A number of factors can limit the comparability of statistics on employment by economic activity between countries or over time.
Comparability of employment statistics across countries is affected most significantly by variations in the definitions used for the employment figures. Differences may result from age coverage, such as the lower and upper age bounds for labour force activity. Estimates of employment are also likely to vary according to whether members of the armed forces are included. When the armed forces are included in the measure of employment they are usually allocated to the services sector. Therefore, in countries that do not include armed forces, the services sector tends to be understated in comparison with countries where they are included.
Another area with scope for measurement differences has to do with the national treatment of particular groups of workers. The international definition of employment calls for inclusion of all persons who worked for at least one hour during the reference period. Workers could be in paid employment or in self-employment, including in less obvious forms of work, some of which are dealt with in detail in the resolution adopted by the 19th ICLS, such as unpaid family work, apprenticeship or non-market production. The majority of exceptions to coverage of all persons employed in a labour force survey have to do with slight national variations to the international recommendation applicable to the alternate employment statuses. For example, some countries measure persons employed in paid employment only and some countries measure “all persons engaged”, meaning paid employees plus working proprietors who receive some remuneration based on corporate shares. Other possible variations to the norms pertaining to measurement of total employment include hours limits (beyond one hour) placed on contributing family members before for inclusion in employment.
Comparisons can also be problematic when the frequency of data collection varies. The range of information collection can run from one month to 12 months in a year. Given the fact that seasonality of various kinds is undoubtedly present in all countries, employment figures can vary for this reason alone. Also, changes in the level of employment can occur throughout the year, but this can be obscured when fewer observations are available.
It is also important to note that different versions of the ISIC can be used across countries, with countries moving to adopting the most recent version at different paces. A country may continue to use the previous version even after starting a new data series according to the most recent version. Although these different classification systems can have an impact on comparability at detailed levels of economic activity, changes from one ISIC to another should not have a significant impact on the information for the three broad sectors presented in ILOSTAT.