IMDi has commissioned Fafo to create a state of knowledge report about parallel societies in Scandinavia, focusing on characteristics, prevalence and measures. This report is based on our review of 89 research contributions from Norway, Sweden and Denmark. We have searched for the term ‘parallel society’ in various portals and research databases, in English and in the three Scandinavian languages. Because parallel society is a little-used term in social research, we have also searched for the related phenomena ‘vulnerable areas’ and ‘residential segregation’, as research on these phenomena can shed light on some of the challenges that the concept of parallel society is intended to capture.
Our review shows that there is relatively little research that uses the concept parallel society in Scandinavia. The understanding of the term varies in research, including everything from poor integration to structures that allow criminal gangs to maintain control and power, little contact between minority and majority, or minorities' lack of adherence to the institutions of the larger society. For some researchers, parallel societies are groups that form their own institutions, such as parallel legal systems or educational institutions. The research is consistently critical of using this term in Scandinavian contexts and we find few researchers who have empirically investigated whether parallel societies exist in Scandinavia. The research therefore does not provide a basis for saying anything about the prevalence of parallel societies, in Norway, Sweden or Denmark.
In Denmark, the concept of parallel societies has for many years been an integral part of political rhetoric and legislation aimed towards vulnerable residential areas. However, Danish authorities have a definition of a parallel society that differs from those found in research. From the perspective of Danish authorities and according to Danish law, a vulnerable area is an area where two of the following criteria are met: high unemployment, high criminal conviction rates, and/or low higher education attainment rates and/or low average gross income. A parallel society is a vulnerable area where the proportion of immigrants and descendants from non-Western countries exceeds 50 percent. Each year, the Danish authorities prepare lists of areas defined as parallel societies according to these criteria.
Our review of research on vulnerable areas shows that who define an area as vulnerable and what vulnerability criteria are used, are crucial for what is found and monitored. For example, it is the police that maintain lists of vulnerable areas in Sweden, which contributes to crime rates being heavily weighted. The sizes of the areas also vary. In Norway, there are no official lists of vulnerable areas, but challenges in living conditions are emphasized in Norwegian area initiatives.
Defining areas as vulnerable according to certain set criteria can be a useful tool for monitoring developments and for implementing relevant measures. An advantage of defining areas as vulnerable is that measures can be implemented only where they are needed. At the same time, the research we have reviewed shows that it is important that the criteria and justifications for vulnerability match the actual challenges that exist, and that different criteria are not mixed together. Otherwise, there is a risk that measures to combat crime or reduce school dropout rates are implemented in areas based on immigrant population sizes, rather than in those where crime or dropout rates are the highest. Research also shows that politically defining an area as vulnerable can contribute to increasing the stigmatization of the area and those who live there and further exacerbating the problems that exist there. It can also legitimize the implementation of very intrusive measures that would not be implemented elsewhere.
When it comes to residential segregation, that people live close to people who resemble them, research shows that segregation is increasing in Norway. This applies to both socioeconomic and ethnic segregation. When we compare Norway with other countries, we still see that the levels are lower than in Sweden and significantly lower than in, for example, the United States. Exactly how the picture looks depends on how segregation is measured. Oslo, for example, is relatively segregated at the district level, but segregation is much lower in Oslo than in Stockholm and Copenhagen at the neighborhood level. Research on residential segregation also shows that there are several causes of segregation, including housing policies, housing prices and preferences of both minorities and the majority.
Complicated societal challenges are rarely solved easily. Since there are so many different understandings of what parallel societies are, and a lack of empirical basis for saying that they exist, it is difficult to identify effective measures to limit or prevent their emergence. Which measures are relevant depends on how the phenomenon is understood. If parallel societies are understood as a form of antithesis to good integration, the entire immigration and integration policy area is relevant. If parallel societies are understood as the foothold of criminal networks, the entire justice sector warrants review.
Our search reveals measures within four categories: 1) area initiatives, 2) measures in kindergartens and schools, 3) inclusive leisure activities and 4) crime prevention. Research does not always provide clear answers to the effects of these measures. There is also uncertainty about whether the measures reach the most vulnerable groups.
The type of measure we find most discussed in our search is area initiatives. Research in all three countries shows that area initiatives can improve living conditions for residents, but that they do not reduce segregation. The parallel society agreement and the area initiatives in Denmark have contributed to reducing the number of parallel societies, through very intrusive measures such as the demolition of large residential areas and a ban on certain groups from moving into certain areas. The evaluations of these measures show that the areas in which the measures have been implemented have become more mixed, but segregation in Denmark still persists. Also, the measures do not seem to have improved the lives of the most vulnerable. A common challenge in evaluations of area initiatives in all three countries is related to isolating and measuring the effect of measures. Area initiatives are often extensive and complex, with many objectives and activities. This makes it difficult for research to identify which specific measures within the initiatives actually have a measurable effect.
The measures implemented in Norway, Sweden and Denmark must be understood in light of the different national contexts and structural differences, particularly in housing policy and segregation patterns. What works in one country is not necessarily transferable to another. In addition, research shows that various actors interpret challenges in vulnerable neighborhoods differently, affecting which measures are proposed. Regardless of context, the research points out that complex structural conditions require comprehensive solutions – not individual measures.
A set of measures to address parallel societies cannot be derived from the review we have done. For that, the phenomenon is far too poorly defined in research. We recommend that the authorities refrain from using this term, which our review shows is not clearly defined and potentially stigmatizing. Rather than grouping together many different problem areas, research suggests that measures are more effective when focused on specific concerns, whether those are segregation, poverty, exclusion, lack of integration or crime. When the problem you are seeking answers to is more clearly defined, it is also easier to use research to find good solutions.
