| dc.description.abstract | The Swedish Consumer Price Index (CPI) is supposed to measure both inflation and changes in
costs of living. Using a single measure is problematic since costs of living may not change at an
equal pace for all members of society. Because of their varying patterns of consumption,
individuals and households can have higher or lower costs of living than a general CPI indicates.
The purpose of this study is to examine how large these discrepancies are and who is affected by
them. This is done by constructing experimental price indices where goods and services are
given different weights for different categories of households. These weights are based on
household budget surveys, where households are categorised along three dimensions: income,
socio-economic group and composition of the household. A supplemental study is performed
based on Danish data for comparison, because of limitations in the Swedish material. Indices are
constructed for the period 2003–2013 for Sweden and 2000–2013 for Denmark. The results
indicate that small households, households with low incomes and households with higher ages in
both Sweden and Denmark have costs of living that are underestimated by official CPI:s. This is
primarily due to rents and other housing costs making up a larger share of the budgets for these
households, and that these types of costs have risen faster than the average inflation.
Furthermore, the experimental price indices are steadily diverging, at least since 2008. | sv |
