Sweden has received international criticism from the UN and the Council of Europe, who consider that there is a lack of knowledge of what conditions are like for different groups in Sweden. In countries such as the United Kingdom, USA and Canada, questions about ethnicity and religion are almost routine in surveys, but not here.
“It is linked to our history. The race registration done by the Swedish race biology institute fell into disrepute after the genocide carried out by Nazi Germany, and a decision was made not to collect such data. The state should not divide up inhabitants according to their race and ethnicity,” says Edda Manga, researcher in history of ideas at the Multicultural Centre.
She is leading the research project Metodologiska Laboratorier, which is looking for ethical and scientifically sustainable methods for mapping discrimination. With this project, they want to develop methods for creating statistics-based knowledge of how the living conditions of different groups differ in relation to skin colour, ethnicity and religion.
The issue of equality data has been much debated by researchers and the anti-racism movement, and is complicated by views differing widely between different exposed groups.
“Among Afro-Swedes, many are asking for this type of statistics, while groups that have been subjected to state repression, such as Jews, Sami and Roma, have strong views against this,” says Edda Manga.
In the last six months, equality data have become a hot political issue all the way up into the Riksdag. The Left Party was the first political party to take a standpoint in favour of collection at last year’s party conference.
They want Statistics Sweden to ask questions about ethnicity and skin colour, in the same way that the agency today asks about gender. The motion behind the decision describes such statistics as necessary for developing political measures against discrimination.
Problems with survey questions
Surveys can be a tool, but there is a basic problem according to Edda Manga:
“By asking questions about skin colour and ethnicity, we establish a way of thinking where we divide people up according to racist categories.”
She points to several ethical and scientific challenges of asking people to identify themselves ethnically or racially. Studies show that many respondents find it unpleasant or offensive to be asked this type of question. This is a problem in itself, but is also means that many people will probably skip the questions. It is not even the case that everybody could answer, says Edda Manga.
“In practice, questions of ethnic identity or racial classification are context-bound and complicated. I am not quite sure myself what answer I would give to the question of what ethnicity or ‘race’ I ‘belong to’.”
In the United Kingdom and USA, researchers use a number of pre-determined categories for collecting the data, but in Sweden there is a proposal to let everybody describe in their own words how they define themselves. Then the survey itself does not define any racist categories. But this design places great demands on researchers, who must find a way of grouping the answers they receive.
Measuring differences in living conditions or exposure to discrimination is, in short, difficult. Different researchers and research teams have tried various methods. For example, experiments using fictional job applications have been conducted, showing that the chance of being called for an employment interview declines if your name has Middle Eastern associations.
Using register data instead
At the Centre for Multidisciplinary Research into Racism (CEMFOR) at Uppsala University, researchers have mapped Afro-Swedes’ situation on the labour market using the population register.
Their report Antisvart rasism och diskriminering på arbetsmarknaden (“Anti-black racism and discrimination in the labour market”) shows clearly that persons with African background has a worse position in working life than the rest of the population. The group receives less pay, has more unemployment days, and low representation in high-status jobs and in managerial positions.
The study is ground-breaking in its method, but this has some shortcomings. The researchers have identified the Afro-Swedes as a group by selecting persons who were born in sub-Saharan Africa, or whose parents were. This means that white persons are included in the group, while it misses black persons whose background is from other parts of the world, such as North or South America.
“I still think it is lightyears better than using surveys. Methodologically, register-based data are difficult to beat, as we get the whole picture. Using surveys, we always have dropouts, and they are answered by persons in a subjective way,” says Irene Molina, one of the researchers behind the study.
More qualitative studies needed
Collecting equality data by asking respondents questions about ethnicity, skin colour or religion is the wrong way, according to her.
“I think it is a pity that the methodology discussion only has concerned surveys, when there are so many other methods that we should try.”
She particularly emphasises the need for more qualitative studies.
“Statistics are needed, but I also want to understand the underlying mechanism: How does discrimination function? What situations open the door to discrimination? These are questions we could never answer based on statistics.
Testing different methods
Irene Molina feels that there is a great lack of knowledge of how register data can be used to visualise differences in living conditions between different groups. On the other hand, she establishes, there are limits to the groups that can be identified using the data. For example, it is difficult to separate out a specific religious group – but perhaps not impossible?
Within the project Metodologiska laboratorier, researchers are investigating whether it is possible to produce figures for the living conditions of Muslims, using a model similar to the one used by the researchers at CEMFOR.
“We have discussed looking at country of origin in the register data, and weighting this with the proportion of Muslims in the different countries. This is not an ideal method, but it is worth testing,” says Edda Manga.
She and her colleagues are half-way through the project, and will continue testing different investigation methods for another two years. Although using surveys involves many challenges, she is not discounting this as a useful method.
“I believe that we will conclude that several different methods are needed to map discrimination,” says Edda Manga.