The main goals of the paper are to describe the index (launavísitala), the data used for its calculation and to provide information on the statistical methods used by Statistics Iceland for the construction of the index. Furthermore, some of the most relevant errors and biases known to affect general price index numbers are analysed and assessed for the case of the Icelandic Wage Index, i.e. the chaining effect and the influence of life cycles.
The main findings of the present analysis are the following:
- The wage index is found to be of good quality, using standard methods of price index theory utilising comprehensive data. Its main challenges are lack of coverage and adequate sample refresh.
- The wage index is a price index based on data of the Statistics Iceland's wage survey and administrative data from tax authorities.
- Statistics Iceland compiles its wage index based on a Törnqvist index formula at lower aggregation levels using a matched sample model with the smallest aggregate being the item, i.e. the hourly wage paid to an employee for fixed working hours within the same occupation the same economic activity and the same employer. At top level the sub-indices are aggregated according to Laspeyres' formula since weights at this level are updated once a year.
- The drift effect is of the same size as the rounding errors of the index, due to the monotonous character of the wage data and to the choice of the index formula.
- There exists a very weak effect of the items' life cycle on the values of the sub-index values.