Producing decent work indicators: contested numbers at the ILO

John Berten (2022)

Policy and Society, Online first

Abstract. The article investigates the production of decent work indicators within the ILO, to demonstrate that developing measurement infrastructures in global policymaking requires political work. The concept of decent work responds to the perceived marginalization of the ILO in social and labor policy and was supposed to provide a new unifying normative framework for the organization. The article shows that creating decent work indicators encountered challenges due to its highly politicized production process. Proponents of quantification (mostly workers’ representatives) and opponents (mostly employers’ representatives) disagreed about the function of indicators: should they be country-specific or allow for universal assessment of progress from above. In effect, although indicators of decent work have been integrated into the Sustainable Development Goals—mostly as part of goal no. 8, many are still incomplete. As a result, the indicators did not establish a “framework of assessment,” which would have been guided by universal standards of progress allowing the ILO to “govern at a distance,” and could not initiate a paradigmatic policy shift, impeding the infrastructuralization of measurement. Theoretically, the article advances our understanding of policy formulation and design on the transnational level by showing the political foundation of knowledge-based instruments. Empirically, it rests on a Grounded Theory-based analysis of key ILO documents, including Governing Body minutes, conference and expert meeting reports, and official publications, mainly from the period from 1998 to 2015.

Extract: “Indicators usually make invisible how they are constructed (Merry, 2011, p. S86), but “standardization and evidence are both socially embedded and politically contested processes of knowledge creation” (Elgert, 2015, p. 342). According to Desrosières (2016, p. 184; emphasis in original), producing numerical knowledge “entails firstly, agreement, that is deciding on the conventions, choice of objects and modes of equivalence and then, once these conventions have been finalised, proceeding with the measurement operations proper.” However, to make something comparable is an “active” construction of similarity, requiring epistemic and political labor, including potential contestation […]” (p. 4)

Lien / Link