British universities are producing a greater depth of world-leading research than previously realised, the results of a large exercise examining the output and real-world impact of 76,000 academics in the UK suggests.

The findings of the research excellence framework (REF), looking at work produced from 2014–2021, is based on rating nearly 186,000 pieces of academic research, with the results used to divide up about £2bn in annual government funding.

According to experts, the results show that Oxford and University College London are likely to get the largest share of government research grants based on the high proportion of their departments rated as “world leading”.

But the overall results showed a small decline in the proportion of top-class research by the “golden triangle” of Cambridge, Oxford and London universities, and showed larger pockets of high-quality research at smaller institutions.

The University of Northumbria was one of the biggest winners, with its research output in areas allied to health, such as nursing and engineering, receiving high proportions of top ratings.

Andrew Wathey, Northumbria’s vice-chancellor, said his university’s increase from 52nd to 28th in market share of future funding “moves us clearly into territory formerly the preserve of the Russell Group” of universities as a global force in research. “Northumbria is the first modern university to cross the clear blue water that separated the old and the new parts of the sector, and others are following,” Wathey said.

The two-year-long peer review found that 41% of research submitted was worthy of a “world leading” rating of four stars, while 43% received three stars as “internationally excellent”. The 2014 REF rated 22% of research as world leading but senior leaders said the increase was the result of improved submissions and changes in how the exercise was run.

Reforms since 2014 forced universities to include all academic staff with significant research involvement in their submissions. But it also them gave greater flexibility in how many pieces of research could be submitted for each member, including for staff whose careers had been disrupted during the period.

For the 2014 REF, about 52,000 academics submitted 190,000 pieces of work, while in 2021 76,000 staff submitted 185,600 pieces of research, including books and journal papers.

David Sweeney, the executive chair of Research England, said the reforms had made it impossible to compare the results. “I think we are seeing that research in the UK continues to be very, very strong, and perhaps a little bit stronger. But this is about a different measure,” he said.

Prof Dinah Birch, who chaired the REF’s arts and humanities review panel, said the differences were far-reaching. “This is a different exercise and lining up the REF 2021 results against those of 2014 will be misleading,” she said.

Birch said a “dizzying variety” of research was submitted. “Everything from the fiction of Anthony Trollope, to the nature of the particles that make up the substance of the universe, to the identification of new materials for manufacturing innovation,” she said.

Rankings calculated by Research Professional News, a specialist publication that tracks funding, found that Oxford’s “market share” of funding was likely to fall from 6.24% in 2014, after the previous REF, to 5.7%.

The rankings also showed that Lancaster and Loughborough universities had better results than two members of the Russell Group of research-intensive universities, with Lancaster above Queen’s University Belfast and the London School of Economics.

Sarah Richardson, editor of Research Professional News, said the REF outcomes were crucial in giving universities access to longer-term, more flexible streams of “quality-related” funding. “It’s crucial for universities and researchers because it pays for their ongoing costs, like salaries and early-stage seed funding for projects, and it helps them to plan for and invest in research capacity, rather than just financing research project by project,” Richardson said.

Jo Grady, general secretary of the University and College Union, said the REF was a drain on staff time and resources, describing it as “emblematic of a research culture obsessed with arbitrarily designating institutions or departments as winners or losers”.