Methodological challenges in integration reporting

Integration Department

Project head: Dr. Niklas Harder

Project team members: Samir Khalil

Running time January 2021 until December 2022
Status Completed project

Integration reports and monitors are a popular tool to inform debates around migration and integration in the public, media and political spheres. Moreover, such reports are often used to develop or evaluate integration concepts. However, these reports, which at first glance appear to be neutral, face various methodological problems. Through the work on the Thuringian Immigration and Integration Report and the work on the 1st report on indicator-based integration monitoring, DeZIM is familiar with the various challenges that arise when writing integration reports. The aim of this project is to illustrate the identified problems and possible solutions separately and to make them accessible to the general public through individual publications. In this way, the standard of integration reporting as a whole is to be raised.

Two methodological challenges will be at the heart of the project. Both arise from the group formations and group comparisons that are a central element of integration reporting.

In terms of scientific theory, it is repeatedly postulated that categories and groups result from specific questions. This is in contrast to the widespread practice in integration reporting of examining all integration questions on the basis of simple groups such as Germans and foreigners or migration background and no migration background. We examine how this can lead to erroneous conclusions and make suggestions on how more appropriate groups can be defined depending on the topic area and the question.

Given the groups mentioned above, group comparisons are the core of integration reports. It should be noted that such groups can differ significantly on third-party variables and this can strongly distort the comparisons shown. Linear regressions have been established to control for these biases. We show that new problems arise from the use of linear regressions and propose alternatives that also solve the problem of bias while introducing fewer new drawbacks.

Funding: Federal Ministry of Family Affairs, Senior Citizens, Women and Youth (Institutional funding)