Bibliometrics: Building an equitable global science system

Robert Maxwell’s body was found floating off Gran Canaria on 6 November 1991. He had disappeared overnight from his luxury motor yacht, the Lady Ghislaine. Amid fevered speculation about the cause of death, attention focused on the huge debts facing his media and business empire.

Two weeks later, the Mirror newspaper, having initially run with the headline ‘The man who saved the Mirror’, revealed that he had stolen £526 million from his Mirror Group of companies: most of this was from the pension fund. Maxwell is now remembered for his ambition, his ego and his fraud. Less recognised is that his sprawling business empire was built on the profits and success of Pergamon Press, the academic publishing venture he began in 1951.

Observers of contemporary global higher education tend to focus on the latest fast-moving developments. However, the commercial landscape of today’s global science communication can be traced back to the policy foresight of Vannevar Bush, the deal-making of Robert Maxwell and the data skills of Eugene Garfield.

One place to start is with the influence of Vannevar Bush on US government policy. Bush was an American engineer, inventor and science administrator, who during World War II helped set up the US Office of Scientific Research and Development that oversaw military R&D.

Bush made the case that basic research was the ‘pacemaker’ that underpins scientific progress that required dedicated funding and support. His tenure and influence were marked by a massive expansion in US science funding, and a few years after the war, the National Science Foundation was launched.

Vannevar was a skilled administrator, but not an entrepreneur. Robert Maxwell was one of the first entrepreneurs to realise just how profitable scientific publishing could be. For more than 30 years he convinced numerous editors and societies to sign contracts, rapidly scaling Pergamon’s journal publishing programme from six in 1951 to more than 700 at the company’s peak in the 1980s.

Eugene Garfield, founder of the Institute for Scientific Information, was a post-war entrepreneur of a different sort. Amid a rapid growth in research flows, he was determined to find ways to automate and speed up information searching. His initial experiments included Current Contents, an abstracting service that started as a photocopied pamphlet sent out on airmail paper. Garfield needed money to fund his ideas, and he worked hard to market his products to the commercial sector.

Garfield is best known as the creator of the Science Citation Index (SCI) – now known as the Web of Science. A searchable research index, based on citations rather than keywords, was controversial, technologically demanding and expensive.

His first step was to build a database of all the references in a select group of 600 scientific journals. Launched in 1963, the number of indexed journals doubled in three years. As it grew it began to define ‘reputable’ academic knowledge.

Inclusion mattered for publishers, who were also prepared to pay significant subscription fees. With ever more ‘international’ journals being launched by Pergamon Press and Elsevier, Garfield held a powerful gatekeeping role.

It was not Garfield’s intention for universities, academics and publishers to use the Science Citation Index to compete. But citation data allowed users to score and rank journals based on their citation ‘impact factor’. Unwittingly or not, Garfield had created the tools for academic game-playing and institutional performativity. His index moulded the unequal and stratified research terrain we now inhabit.

Growing pains

According to the Ulrichsweb periodical database, there are now more than 100,000 academic journals published worldwide. This is likely to be an underestimate, as the global community of higher education continues to expand.

Four multinational companies dominate, each publishing more than 2,000 journals each – Springer Nature, Elsevier, Wiley-Blackwell and Taylor and Francis. They are based in London, Amsterdam, Hoboken (New Jersey) and Oxford, but have global profiles. Together, they publish more than 70% of all social science journals, and 50% of journals in the natural sciences. Sage is in fifth place with more than 900 journals.

In a growing global tertiary education sector, new market opportunities constantly emerge. For example, Hindawi was founded in Egypt in 1997. It, like MDPI, launched around the same time, made the most of the internet and the shift to digital to scale up the author-pays open access business model.

Two hundred and thirty journal launches later, it moved to London and was bought by Wiley in 2021 for US$300 million. MDPI has had a similar commercial success, and in 2022 published 235,000 articles, second only to Elsevier.

