Current international guidelines

Current guidelines on gender mainstreaming were adopted by the Seventeenth International Conference of Labour Statisticians in December 2003. Checklist of good practices for mainstreaming gender in labour statistics stresses the need for support at all levels in the responsible agencies to ensure that relevant topics are measured, including informal employment and unpaid (non-SNA) work, that all workers are covered and that statistics are published to reflect the particular circumstances of workers, including family context. It is expected to significantly assist the ILO in enhancing its technical advisory capacity, as well as countries in improving national labour statistics.

According to this checklist, mainstreaming gender in the production of labour statistics should affect each stage of the data collection and production process. This means that when setting up definitions, designing measurement methodologies and deciding on dissemination procedures, the existing differences between men and women in the labour market are being taken into account and reflected. These differences relate to women’s and men’s types of contributions, their access and control over resources and benefits and their needs, constraints and opportunities.

Statistics on conventional labour subjects, such as employment, unemployment, strikes and occupational injuries, when disaggregated by the sex of those involved will, as a general rule, always be useful to describe gender issues. But labour statistics that are gender mainstreamed are more than information by sex. In addition they need to satisfy at least the following four characteristics:

  • First of all, they should relate to issues or areas that are relevant to enhancing the understanding of men’s and women’s positions and interrelations in the labour market. Statistics are needed on subjects where there are important inequalities between women and men, including areas related to how men and women balance their working life with their other obligations, such as family life; the full extent of men’s and women’s participation in productive activities, covering their participation in the paid labour market as well as in unpaid production of goods and services for consumption by their own household; the different contributions of men and women to the labour market and how this translates into differences in tasks and duties (occupations), working time arrangements, status in employment, occupational injuries and income inequalities.
  • Second, labour statistics should cover and adequately describe all workers and work situations. The identification and adequate description of “atypical” work situations – i.e. those which do not reflect a common view of what “working” and “joblessness” are all about - is the most important challenge for conventional labour statistics. It is more difficult to identify and describe work situations which are informal, irregular, short time and unpaid than work which is paid, full-time, regular and in formal sector establishments. Measurement definitions need to be based on criteria that do not exclude groups of workers or work situations, and measurement methodologies need to apply special procedures when there is a risk that groups of workers or work situations may be overlooked.
  • Third, labour statistics should be sufficiently disaggregated to show meaningful distinctions between men and women. Broad population groups can be very heterogeneous and comprise a diverse set of employment situations where men and women are present to different extents. For example, analysing the managerial group as a whole will not reveal the fact that women may be concentrated in managing small enterprises, while most of those managing larger companies are men. Similarly, analysing earnings as a whole may hide the fact that it is most often men than women who receive family allowances and other benefits.
  • Finally, the way labour statistics are presented and disseminated should reveal significant differences and similarities between men and women, and the factors that may cause them. This implies relevant cross-classification of labour statistics by variables which present the demographic, economic, social and family context of workers, including, in addition to the workers’ sex, at least their level of education, their marital status and most importantly, the presence in the household of small children and other persons requiring care.