In 2004 Mexico managed to consolidate a highly complex and contradictory large-scale assessment (LSA) landscape comprised of three particular programmes run by different institutions, each with their own special interests and their own distinct scalar architecture (global, regional and national), yet mutually focused on seeking to shape education practice improvement and policy decisions through the production of data regarding student learning.

Existing literature in the field of LSAs and governance tends to address this phenomenon through the lens of one or at best two programmes, and their effects on single cases. As a result we do not know enough about the origins, nature and problems generated by these sorts of arrangements and their implications for education. Adopting a critical theory stance, this thesis develops an empirically driven multi-scalar approach to explore large-scale assessments and governance in three layers of the Mexican context: local, sub-national and national.

To this end, it makes use of a multi-scalar embedded case study research design, and a multi- phase mix-methods methodology which draws upon an original data set comprised of one survey applied to 319 educators; semi-structured interviews with Senators, Secretariat of Education authorities, Governing Board members from the National Institute for Education Evaluation, sub-national Education Authorities, educators and school leaders; and a set of LSA publications.

Findings suggests that the assessment regime in Mexico came into being as the result of wider global efforts to improve education systems, at the same time that neoliberalism as a political project emerged as a way of rethinking state-society relations. The current arrangement is comprised of three LSAs namely PISA, ERCE and PLANEA, all of which operate at particular scales, and are underpinned by distinctive programme ontologies: respectively, human capital, human rights, and public accountability. There are significant variations in each case regarding how LSAs govern the education sector in Mexico. Even so, these relationships are underlined by a set of governing tools which set in motion competitive comparison, principles of classificatory judgment and flattening practices, which help frame, reinforce, amplify and reproduce beliefs and judgments amongst participants, consequently influencing how education issues and problems are approached and viewed and how their solutions are framed and engaged.