In 2011, the European Research Area (ERA) outlined several principles of innovative doctoral training such as research excellence; exposure to industry and other relevant employment sectors; transferable skills training; and quality assurance (Vittorio, 2015). Similar policies and guidelines can also be found in doctoral education across the globe, where the request for innovation and societal collaboration is urgent (e.g. Association of American Universities, 1998/2017; Australian Council of Learned Academies, 2016; HEQSF, 2013; Swedish Government Bill, 2008, 2016). At the same time, it is known that the general conditions for innovation vary across countries (Meeus & Edquist, 2006), and that global policy trends are construed and organised differently at national level in doctoral education (Andres et al., 2015). Furthermore, studies have shown that doctoral students’ inter-sectorial work could be hindered by the fact that the universities are not always near to knowledge-intensive industries, or that the industry is ill prepared to make use of the doctoral students’ qualifications (Vittorio, 2015). Combined with the fact that gender differences have been found in a number studies on doctoral education in general (Jones, 2013), it is accordingly significant to ask: What are the contextual and gendered conditions for doctoral students to develop their innovative and collaborative capability? Current research has no satisfying answer to this complex question yet. Against this background, our conference contribution is founded in a newly started project entitled "Developing innovative and collaborative capability in doctoral education from a gender perspective: Conditions, processes and outcomes in Sweden and South Africa". Based on social rule theory (Burns & Carson, 2002), we assume that learning practices are governed by a range of social rules founded in policies, organisational leadership, scholarly norms, and societal stakeholders. In particular, we are interested in how diverse levels of the doctoral educational system are related (or not) to each other, and how different systems affect the students’ innovative and collaborative development. While our overall project consists of several self-contained but interrelated studies to fulfil this holistic and systems theoretical approach, the first part of our study, reported on here, focuses on the macro level only by analysing national policies and guidelines related to doctoral education. Hence, based on summative content analysis (Hsieh & Shannon, 2005), this conference contribution will illuminate how the expressions of innovation, societal collaboration, gender and related concepts have occurred, converged and developed over time in Swedish and South African national policies on doctoral education from 1970-2017. Certain attention will be given to differences and similarities in these regards when comparing the two nations. Comparing Swedish and South African doctoral education is well justified. In contrast to Sweden, with its long traditions of societal welfare and of producing doctorates, South Africa is now in a phase of significant expansion and construction of doctoral education – with the political aim to strengthen the economy and democracy of their nation (National Planning Commission, 2011). Due to these national differences, our project can contribute to a deeper understanding of both context-specific and global issues within the problem field.