Demographic data on unemployment, age and foreign background can be an important complement to the police's definition of vulnerable areas, says Mia Puur in her criminology master's thesis.

“When the police map vulnerable areas, they make a larger analysis that is based, among other things, on the crime in that specific area,” says Puur.

Her survey, conducted in Sweden, is a national micro-unit analysis based on a statistical grid consisting of squares with the dimension of 250 x 250 metres, where each square has information about socio-demographic data.

My method broadens the picture, and can mean a streamlining of the police work and open the way for more targeted efforts

Mia Puur

The main purpose has been to identify a statistical model that can more objectively identify areas that are vulnerable or not.

“It can be a big difference between what a police officer in a small town sees as serious, and what a police officer who works in a vulnerable area in a large city sees as a problem,” she says.

Divided into ‘vulnerable areas’, ‘risk areas’ and ‘particularly vulnerable areas’, the police has identified 60 such areas in Sweden. Using the grid method, Puur has compared demographic data for these boxes with other areas in the country. Is it statistically possible, based on socio-demographic data, to identify vulnerable areas in the police's national database, she wonders?

As a basis, she has used data from the Central Bureau of Statistics in Sweden on factors such as unemployment, age distribution, the proportion of residents with a foreign background, and households with single parents.

“Analyses have shown that these indicators are more prevalent in vulnerable areas," she adds.

According to Puur's model, at least four continuous squares are required for the area to be classified as exposed to vulnerability. The results showed 13 new areas that are completely independent areas that risk being exposed in the future, which authorities should therefore be vigilant about.

She also found 16 new ones that border other vulnerable areas. But there is also the reverse relationship: some areas deemed as vulnerable are, according to her model, no longer regarded as such.

“In the city of Malmö, some neighbourhoods are emerged and classified as a single vulnerable area, whereas in other parts of Sweden, areas from the police’s national compilation include anything from four to 76 squares. If areas become very large, the risk of stigmatisation increases,” says Puur.

The results of her essay are generally quite in line with the police's national compilation. Puur emphasises that she did not have access to crime data, which would probably further strengthen the coherence.

“My method broadens the picture, and can mean a streamlining of the police work and open the way for more targeted efforts. Social vulnerability is a comprehensive problem that cannot be solved by the police alone,” she concludes.

Text: Magnus Jando