Social Transformation in South African Science and Innovation – STIS

The project uses a combination of methods. To a great extent our empirical work is based on data acquired from the National Research Foundation in South Africa. These are data from their "rating system" and comprise in essence the CVs of (almost) all research active university faculty in the country. These data are complemented by more aggregated data regarding the racial and gender composition of faculty and graduates in South Africa (supplied by our South African partner. Considerable effort has been put into cleaning these data and making them compatible with standard data structures.

We also make use of computer simulation. One of the papers is a simulation model of how university departments evolve as they hire PhD graduates from their own and other departments. Exploring this with regard to multiple policy objectives and multiple policy levers involved construction of a detailed model which was calibrated to available data on racial composition of graduates and departments.

The main results are captured in chapters of PhD theses and in published articles.

Papers published in refereed journals

Cowan, R., M. Müller, A. Kirman and H. Barnard, “Overcoming a legacy of racial discrimination: Competing policy goals in South African academia” Socio-economic Review, 22(3): 1413–1449, 2024. doi.org/10.1093/ser/mwad043

Rossello, Giulia, Robin Cowan and Jacques Mairesse, “Ph.D. Publication Productivity: The Role of Gender and Race in Supervision in South Africa” Journal of Productivity Analysis, 61, 215–227 (2024). doi.org/10.1007/s11123-023-00681-4

Müller, M., R. Cowan and H. Barnard, “International scientific collaborations as a source of social capital: Peer effects in developing country scholars’ establishment of foreign ties” Industrial and Corporate Change, Vol.32(5):1077–1108, 2023. doi.org/10.1093/icc/dtad043

Mario González-Sauri, Giulia Rossello,(2022) “The Role of Early-Career University Prestige Stratification on the Future Academic Performance of Scholars” Research in Higher Education 64(4) DOI:10.1007/s11162-022-09679-7

Working papers

Rossello, Giulia, Robin Cowan and Jacques Mairesse, “Ph.D. Publication Productivity: The Role of Gender and Race in Supervision in South Africa” National Bureau of Economic Research working paper 31346, June 2023, www.nber.org/papers/w31346

Cowan, Robin, Mortiz Müller, Alan Kirman & Helena Barnard, 2021, Overcoming a legacy of racial discrimination: Competing policy goals in South African academia, UNU-MERIT Working Paper 2021-040 | Abstract

Rossello, Giulia, Robin Cowan & Jacques Mairesse, 2020, Ph.D. research output in STEM: the role of gender and race in supervision, UNU-MERIT Working Paper 2020-021 | Abstract

Rossello, Giulia & Robin Cowan, 2019, Far from random? The role of homophily in student supervision, UNU-MERIT Working Paper 2019-024 | Abstract

PhD thesis

Giulia Rossello “Social Transformations and Labour Market Entry: An Investigation into University Systems in Emerging Economies”, Ph.D. thesis, Maastricht University, 2021.

Lorena Rivera-Leon “Unveiling the Determinants of Scientific Productivity in Middle-Income Countries: An Economics of Science Perspective”, Ph.D. thesis, Maastricht University, 2021.

The project has opened the door to several avenues of future research. One observation made through the project is that different groups (by gender or race) tend to have different academic interests, that is, they focus on different topics. As the system transforms, the racial and gender mix will change. Does this imply a fundamental shift in the (aggregate) topics addressed by South African researchers? If it does, what does this imply for the relations between academia (and academic research and training in particular) and the rest of society? This is an agenda that we have begun to take up, assessing the relationships between "science" (or more properly "academic research", policy (seen through the lens of government interests) and society more generally. Work in this area so far has tended to look for one domain citing another in its writings. We believe that there are better approaches using modern text analysis of the documents themselves rather than just their references.

Achieving a balanced representation of racial and other specific social groups in desirable occu- pations is important both politically and socially. Failing to tap the talent of large parts of the population to the benefit of society is limiting economic growth in terms of human capital. This is true in many different social and economic milieux, but it becomes particularly compelling in the university systems of knowledge-based economies. Participation in the education system is now a precondition for equal opportunity in all other spheres of life, and the science system is an increas- ingly important pillar of the innovation system driving economic growth. This project investigates the integration of formerly excluded groups into socially attractive occupations and positions, with an empirical focus on the racial transformation of the South African science system. We address this issue from two perspectives: the processes by which this change happens; and the consequences of the transformation on the structure and content of the research system.

A large literature argues that successful entry into a profession depends on features of a person’s social network. In contrast to much of the literature, we treat this issue using multiplex network structures in which we do not collapse all the different types of relations individuals might have (student-supervisor, co-author, citation, co-location) into a single network but treat them as dif- ferent, interacting networks. Further, we have access to a unique dataset from the South African National Research Foundation, containing rich biographical information on almost all research- active academics in the country. Thus our empirical analysis not only observes different types of links between academics, it also takes account of the role of individual characteristics (age, race, gender, education, discipline . . . ). We do a much more nuanced empirical analysis than has hith- erto been possible. In examining this social transformation we ask: does transformation move from layer to layer; is there a typical pattern for a single individual; does one layer transform before an- other. We treat the work structure as just another layer in the multiplex network of social relations wherein two people have a link if they share an employer. Integration at the institutional level is thus not a simple dependent variable depending on labour market dynamics, but rather just one of several layers interacting as a complex system. This conceptualization permits a more complete empirical analysis which in turn informs our modelling. In the latter we build an agent-based model of the academic system as it undergoes this social transformation. The model includes different types of relationships among academics (a multiplex network), and examines how evolution in one layer interacts with evolution in another. Racial integration can be considered in any layer, even if the policy discourse has centred on the workplace or institutional layer. Calibrating the model to the SA case permits us to examine carefully which factors could be valuable policy levers to speed the process. The last part of the project begins from the observation that currently black and white academics gravitate towards different academic disciplines. If this disciplinary “preference” remains unchanged, this implies that social transformation of the science system will be accompanied by a transformation of the type of science done, with further implications for the entire innovation system. We examine this hypothesis using the model we develop earlier, to ask whether social transformation will be accompanied by transformation of the science system itself.