2 ms·
TLD's like this one is why statistics are often described as only tools for spreading bias, rather than facts. Depending on which data points one choses to dra
by belorn 10y ago
TLD's like this one is why statistics are often described as only tools for spreading bias, rather than facts.
Depending on which data points one choses to draw a line, one can either say that crime has increased or decreased. During the European migrant crisis of 2015, crime did increased compared to 2014. Gun violence in Sweden also increased significant compared to 2014.
But is comparing 2014 with 2015 a relevant data set? The article provide additional data sets, like the 1995-2015 data set and 2005-2015 dataset. Those say a different story than 2014-2015 dataset. The article do not explicitly say which data gives a correct answer to the claims, only that multiple data sets exist.
We do not have 2014-2015 dataset of the nationality background of criminals. The latest, as mentioned in the article, was in 2005. The argument against doing new such report after the migration crisis is that the government don't expect the result to be different and that the biggest predictor towards crime is socioeconomic, and any group which rank low on that (such as immigrants) will be overrepresented. If immigrants had above average socioeconomic status, then crime should according to this theory go down. If they have below average (which all data points to), it will go up. "Refugees are behind the increase in crime" is not directly opposed as a claim, but the implied cause that is suggested by it is.