Association for Public Policy Analysis & Management

This paper investigates the impact of strengthening local governance of schools, by engaging school actors using Information and Communication Technology (ICT), on school performance in a developing country context. Evidence on the effectiveness of local committees in improving school performance is mixed at best. For example, some studies highlight that giving financial and operational autonomy to school-level actors has a positive impact on process outcomes, like improved community involvement in schools, but has no impact on student achievement (Blimpo, Evans, & Lahiri, 2011; Beasly & Huillery 2016). Other studies show positive effects of engagement on test scores (Duflo, Dupas, &Kremer, 2015), while some studies show no effect on intermediate or longer-term outcomes (Banerjee, Banerji, Duflo, Glennerster, & Khemani, 2010; Leer, 2016).

I study a unique program in Pakistan, the School Council Mobilization Program (SCMP), in which the provincial government provided information to school councils (SCs), comprising of teachers, parents and community members, on their roles and responsibilities in school management via sustained and targeted phone calls. I study the scaled up SCMP in 21 districts (approximately 22,000 primary and middle schools) of Punjab, Pakistan’s largest province, using rich administrative data from the School Education Department. In the scaled up program, calls were made to SC members every month for seven months in 2015, and focused on encouraging SCs to spend funds given to them every year.

A school was identified as SCMP eligible if it had 50% or higher enrollment in each district, school level, and gender (schools are segregated by gender in Pakistan) cell. I leverage the school selection criteria to generate evidence for action, by using a Regression Discontinuity Design. I am able to estimate a causal impact of providing information to SC members via phone calls, on school and student outcomes.

I find that expenditure in middle schools went up as a result of the program, however, that did not translate into improved school outcomes, such as, enrollment, attendance, and improved school facilities. I also find no impact on test scores a year post intervention. My study contributes to the literature in several ways: The design features of SCMP (i.e., a one-to-one, low-cost, and sustained ICT-based engagement mechanism between the provincial government and the SCs to encourage citizen participation) have not, to my knowledge, been evaluated in local governance settings in education. Moreover, my study provides evidence that ICT-based engagement programs may only be able to influence outcomes that are directly targeted through the program (in this case, expenditure of SC funds). Findings from an earlier evaluation of the pilot SCMP program corroborate this evidence; calls in the pilot program targeted enrollment and we found positive and significant improvements in enrollment in Punjab, especially for girls (Asim & Dee, 2017). Overall, findings from the pilot and scaled up intervention provide insights and recommendations for policy makers on adapting program designs to specific contexts. Finally, this paper contributes to the broader literature on strengthening local governance to improve public service delivery in general, not limited to education.