Although the Swedish constitution states that the parliament should represent the people, many researchers have questioned if this is true. Does the Swedish parliament really represent the people? This essay aims to provide an increased knowledge of the descriptive representation in Sweden. To do this, the composition of the parliament and the composition of the entitled voters between 2002 and 2014 are being studied based on three demographic dimensions. The demographic dimensions studied are gender, foreign-/domestic born and age. The main question of the essay is “To what extent does the Swedish parliament represent the entitled voters?”. It intends to be answered using these three, more precise, questions: •  To what extent does the Swedish parliament represent the demographic composition of the entitled voters? • To what extent does the Swedish parliament represent the entitled voters when the three dimensions are merged into a total dimension? • To what extent does a time shift occur in the Swedish parliaments descriptive representation?

A quantitative method is used to answer the questions. Statistics from Statistiska Centralbyrån (SCB) are used as empirical material and analyzed using a mathematical formula called Gallagher index. Gallagher index is being used to find out the representativeness between the composition of the parliament and the composition of the entitled voters for each demographic dimension. Another formula is used to give an indication of the overall representativeness. The essay also applies a comparative design to find out if there is a time shift in representativeness.

The study shows that the descriptive representation of the entitled voters by the Swedish parliament varies in the studied dimensions and over time. The demographic composition of the Swedish parliament represents the composition of the entitled voters between 73,7-97,7% in the period between 2002 and 2014. When the dimensions are merged together, the result shows that the overall representativeness is 68,3%. The study also shows that there is a time shift in the representativeness in 16 out of 24 cases studied.