A move to full "gold" open-access publishing will "cripple university systems" by incurring large extra costs without significantly improving access to research, leading research universities have warned.
The government-convened Working Group on Expanding Access to Published Research Findings last week said that adopting the "gold" system - under which authors pay fees to make their articles open access - would be the best way to increase access to publicly funded UK research.
But the group, chaired by former Keele University vice-chancellor Dame Janet Finch, estimates that the transition could cost UK higher education an extra £60 million a year. This is because expenditure on article fees is likely to increase faster than subscription charges fall, particularly if the rest of the world does not follow the UK's lead.
It suggests that universities meet article fees using dedicated finance provided by research funders, plus "other available resources".
The research councils are expected to confirm soon that they will make it easier for universities to bill them for article charges relating to research they have funded.
Meanwhile, the Higher Education Funding Council for England, which is expected to require all submissions to future research excellence frameworks to be open access, has suggested institutions could use some of its funding to cover article charges "at a reasonable level".
But David Price, vice-provost for research at University College London, expressed surprise that the Finch report does not propose any regulation of article fees, without which "publisher profits will continue to be high at the expense of the public purse".
He endorsed the Finch group's desire for increased access, but said its recommendations would "cripple university systems with extra expense", which would require other activities to be scaled back.
"Finch is certainly a cure to the problem of access, but is it not a cure which is actually worse than the disease?" he asked.
He said financial modelling by the UK Open Access Implementation Group for a forthcoming report indicated that the most cost-effective route to open access for research-intensive universities was the so-called "green" model, under which authors self-archive their published papers in open-access repositories - usually after an embargo period.
Green open access was also more in keeping with the access requirements of the European Union, and Professor Price suggested that until it became "embedded", the UK should follow other European countries in negotiating national site licences for journals.
"[The] Finch [group] claims this would be too expensive. But where is the financial modelling to prove this?" he asked.
His concerns were echoed by Ian Walmsley, pro vice-chancellor for research at the University of Oxford.
Professor Walmsley said that in the worst-case scenario, full gold open access could see the University of Oxford's expenditure on publishing rise by a "staggering" 350 per cent.
Costs for top universities would be exacerbated by the particularly high article fees charged by the high-prestige journals in which their researchers typically publish, he added.
"The increased costs accruing to UK researchers will likely have to be borne at the expense of research itself, but the cost-benefit ratio of this has not been assessed," he said.
Times Higher Education understands that similar concerns are shared by other research-intensive universities. One pro vice-chancellor for research, who did not want to be named, said that although the Finch group contained publishers, funders, librarians and academics, there was a perception among his colleagues that its conclusions had been "overly dominated by the views of publishers", many of whom view green open access as a threat to their business models.
But Professor Walmsley said the professional society publishers he had worked with considered green open access with a 12-month embargo to be "perfectly acceptable". He was unconvinced that universal gold open access would be a significant advance.
"We are going to add considerable cost to providing a route to access that is only a marginal improvement on what already exists," he said.