Since the 1990s, international and regional organisations have responded to calls to open up to civil society. Some, however, remain relatively inaccessible. The Southern African Development Community (SADC) is one of those regional organisations that seems to have resisted opening up, making it a “tough test” for civil society engagement. Yet, even in this difficult regional context, we observe the existence of transnational civil society networks that aim to engage and influence regional governance, although there are notable differences across policy sectors. It is likely that the characteristics of these civil society networks have an effect on civil society participation in, and engagement with, regional governance. Against this background, this paper asks: How do the characteristics of transnational networks contribute to civil society engagement in regional governance in SADC? The paper employs a comparative case study design focussing on civil society engagement in two policy sectors: gender, and employment and labour. Using an interview-based approach to social network analysis (SNA), we map the two policy networks surrounding the implementation of Article 8 of the SADC Protocol on Gender and Development, which aims to end the practise of early marriage, and the ongoing ratification of the SADC Protocol on Employment and Labour. We complement the social network analysis with semi-structured interviews with a variety of stakeholders, including civil society, donors, researchers, and national and regional policy-makers in Botswana, Mozambique, South Africa and Zambia. The findings reaffirm the state-centric nature of SADC and the difficulties for civil society to engage meaningfully. We also find notable differences in the civil society networks involved in the two policy sectors: The gender sector is driven by civil society organisations (CSOs) and financed by donors, with member states playing a relatively minor role, whereas member states are central players in employment and labour. The gender network is highly centralised, with one central CSO performing a coordinating role, whereas the labour and employment network is very dense and shaped by many interactions between different actors with diverse political aims. The findings suggest a trade-off between a hierarchical, centralised network that is efficient when it comes to sharing resources, versus a dense, consensus-finding network that mitigates potential conflicts. The networks for both policy sectors also suggest that the SADC Secretariat is more accountable to donors than CSOs, a reflection of SADC’s dependence on donor funds. We find that many of the challenges to civil society found at the national level in developing countries are replicated at the regional level. Questions surrounding the extra-regional funding of CSOs, their representativeness and their legitimacy pose great challenges to civil society networks. Nevertheless, civil society networks have the potential to act as drivers of people-centred regionalism, but so long as the institutions and organisational culture of SADC remain a “closed shop”, their potential will go unrealised.
