the benefit of the doubt and trust you to succeed, you may see your good choices—and your ability to make them—as privately produced. From this perspective, government intervention to support agency might seem unnecessary, or even like overreach. But many people face invisible deprivations that limit their ability to set goals, navigate towards them, or even to imagine a better future in the first place. For example, in a community where no one has moved beyond subsistence farming, it’s exceptionally difficult for any individual to envision alternatives for themselves without specific support. While agency is manifested in individual choices, it’s always situated in social context. This, we argue, is the paradox of agency. In a recent article, The Brookings Institution, James Walsh and Karla Hoff take on the Agency Paradox, arguing that while policymakers shouldn’t manipulate choices, they can and should intervene to expand people’s capacity to make choices and envision better futures. 📖 Read the full piece here and share your thoughts in the comments: https://lnkd.in/g5bEez9h

💡Our friends at Youth Impact have produced an excellent guide to using A/B testing for social sector impact! In the last few decades, growing use of RCTs has helped steer investment towards interventions with the greatest social return. But while RCTs establish broad evidence on what works, they often take years. Adding A/B testing to the toolbox allows for real-time, iterative refinements within program cycles and specific contexts. For example, as part of one of its remedial education programmes in Botswana, Youth Impact experimented with refining student groupings by adding an additional level based on digit recognition. This small tweak caused a 0.2 standard deviation in learning within a single school term, with only a slight increase in cost. This example highlights the potential of A/B testing to drive continuous, cost-effective improvements in program delivery, enhancing interventions as they scale. It’s an important growth area for social sector organizations and a major focus of our work with grantees. Have you been tapping into the potential of A/B testing for social impact? 🔗 Read the full review and share your thoughts in the comments: https://lnkd.in/gCdnXJ_W Authors: Noam Angrist, Amanda Beatty, Claire Cullen and Moitshepi Matsheng

📢Funding Opportunity in Women's Health Analysis of the global disease burden published this year suggests that, on average, women spend about nine years of their lives in poor health—25% more than men. The report, published by McKinsey & Company and the World Economic Forum, attributes the so-called gender health gap to inequities in care delivery, but also to disparities in science, data analysis, and R&D investment. Historically, the medicines and therapies used to treat women’s health concerns have not been based on adequate scientific understanding of women’s bodies or outcomes, with serious implications for many health issues, from cardiovascular disease to pain, and from psychiatric conditions to sports injuries. To help close the gap, Melinda French Gates and Pivotal Ventures have announced a $250 million global open call, which will fund organizations working to improve women’s mental and physical health. Applications are invited from organizations that: ➡ Serve women and have a record of improving women’s mental or physical health ➡ Center equity in their approach ➡ Are poised to scale their work to strengthen the health of more women. More details here: https://lnkd.in/gTqeZzvN Know any organizations that fit the bill? Tag them in the comments.