Involving universities from the developing world in international collaborations is often seen as a key way that scholarship can have a lasting impact on some of the poorest societies in the world.

But how often are the world¡¯²õ top universities actually working with researchers in such countries? And when they do, is the nature of the relationship fair and set up in such a way that it will benefit the developing world?

According to a ³ÉÈËVRÊÓƵ analysis of data in Elsevier¡¯²õ Scopus database, it is striking that partnerships between the poorest nations and the world¡¯²õ research elite form a very small slice of their international collaborations.

For instance, at nine out of 10 of the top-ranked universities in THE¡¯²õ World University Rankings 2018?(WUR), less than 3 per cent of cross-border research featured a partner from nations . At four of the universities in the top 10 of the World University Rankings, this share was lower than 1 per cent.

³ÉÈËVRÊÓƵ

According to the data, analysed using Elsevier¡¯²õ SciVal tool, there were about 57,000 pieces of cross-border research produced between 2013 and 2017 that featured at least one author from a least-developed country (LDC),?which are mainly made up of nations in Africa but also include some of the poorest countries in Asia such as Laos and Cambodia.

However, out of the world¡¯²õ top 10 universities, only Harvard University (1,775) and the University of Oxford (1,181) were involved in more than 1,000 of those papers.

³ÉÈËVRÊÓƵ

Allowing for the differing volumes of international collaborations among the World University Rankings top 10, Oxford had the greatest share published with the LDCs (3.2 per cent), followed by Harvard (2.9 per cent) and Imperial College London (2.2 per cent). Every other university had a share under 2 per cent.

Much of the variation can be explained by collaborations with the LDCs being heavily skewed towards medicine, which makes up 42 per cent of all the LDC collaborative research. As a result, universities with a lot of medical research tend to have the most LDC collaboration.

But in a way this only highlights how, away from medicine, collaboration involving the LDCs is at a very low level.

In engineering, for example, there were about 4,500 internationally co-authored papers involving the LDCs but none of the World University Rankings?top 10 appear among those institutions collaborating the most with the group. In terms of nations co-authoring the most engineering papers with the LDCs, Malaysia was top, while the US was second.

Kevin Marsh, professor of tropical medicine at the University of Oxford and director of the Africa Oxford Initiative, which aims to develop more sustainable and equitable collaborations with academics on the continent, said that medicine dominated both?owing to the importance of global health issues but also the sheer amount of funding available.

There are ¡°many big traditional funders that fund medical research¡± involving the developing world, he pointed out, ¡°but it is important to make sure that we support all areas [such as] climate change, sustainable food production [and] engineering, [which is particularly] core to international development¡±.

Maggie Dallman, vice-president (international) at Imperial College London, who has called on top universities in the West to work more with emerging countries, said that the shift towards other subjects was beginning to happen as more opportunities emerged to apply research to local challenges.

¡°Certainly our engineers are getting very interested in trying to produce technological solutions for local problems in Africa,¡± she said.

³ÉÈËVRÊÓƵ

Professor Dallman added that such applied research might also not always show up in the best journals, despite its potentially having the greatest societal impact.

¡°There is a piece that we should be thinking about in terms of [measuring] impact locally¡­as that is every bit as important, if not more important, as getting a paper in Nature.¡±

For the medical research that makes up the largest chunk of LDC collaborations, there is also the question of how equitable the partnerships are.

In the past few years, the priority for many Western academics and funders has been ¡°shifting the centre of gravity¡± towards the poorer nations themselves to help boost their own research capacity.

For Professor Marsh, who is also a senior adviser to the African Academy of Sciences, ¡°early engagement¡± with research partners in the developing world was crucial ¡°so that everything is co-developed and co-produced¡±. Too often in the past, he added, Western universities would ¡°set up on their own track and then try to engage partners¡±.

³ÉÈËVRÊÓƵ

The ¡°really important shift¡± is to go further and fund developing world researchers directly, so that they make the decisions, he said. ¡°People are very nervous about this due to the issue of financial accountability and so on but that is a really important move.¡±

This is an approach that is already central to the strategy of one of the biggest medical research funders, the Wellcome Trust, which took a decision about?10 years ago to target funding directly at principal investigators in developing countries rather than the other way around.

¡°We are totally agnostic if he or she [then] decides to have a UK, US, Australian or [other national] partner¡­it is totally their decision¡­rather than an assumption that the [Western] institution would make that decision,¡± said Simon Kay, Wellcome¡¯²õ head of international operations and partnerships.

¡°They obviously do pull in expertise,¡± he added, ¡°but the decision and the leadership is always in Africa. Then you almost don¡¯t need a discussion about equitable partnerships.¡±

In addition, Wellcome is also one of the major backers of the Alliance for Accelerating Excellence in Science in Africa (Aesa), an agency set up to lead and manage distribution of research funding in Africa.?

Dr Kay accepted that it was easier for an organisation?such as the Wellcome Trust to take such an approach?owing to its independence and that he was ¡°totally sympathetic¡± as to why taxpayer-funded agencies might have more concerns.

¡°Obviously we audit them [Aesa] but we don¡¯t operate from a point of view of a huge fear of fraud or corruption, our operating point of view is that it makes a huge amount of strategic sense and we¡¯ll manage the risk that¡¯²õ around that,¡± he said.

But if the approach works, it could mean that the networks of research collaboration for the world¡¯²õ poorest nations are, in the future, firmly rooted in the developing countries themselves.

¡°As far as possible, we are shifting the decision-making and ownership as close as possible to where it¡¯²õ needed,¡± Dr Kay said.

simon.baker@timeshighereducation.com????

For the biggest impact, work with nations?at every economic level, says study

International collaboration that involves academics from countries at every level of economic development can achieve the best citation impact, the latest research suggests.

Xinying An and Ping Ni, from the Chinese Academy of Medical Sciences, used the Web of Science bibliometric database to look at internationally co-authored research in public, environmental and occupational health.

They defined collaborators as being from either high-, middle- or low-income countries and then categorised papers into subsets depending on the make-up of each collaboration.

Their results, , suggested that ¡°citations of international collaboration papers are signi?cantly different according to different international collaboration types¡±.

International collaboration involving countries from each income category achieved a citation impact that was ¡°statistically higher¡± than most other types, although papers co-authored only by academics from high-income nations also achieve very high citation rates.

The authors also suggest that the results could be?explained by the fact that issues in public health ¡°are global events or social hot issues, like the Ebola infection, the maternal mortality problem or the malaria epidemiology problem and so on¡±.

³ÉÈËVRÊÓƵ