Meanwhile, the elite journal ‘brands’ have become tradeable marketing tools for their commercial owners. Where there was once one The Lancet, there are now 22 Lancet-branded journals. Springer-Nature’s ‘brand expansion’ strategy has meant there are now more than 30 journals within its portfolio, all with ‘Nature’ in their title.

Nature publishes the very strongest submissions it receives, but the publisher ‘cascades’ rejected articles to other open access Nature-branded journals, many with high publication fees.

Springer Nature has also created their own journal ranking index, and publish an increasing number of branded supplements, special sections and ‘advertorials’. The journal impact factors of several Nature journals have increased by almost 50% over two decades.

The commercial academic publishing model requires consistent growth, whether from publishing more in each issue, soliciting more special issues or launching more journals. At the same time, several long-established – and once-prestigious – scientific journals owned by professional societies have seen submissions and income plummet.

In an unequal global research system, acceleration and productivity become survival strategies. Universities are ever more focused on reputation and international rankings as they compete for students, funding patronage and reputation. Many incentivise their staff to publish through financial incentives and promotion pressures, changing academic practice.

Subsidies and institutional expectations foster an acceleration of the research publication cycle. The logics of reputational stratification across a hierarchical global science system require those at the peripheries (especially precarious junior and adjunct staff) to publish more and faster to stay visible, putting yet more pressure on the system.

In this context, it is not surprising that some academics take short cuts to survive. In China, as in other emerging economies, doctors and other professionals need to have academic publications in ‘top’ journals to get promoted. If they have no research or writing experience, the chances of getting their work into SCI journals is slim. The only option may be to purchase authorship, and there is thus a burgeoning demand for brokers and agents who can help with this process. This has led to a series of high-profile mass retractions.

When academic publishers talk about the importance of integrity and trust, this is because they are acutely aware that academic credibility and reputation are a precious asset. In sensationalist media reporting, publishing fraud gets portrayed as an existential threat to science. While such cases remain rare, the policy community doubles down on research integrity as the only way to solve a seeming crisis in scientific communication.

In response, many publishers are introducing elaborate AI-driven ‘fraud detection’ tools. They also rely on citation benchmarks and metrics indicators to detect journal ‘outliers’. This use of aggregated publication metrics means that journals with more distinctive profiles – whether because of their history, remit or editorial model – risk being seen as potentially fraudulent or fake.

There is less reflection among publishers as to whether the tactics used by those on the margins are just extreme forms of the ‘gaming’ built into a metrics-based system. Few ask if the accusations of fraud are appropriate, given that all the actors within this ecosystem are caught up in an integrity-technology ‘arms race’.

Spurred on by integrity watchdogs and scientific sleuths, the focus is on tracking down misconduct rather than a discussion of the features of mainstream science generating these mimetic practices.

More broadly, the discourse of ‘fraudulent’ science also serves to reassert the boundaries of genuine science, and to shore up the legitimacy and value of its key currency: citation data.

What gets left out?

The unequal geographical representation of scholarly journals by the Science Citation Index was first pointed out more than 50 years ago. Today, thanks to the business models developed by Maxwell and his rivals, the hold of commercial publishers on the global research ecosystem is stronger still.

Despite calls to decolonise open access and promote bibliodiversity, the two commercially owned citation indexes cast a long shadow across academic publishing across the Global South.

Toluwase Asubiaro is a Nigerian information scientist who is documenting the impact on the visibility and status of African academic publishing. With a group of colleagues, he has set up the African Research Visibility Initiative, using their bibliometric skills to evaluate these indices, and to find alternative ways to measure and assess African research, such as the use of Altmetric, Google Scholar or Crossref.

He has co-authored a study which estimates that there are currently around 2,200 active journals published in Sub-Saharan Africa. Of these, only 166 were indexed in the Web of Science (and 174 in Scopus), around 7.5% of such journals.

Of the 166 in Web of Science, around 75% were published from South Africa. This means that only around 50 journals are indexed in Web of Science from across the rest of Sub-Saharan Africa. Many African countries have no journals in the index. Only 21 Nigerian published journals are indexed, four from Ghana, and five each from Ethiopia and Kenya. Very few journals from francophone Africa are indexed. This shows just how much research knowledge is made invisible by these indices.

