Creating a System for Data-Driven Decision-Making: Applying the Principal-Agent Framework
This study used principal-agent theory to inform understanding of how systems (districts) can promote data-driven decision making. The study is based on case studies in four urban school systems and focuses on elementary schools’ data use. Study authors highlight three major study findings: First, school level educators need not only systemic support for data use but also enough decision-making autonomy to make site-level decisions on the basis of data. Second, building school site level expertise for data-driven decision-making is a necessary but not a sufficient condition for success. Finally, the design of accountability systems must accommodate the imbalance in information available at central office and the schools. The article discusses implications for further research and policy.
All four systems set a strong foundation for data-driven decision-making by aligning goals, curriculum and assessment, and by creating a culture of data use. Importantly, the implementation of data-driven decision-making was not treated in isolation but rather was a part of a strategy for system-wide change. The systems and the schools co-created similar objectives and values to combat key challenges, as described below (see also Table 3 in article for succinct summary of strategies and how they were implemented):
Divergent objectives by aligning goals and curriculum and assessment across the system and establishing a common language and culture for data use.
Information asymmetry between districts and schools by establishing a structure for the bottom up flow of information and investing in training and data management personnel to support schools
Weak incentives by linking data use with federal and state accountability, school improvement plans, and performance-based compensation systems
Limited local site decision rights by balancing school-site decision making power with cross system consistency, for example by making clear school leaders’ decision authority and providing educator autonomy in instruction even with system-wide curriculum and assessment.
Adverse selection by hiring agents with data capacity and commitment, providing targeting professional development and structuring time and opportunity for collaboration over data within and across schools.