In both research and policy circles, high-stakes accountability is increasingly perceived as a generator of side effects. In contrast, low-stakes accountability is conceived of as a source of formative assessment, educational improvement and, accordingly, as a more desirable policy option. Nonetheless, our research shows that the reactivity of school actors to accountability does not only depend on the formal consequences of external evaluations, but on the very fact of school actors being externally surveilled. Through a survey experiment, we have been able to analyse the effects of different types of incentives (individual or collective, symbolic or material) on the likelihood of teachers trying to inflate assessment results. Counterintuitively, we show that the symbolic consequences of accountability predispose teachers to over-react, even more frequently than material consequences. Teachers’ professional reputation concerns, and the fact that national assessments, independently of the stakes, work as a form of commensuration are some of the social mechanisms explaining this counterintuitive finding. A second survey experiment has captured teachers’ preferences regarding school governance policy in three different countries. The experiment shows that, in all country contexts, material incentives, in the form of individual salary bonus, are not a main driver of teachers’ motivation, and that teachers prefer their work to be assessed through methods of a qualitative and procedural nature (such as classroom observations), as well as work environments with clear objectives and cooperative relationships.
Our research findings invite policy-makers to rethink accountability policy, and to give greater consideration to the locus and space of accountability. New forms of network and incumbent accountability, grounded in local education spaces that transcend individual schools, could contribute to generate organisational improvement, quality assurance and policy learning dynamics among schools, without necessarily generating competition and segmentation within educational systems.
Finally, the REFORMED Survey delivers large-scale datasets that can be used by scholars to explore and analyse the enactment of school autonomy and accountability policies in different contexts, by taking into account context-, school- and individual-level relevant variables that have been overlooked in previous research. The conception and development of the REFORMED questionnaires, as well as the sampling strategy, have involved a number of steps aiming at comparability across cases and at obtaining data of a very high standard.
