Implementation at scale: The role of subnational actors

• In Event: Reimagining the middle tier part I: Activating potential for implementing and scaling education innovations - Investigating roles and structures

Proposal

How can pilot programs be successfully scaled up in education systems in low-income countries? Recent research suggests that along with schools, national and district-level leadership and management are crucial to bring about sustained and large-scale improvements and reduction in disparities in educational outcomes (Antoninis et al., 2024; Mundy et al., 2024). This paper examines why the scale-up of a school-level program in Pakistan did not lead to improved school and student outcomes, despite a successful pilot. The findings focus on the role (or lack thereof) of subnational actor engagement in the large-scale implementation of the program.

The government of Punjab, with funding from the World Bank, started a pilot program called the School Council Mobilization Program (SCMP) in April 2013 in 5 out of 36 administrative districts of Punjab, Pakistan. The program used mobile-phone calls to provide sustained and targeted guidance to local school-council members (head-teacher, teachers, parents, members of the community) on their responsibilities and authority in managing the school (Author, 2022). The study found that the initiative led to meaningful increases in primary-school enrollment (i.e., a 4.0 percent increase), especially for young girls, and the improvements were sustained in the months after the program concluded. After policy advocacy, the pilot program was scaled to all districts in the province. However, a follow-up study found that, after scaling, there were no significant effects on school inputs such as enrollment and attendance or academic achievement in Math, Urdu, or English (Author, under review).

The paper employs descriptive data and informal conversations with implementing partners to account for why the scale-up of the program failed to have the desired effects. As a contrasting case, the Tusome program in Kenya is used as an example of a program that was designed and scaled up after successful randomized control trials to understand the SCMP. Unlike programs explicitly designed for scale from the start, the scale-up for the SCMP was not organic, and hence, not resilient to political and operational challenges (Angrist et al., 2025; Buccini et al., 2023; Piper et al., 2018). A program like Tusome succeeded because it was fully integrated into the government system from the beginning, used data for improvement, and sustained support from the partners. Compared to Tusome, the SCMP program was funded by the World Bank. Informal conversations with district staff revealed low levels of interest and motivation among the staff to oversee yet another donor-funded program. Additionally, there was no internal accountability among subnational actors, despite them monitoring schools on a regular basis. They were unfamiliar with the program objectives, which led to misalignment between what the program intended to achieve and the actors’ daily tasks. While donor-driven programs at a small scale can bypass challenges of implementation at subnational levels, it is more difficult to do at scale. This paper illustrates that scaling programs to improve outcomes for students requires practice and ongoing engagement with and collaboration across all levels of the education administrative hierarchy.

Author

• Minahil Asim, University of Ottawa