by Theo Sparreboom, Statistics Department, ILO

Within the United Nations system, the ILO is the only organization to have a general, explicit mandate to work on co-operatives. The Co-op- eratives Unit of the ILO (COOP) was established in 1920 in recognition of the role of co-operatives in employment creation and their contribu- tion to the advancement of social justice. Located in the ILO's Enter- prises Department, COOP contributes to ILO's overall mission of creating decent work for all through the promotion of sustainable en- terprises of all types and sizes. Within this broad mandate, COOP is re- sponsible for ILO's work on job creation and growth through co-op- erative enterprise development, including applied research, policy ad- vocacy and advisory services, technical co-operation, and partnerships.

Of particular importance, the implementation of ILO Recommendation No. 193 concerns the promotion of co-operatives through advisory services on legal provisions and policies at the national and international levels. Paragraph 8 of Recommendation No. 193, adopted by the International Labour Conference in 2002, encourages governments to ensure that national policies " seek to improve national statistics on co-operatives with a view to the formulation and implementation of development policies."

Statistics on co-operatives are essential, too, when it comes to quantifying and analyzing the impact they have on their owner-members, as well as on the economy as a whole. They can provide evidence, for ex- ample, on how co-operatives fare with respect to employment and in- come in comparison with other economic units and throughout economic cycles. Statistics on co-operatives also help to elucidate how different business models affect communities and workers, which is a topic of growing interest. At a minimum, such statistics should include: (a) the number of co-operatives and their characteristics; (b) the number of members of co-operatives, the number of workers engaged in co-operatives and their characteristics; and (c) statistics on the value of production by co-operatives.

Compilation and analysis of statistics on co-operatives has been carried out in many countries, but has generally been based on administrative information drawn from co-operative registers, and much less on establishment and household surveys. Administrative information serves as an important source of statistics, but may also suffer from shortcomings in terms of, for example, timeliness, coverage, and (inter- national) comparability. Furthermore, registers do not necessarily include information on membership of co-operatives or the number of workers employed in co-operatives. Nevertheless, surveys targeting registered co-operatives represent a rich source of information on current activities and challenges, which may relate to both economic and social data.

A second source of data could be household surveys that have the advantage that they allow for the production of complete statistics on the number of members in co-operatives, as well as the number of workers engaged in them. In contrast to administrative information, in principle, household survey-based information prevents double counting of workers engaged in more than one co-operative, and includes socio- economic characteristics of workers that are difficult to obtain through other means of data collection. Thus, regular household survey-based information can be used, for example, to analyze relationships between economic sectors, types of co-operatives, and characteristics of workers. In contrast, household surveys do not allow for the production of ac- curate statistics on the number of co-operatives or their economic value.

Current statistical standards make reference to "members of producers' co-operatives" in the International Classification of Status in Employment (ICSE), which was adopted in 1993. This figure can be easily misunderstood because the ICSE definition of members of co-operatives only includes workers who work in a producers' co-operative of which they are also members. It therefore excludes workers who own a farm and are members of a co-operative that provides services to them, such as the marketing or processing of the goods they produce.

This means that in Indonesia, for example, no members of producers' co- operatives are reported in international statistics, while according to government data there are more than 30 million members of co-operatives in that country.

At the 19th International Conference of Labour Statisticians (ICLS) held in Geneva in October 2013, governments and workers' and employers' representatives reaffirmed the importance of having more comprehensive and (internationally) comparable statistics on co-operatives. It is worth mentioning that this was the first time in its history that a specific item on co-operative statistics was included in the ICLS agenda. This session at the ICLS provided an opportunity to share experiences on the current production of statistics on co-operatives, as well as on the statistical challenges faced in many countries.

The ICLS adopted a resolution concerning further work on statistics of co-operatives, and several country representatives expressed their interest in sharing more details of their experiences and in participating in pilot exercises to test new or modified measurement approaches. The resolution recommends that the Office, in co-operation with the ILO's constituents and interested national statistical offices, carry out further developmental work on the measurement of co-operatives through administrative registers, establishment or household surveys.

Accordingly, data collection on co-operatives in selected countries is being reviewed by the ILO, and pilot activities are under consideration.