Corporate social responsibility (CSR) has been pervading firms’ actions and discourses for more than a couple of decades. One of the earliest definitions of CSR in the early 1960s was given by McGuire (1963, 144): ‘The idea of social responsibilities supposes that the corporation has not only economic and legal obligations but also certain responsibilities to society which extend beyond these obligations.’ Since then, definitions have proliferated, but one consensual point — if not the only one — is the emphasis on the voluntary character of actions undertaken to go beyond legal obligations (Carroll, 1999; McWilliams and Siegel, 2001; Vogel, 2005; Crane et al., 2008). The voluntary character of CSR regulation is particularly critical for transnational corporations having to comply with different domestic laws but also international regulations, norms and soft law (McBarnet et al., 2007). CSR can be viewed as a situation of ‘organizational hypocrisy’ (Brunsson, 2002) where contradictory expectations and demands that are social, political and economic cannot all be met. If satisfying all of them seems unrealistic, the risk of a double decoupling is on the contrary very real: firms have to prevent contradictions between policy and practice (what they say and what they do) on the one hand, and between means and ends (actual results can be at odds with initial aims and plans) on the other (Bromley and Powell, 2012).