Although gender mainstreaming as such has only been a recent requirement when producing labour statistics, it has by and large been inherent in these activities. This is due mainly to two reasons.
The first is that in social statistics, the distinction between men and women is the most efficient way to minimize variability within groups while maximizing variability between groups, so as a general rule they have always been disaggregated by sex. Thus, national statistics tend to be produced and disseminated by sex when this information is meaningful.
The second reason has to do with the statistical principle of comprehensiveness. For labour statistics to be useful they need to provide an accurate and complete description of the labour market. This requires definitions and classifications which taken together can reflect well the different work situations of all participants in the labour market; that measurement needs to use sound methodologies that will ensure that these particular work situations are clearly and consistently identified; and dissemination practices need to present data in such a way that differences and similarities, and the causes that cause them, are clearly highlighted.
Notwithstanding the above, there is still a lot to be done. Currently, labour statistics as commonly produced tend to be incomplete. Labour statistics are generally weak in identifying and describing “atypical” forms of employment, characterised by being casual, part-time, informal, undeclared to tax authorities, unpaid, carried out at or close to the home, and which are often interchanged with domestic activities. Women tend to be, more than men, in such “atypical” work situations, and as a result, are often underestimated and less well described than men. Furthermore, labour statistics are not always presented in a way which is necessarily useful to identify differences or similarities between men and women workers. Fortunately, mainstreaming gender in the production and presentation of labour statistics is now becoming a core principle of work in many national statistical agencies and it can be expected that labour statistics will improve their overall quality.
In the ILO, all the statistics compiled and disseminated by the Bureau of Statistics, when meaningful and readily available, are disaggregated by sex as a minimum. In addition, the Bureau of Statistics launched a new database in 1993 to measure and analyse occupational segregation between men and women in the world. It is called SEGREGAT and contains statistics on employment by sex and detailed occupational groups. This database was updated in 2002 and now covers 85 (developed and developing) countries. A comprehensive analysis based on these statistics is presented in Richard Anker (1998) (webreference). The database was also used to provide background information on women and men managers, see Linda Wirth (2001) (webreference), which analyses why women find it hard to come into positions of power. In addition, the database has been used for many other applications, and judging from the amount and type of internal and external requests we receive for its use, the database became more valuable than anticipated.