Is open science the answer or the question?

The modern ‘open science’ movement begins in the early years of the internet. Initiatives such as Project Gutenberg sought to make research publications more widely available, for instance. The 2001 Budapest Open Access Initiative was the first of many initiatives to promote open access publishing.

European research funders took a major step forward in 2018 by supporting Plan S, which committed them to publishing all funded research in open access repositories and journals.

Responding to Plan S mandates, the majority of commercial publishers developed ‘transformative agreements’, shifting from funding through subscriptions to author-pays ‘gold’ open access funded by article processing charges. Most have seen their profits grow.

Far from helping to decolonise the publishing ecosystem, this model of open science embeds the dominance of commercial publishers. At the same time, it marginalises researchers working in resource-deprived research systems and seems to be embedding acceleration and productivism as survival strategies for those on the academic periphery.

Today, open science has become a politicised and contested space. The UNESCO 2021 recommendation envisions research infrastructures that are “organised and financed upon an essentially not-for-profit and long-term vision, which enhance open science practices and guarantee permanent and unrestricted access to all, to the largest extent possible”.

In May 2023, the European Council recommended that European member states “step up support” for the development of a not-for-profit publishing platform free to both authors and readers (so-called ‘diamond’ open access).

Several Horizon Europe funding projects – including DIAMAS and OPERAS – are working to build a high quality, sustainable and community-owned scholarly communication system, including institution-funded technological infrastructures and common standards for open access scholarly journals.

It is an ambitious vision that aims to redirect funder resources away from commercial publishers. The ideas are still embryonic. Scaling up a community-owned open source publishing infrastructure would need unprecedented political will and deep pockets, given that Elsevier alone spends billions each year on developing its communication technologies.

If such a model is to work beyond well-resourced European universities, governments across the world will need to step up their funding for national and regional research and publishing infrastructures.

The underlying rationale is that equitable and sustainable community ownership promotes ‘bibliodiversity’: a diverse range of regional initiatives, publishing infrastructures, databases and knowledge ecosystems that support the needs and epistemic pluralism of different research communities.

Latin America offers an exemplar of a strong regional Portuguese and Spanish publishing research ecosystem supported by the community-owned SciELO database and publishing infrastructures.

There are a growing number of ‘diamond’ open access publishing platforms and experiments from which to learn, and much policy interest in sustainable open science. It is an idealistic vision of a more equitable research world in which, as Arturo Escobar argues in Pluriversal Politics, “many worlds might fit”.

What comes after the citation economy?

Sixty years after Garfield launched his first citation index, and more than 70 since Maxwell founded Pergamon, academic journal publishing has been transformed into a profitable global industry. Commercially intertwined, the indexes and the publishers have together built a citation economy.

Today, scholarly reputation and status are measured by journal rankings, ‘impact factors’ and ‘h-indexes’. The reach of these citation indexes and their data has been amplified by digitisation, computing power and financial investment.

Citation metrics reinforce existing academic ‘credibility economies’, built around Euro-American publishing networks and commercial interests. Across the majority non-Anglophone world, journals excluded from Scopus and Web of Science face constant challenges to their legitimacy and reputation.

The pressures on scholars at the margins have led to academic acceleration and research productivism. This in turn provokes media concern about scientific fraud and an endless integrity-technology ‘arms race’, rather than a discussion about the financial and societal unsustainability of a commercially owned research economy.

Is there a way out of this recursive growth loop? Open access advocates, funders and researchers offer a vision of a more equitable global research system built around community-owned publishing infrastructures and standards. The first step on this journey is weaning scholars and universities off the addictive measures of scientific credibility sold by commercially owned citation indexes.

David Mills is associate professor (pedagogy and the social sciences) at the department of education at the University of Oxford, United Kingdom. He is deputy director of the Oxford Centre for Global Higher Education (CGHE) and a co-investigator on CGHE’s research programme on supranational higher education. This is an edited version of his keynote speech at the 2023 CGHE annual conference.