Pakistan has extremely low learning levels and poor education service delivery is a driving factor. The Khyber Pakhtunkhwa (KPK) government has expressed a strong desire for evidence on how to design evaluation and oversight systems for the education sector, with the goal of improving learning.
The government has two evaluation and oversight systems that should, in theory, incentivise teacher effort; in practice, they face substantial problems and are ineffective. First, teachers are evaluated annually via Performance Evaluation Report (PER) that feeds into promotion decisions. However, the PER is not teacher-specific, but general to the civil service, and accordingly does not properly measure aspects of teaching. Additionally, PERs are carried out by the headteacher, who - in Pakistani society - is hesitant to criticise colleagues. As a result, all teachers generally receive the same PER score, rendering scores meaningless. Second, a school inspection system exists, but inspections do not occur regularly and there is no guidance on how often or what inspectors should inspect, and how they should report findings.
We will examine two interventions to improve teacher performance and learning outcomes. First, in a random subset of villages, we will introduce teacher-specific performance evaluations (PERs), to be carried out by an independent third party on an unannounced day. Second, in another random subset of villages, we will ensure that meaningful and informative school inspections occur by establishing clear guidelines and evaluation criteria. Teacher evaluation and school inspection outcomes will be linked to faster and slower promotions for teachers and headteachers, respectively. We will work with the government to decide what data to collect and how to provide feedback to the education department and school. In another random subset of villages, we will undertake both interventions. A control group will receive no interventions.
This is the first study seeking to improve both teacher and school (i.e. headteacher) accountability. Literature shows that both levels are important. Teacher accountability reduces absenteeism and improves learning outcomes, and school accountability through improved monitoring has also been shown to improve school performance. However, it is not clear which level is more important in improving performance, nor whether improving accountability at both levels may provide higher value for money, suggesting complementarities. Knowing which levers and the ensuing rewards and sanctions are most effective (and cost-effective) is important both for education economics and for policymakers.