LEADER

Inequality and the challenge of how to know

The recently released 2026 World Inequality Report¹ provides important and useful information for all scientists and academics, across all disciplines. This third iteration of the report (the previous ones were in 2018 and 2022) shows that income inequality globally is persistent and increasing: the top 0.001% (about 56 000 adults globally) have about three times more wealth than half of the entire world population as a whole.^11(p.14) Income and wealth inequality are intertwined closely with other factors. For example, the wealthiest account for a substantially disproportionate share of global emissions^1(p.16), but are least vulnerable to the effects of climate shocks and calamities. Income distribution is highly gendered, with women, wherever they are in the world, and by various metrics, persistently receiving lower labour income than men in the same regions. Gender income inequality is less extreme in wealthier countries, but in sub-Saharan Africa, for example, the female labour income share stands at about 28%. There is no region in the world showing gender income parity.

South Africa is, as always, a troubling example. In a recently published analysis² accompanying the World Inequality Report, Czajka and Gethin use data from 1993 to 2019 to analyse racial inequality and redistribution in post-apartheid South Africa. They note that, as of the 2019 data, South Africa remained the most unequal country in the world, with a Gini coefficient of pretax income at 0.8. To a small degree, racial pretax inequality declined, these authors suggest, but largely through the very rapid growth of black "very high earners". In 2019, they calculate, the average per capita income for white South Africans was similar to the average per capita national income of Denmark; the average income in the top 10% of black South Africans comparable to the average income in Italy; and the average income of the bottom 90% of black South Africans similar to that of Zimbabwe.

Although there are many debates on how best to measure inequality and its impact on human and planetary well-being, there is little question that inequality is closely related to environmental degradation, gender inequality, poor healthcare outcomes, problems with educational achievement, rates of violence, poor infrastructure, and a host of other vulnerabilities.³ Scientists committed, as our Journal is, to contributing in a range of ways to human well-being and flourishing, must consider issues of inequality and lack of access as these affect different subject fields, and also the research enterprise as a whole. Regardless of what is being studied, and the methods used, questions of inequality must always be considered. As we in the South African research community explore, grapple with, and learn more about artificial intelligence (AI), for example, apart from the legitimate concerns globally about its environmental costs, Wingfield and Wingfield" noted in our pages earlier this year the intertwining of questions about open access and paywalls, with access to and appropriate use of AI.

Many of us who are academics and journal editors have likely been exasperated by the obvious and very clumsy ways in which AI can be and is being used by students and authors at all levels. For example, as an editor myself of another journal (based in the Global North), I dealt with the first case this year in which there appeared to be AI-hallucinated references, and when I asked the authors to respond to my concern, they did not reply. At the South African Journal of Science, thus far, we have not knowingly encountered such problems. We are excited by the potential and promise of AI and LLMs more, probably, than we are worried about abuse. But, as McKenna⁵ has noted recently, the worries we may legitimately have about AI and its use in scholarly work in South Africa map onto other worries which are not new. These are worries about inequality, apart from anything else. McKenna raises important questions about an outputs-driven approach to doctoral education in South Africa - questions which can be expanded to broader concerns about science and its purpose in our context.

People enter, or try to enter, the world of science and academia from very different places. And the world they enter, certainly in South African universities currently, despite some nascent encouraging pushback, is a world of metrics, publication counts, rankings and awards. As Morrell and Roos⁶ have noted, metrics, awards and prizes may become incentives to pursue fake credentials. Indeed, the Facebook page for AD Scientific Index (https://www.facebook.com/adscientificindex1/) promises that it is "beyond rankings", offering "transparency", "equality" and "real academic value", while relying for its rankings and scores mostly on Google Scholar (https://www.adscientificindex.com/methodology/). Google Scholar as a tool has many advantages, but it does not pretend to have quality control. There is a serious problem when South African researchers use rankings from AD Scientific Index to make claims as to their academic prominence, and when the use of such rankings is rewarded.

Underlying this problem of reliance on what is easily counted are much more fundamental questions, as McKenna suggests. When McKenna insists that, at the heart of doctoral education is not the question of "outputs" but the question of what knowledge is, she is raising again the issues of inequality in higher education and research. There are deep and continuing inequalities in who is allowed to know what about what subject, who is allowed to be an expert on which topic, and who has the confidence to speak their mind, or even to feel that they have a right to voice their views. Knowledge is contested and relational, and subject to change. All that we know is provisional. To create and sustain environments in which there is space for robust and safe contestation is easier said than done. It is not easy to find the best way to acknowledge and take due account of epistemic exclusion, but this is something we must pursue. And one category of excluded knowledge is knowledge from below. As we strive to build a more just and diverse science system, we need to understand the impact of inequality in all its forms upon who has the best opportunities to learn and debate in a robust and rigorous way. We need different perspectives in order to address different aspects of inequality. Simply counting heads or outputs will not solve the problems. Technologies and tools will not solve the problems, helpful though they are. We hope that our Journal will continue to be a space in which the practice of thinking together, which includes the practice of respectful contestation and disagreement, will be at the fore. For this to work, we are dependent on the contributions of our authors and reviewers. We all hold a mutual responsibility to care for one another; part of this care is holding one another to account.

References

1. Chancel L, Gómez-Carrera R, Moshrif R, Piketty T. World inequality report 2026. Paris: World Inequality Lab; 2025. Available from: https://wir2026.wid.world/www-site/uploads/2025/12/World_Inequality_Report_2026.pdf        [ Links ]

2. Czajka L, Gethin A. Racial inequality and redistribution in post-apartheid South Africa. World Inequality Database News. 2025 December 07 [cited 2025 Dec 25]. Available from: https://wid.world/news-article/racial-inequality-and-redistribution-in-post-apartheid-south-africa/        [ Links ]

3. Wilkinson R, Pickett K. The spirit level: Why inequality is better for everybody. London: Penguin Books; 2009. Available from: https://pure.york.ac.uk/portal/en/publications/the-spirit-level-why-equality-is-better-for-everybody/        [ Links ]

4. Wingfield BD, Wingfield BJ. The hidden cost of open access: Artificial intelligence, paywalls and the risk of knowledge inequity. S Afr J Sci. 2025; 121(11/12), Art. #23053. https://doi.org/10.17159/sajs.2025/23053        [ Links ]

5. McKenna S. Foregrounding doctoral knowledge and knower in the age of Generative Artificial Intelligence. Transform High Educ. 2025;10, Art. #653. https://doi.org/10.4102/the.v10i0.653        [ Links ]

6. Morrell R, Roos N. Distinguishing the genuine from the fake in South African universities: Scholarly awards, books and academic credibility. S Afr J Sci. 2023;119(11/12), Art. #16491. https://doi.org/10.17159/sajs.2023/16491        [ Links ]