Educational resources and student performance

Researchers on the project
Rune Borgan Reiling (NIFU, project leader), Kari Vea Salvanes (NIFU), Astrid Marie Jorde Sandsør (NIFU), Torberg Falch (NTNU), Bjarne Strøm (NTNU), Eskil Heinesen (Rockwool Foundation Research Unit/International collaborator)

Participating institutions
Nordic Institute for Studies in Innovation, Research and Education (NIFU), Norwegian University of Science and Technology (NTNU) and Rockwool Foundation Research Unit.

International collaborators
Eskil Heinesen, (Rockwool Foundation Research Unit), Björn Öckert (Uppsala University), Diane Whitmore Schanzenbach (Northwestern University) and Victor Lavy (University of Warwick).

Financing of the project
Norwegian Research Council

Project period
2016-2020

Aim of the project
The ultimate goal of this research project is to provide a causal understanding of how teacher density in lower elementary school affects student performance in a Norwegian context. In addition, the project will gain knowledge on how the central government can affect the level of services and output provided by local governments.

Background
The relationship between educational resources and student performance is one of the most debated topics in educational research. Despite increased availability of data and empirical strategies to uncover causal effects in recent years, the evidence of the effect of resources on educational outcomes is still inconclusive. Even for more narrow and popular policy tools, such as class size and teacher density, results are inconsistent both across countries and levels of education. For policy purposes, this implies that studies of the causal effect of resources on student performance for separate levels of education and for separate countries are warranted.

The ideal research design to identify the causal effect of teacher density on student performance is a randomized experiment where one manipulates teacher density. However, this is not how the central government and the school districts operate when resources and teachers are allocated. There will always be a question about the external validity of such experiments, and the understanding of the effect of teacher density will thus be enhanced by other research designs.

One such alternative is to exploit the grant policy initiated by the Norwegian government in 2015. The government decided to introduce a special government grant to increase teacher density in grades 1-4 where the intention was to strengthen early intervention and improve student learning. This grant design provides an opportunity to use quasi-experimental approaches to identify the causal effect of teacher density on student performance. In addition, it can be used to investigate to what extent school districts allocate additional grants in the way intended by the central government, and whether the allocation of the grant and the effect of the grant on student performance depend on school district characteristics

Research questions
In 2015, the 100 school districts with the lowest teacher density were granted 360 million to increase the teacher density for grades 1-4 in primary school. While the intention of the grant was to increase teacher density in targeted school districts, it is unclear to what extent and how this has been enforced. In principle, the targeted school districts had extensive discretion over the allocation of the extra funding. Hence, the design of the policy can be used to increase our understanding of how central and local government policies affect educational spending at the local level. Such analyses are important because the actual effect of the grant in student performance depends on:

  1. The extent to which the grant actually resulted in increased teacher density
  2. How each school district allocates their resources
  3. How each school within a school district allocates their resources

Another important aspect to consider when evaluating the effect of increased resources on student performance is that this might affect schools and student performance differently depending on external conditions. School districts differ in their size, fiscal constraints, labor market conditions, and population, all of which can impact the effectiveness of the grant. In the third part of this project we will investigate whether the effect of increased teacher density and the use of the grant depend on school district characteristics.

Project plan
The project is carried out by researchers from NIFU and NTNU. Project leader is Rune Reiling, NIFU. The project team includes researchers Rune Reiling, NIFU, Kari Vea Salvanes, NIFU, Astrid Marie Jorde Sandsør, NIFU, professor Torberg Falch, NTNU, and professor Bjarne Strøm, NTNU. All researchers in the project team will take an active part in the research and serve as co-authors

International collaborators are professor Eskil Heinesen, Rockwool Foundation Research Unit, professor Björn Öckert, Uppsala University, professor Diane Whitmore Schanzenbach, Northwestern University and professor Victor Lavy, University of Warwick. Heinesen will serve as a co-author, while Öckert, Schanzenbach and Lavy will contribute as keynote speakers at an international conference organized by the project in 2019. In addition, popular science conferences targeted towards policy-makers at both the national and local level will be arranged as part of the project.

The project will be conducted in close collaboration with the other projects funded under the LÆREEFFEKT initiative.

Relevance to society
The project will provide knowledge on the use of non-targeted resources, the labor market for teachers and the effect of increasing teacher density for young pupils, relevant for researchers, policy makers and society as a whole.

Working papers
More to come

Recent relevant publications in international journals
More to come

Recent relevant reports in Norwegian
More to come