Table 2.
Description of the variables used in the regression model.
| Variables | Description | Units |
|---|---|---|
| %_soc.minimum | Percentage of households living below the social minimum threshold | % |
| %_≥_65 | Percentage of individuals over 65 years of age | % |
| %_ ≤ _17 | Percentage of individuals under 17 years of age | % |
| %_ ≤ _bat_45 | Percentage of dwellings built before 1945 | % |
| %_≥_bat_90 | Percentage of dwellings built after 1990 | % |
| %_residences | Percentage of collective residences | % |
| Mean_income | Mean yearly income per person (sum of winsorized* living standards/number of inhabitants) | euro/person |
| Density | Residents per km2 (number of inhabitants in the square/(0.2 km*0.2 km) | pop/km2 |
*Winsorization is a statistical technique used to treat the extreme values of a distribution. In a given unit, an individual's standard of living is reduced to the 95th percentile of the unit distribution if her standard of living is higher than this threshold. Conversely, her standard of living is reduced to the 5th percentile of the unit distribution if her standard of living is below this threshold. If the standard of living is between these two thresholds, no treatment is carried out.