The feminization of unions is underway. It is selective and incomplete, but it is progressing. Many women union leaders describe themselves as feminists and actively defend internal equality policy. But what impact does this feminization have on the way women's interests are represented in the workplace? Is having women unionists defending ‘women's rights’ enough to actually advance gender equality? Research in political science has long considered that a ‘politics of presence’ (Phillips, 1995) was necessary to ensure that the concerns of women (as a group) were put on the political agenda. However, the supposed connection between descriptive and substantive representation (Pitkin, 1977) has been the subject of an extensive literature outlining the pitfalls of such an assumption (Campbell et al, 2010). Critics have pointed out the risk of naturalizing and unifying the definition of what constitutes ‘women's interests’ as such. Scholars have also drawn attention to the distinct social characteristics of women in elected positions, their uneven propensity to formulate and represent the interests of (other) women, and the variations of this representation, depending on the institutional context (Mackay, 2010). Fiona Mackay (2010) contended that the relationship between descriptive representation of women and a better representation of women's interests and preferences is not straightforward but would be better defined as ‘probabilistic’, complex, institutionally constructed, and variable, depending on the problem treated and the political context.

Substantive representation has therefore progressively become a research question, investigating a larger number of representatives acting in the name of women and looking into ‘critical actors’ (Childs and Krook, 2009). These include men (Celis, 2008), but also women outside the parliamentary arena (Squire, 2008), like ‘feminist insiders’ within public administration (Banaszak, 2010) or in equality agencies. The analysis of the representation process itself has been enriched by this literature, moving beyond the idea of ‘a binary relationship between representative and represented that exists prior to the act of representation’ (Dutoya and Hayat, 2016). Scholars have argued that it is the ‘claims making’ process (Saward, 2010) that needs to be investigated, that is, the way in which representatives construct the group that they purport to represent and formulate its interests. Far from assuming a single or mechanical relation, this perspective encourages the analysis of the interactions between multiple actors all claiming to define and defend the interests of those they claim to represent, and of the legitimacy issues they may face in doing so.