Change in the methodology of the data collection system of institutional labour statistics
The change will result in more accurate data
The data collections of interim and annual institutional labour statistics (designating data providers and producing estimates for enterprises surveyed by sampling) are based on the Classification of Legal Forms of Enterprises (GFO) in HCSO. In the classification, the administrative data of organisations are continuously updated, but certain statistical criteria (e.g. staff category) are set and recorded only once a year. In order to keep response burden at a reasonable level, enterprises and nonprofit organisations get into the range of data providers only above a certain number of employees. According to the staff category fixed newly every year, the composition of the reference population may be rearranged, therefore, the comparability of published estimated total values is limited from year to year.
The essence of the methodological correction valid from 2018 is that HCSO also used the numbers of employees estimated from the interim tax returns for fixing the staff categories of the year, so that in case of organisations within the scope of the labour statistical data collection, background information better reflecting the composition of the target population in the reference year should be available for producing estimates for enterprises surveyed by sampling as well as for defining the fixed staff categories used in publications. These pieces of information are more up-to-date than those on the numbers of employees gained from the annual tax returns used earlier, since they refer to the year preceding the reference year, while data from the annual tax returns are from one year earlier. As a consequence, the methodological change described has caused a one-off, major rearrangement in the composition of the reference population, so, in terms of total values (number of employees, total earnings, wages and salaries according to the SNA concept, total hours worked) published since January 2018, there is a significant break in the time series.
The comparability of the target indicators of the institutional labour statistical data collection, i.e. the published specific indicators (e.g. average earnings, working hours per capita) is not basically limited by the above phenomenon. For publishing indices expressing the percentage change compared to the same period of the previous year, from January 2018, we use the estimate filtering out the effect of the methodological change outlined above, therefore, the published indices may differ from the values that can be calculated from the data published earlier.