This summer the Research Councils U.K. adopted a stronger open access policy, as explained by Peter Suber in the September 2012 SPARC Open Access Newsletter. While the strong support for open access, including funding for open access publishing, is welcome news, this policy includes some ill-thought-out provisions and open access advocates are calling for revisions before the policy is implemented; see Peter's article for some objections and recommendations.
In brief, the reason this is a bad policy is because it requires researchers to select a gold open access option when one is available, and provides block funding to UK universities to pay article processing fees. This gives publishers a strong incentive to backtrack on green self-archiving policies, adding to the delay or embargo period or removing this option altogether. By making this requirement and providing funding, this is in effect a "blank cheque" policy which is certain to raises the costs of scholarly publishing.
This is why: if you had a business and customers had to buy what you sold regardless of the cost, how might this impact your pricing policy? What if you're a corporation and legally bound to provide shareholders with the best profit returns that you can? This, from my perspective, is an example of a government just throwing money at a problem without thinking it through - very out of character for the current UK government. If they have cash to spare, for heaven's sakes why do they not use it to subsidize students rather than publishers?
Others have made similar points. The main reason for this post is to ask open access publishers involved in lobbying for this whether they are shooting themselves in the foot, and whether it might be in their own best interests in the long run to join open access advocates in calling for improvements to the RCUK policy before implementation.
Why? The primary reason is that this would be better for open access.
In case any OA publishers are finding it difficult to put the unprecedented public good that is open access at the top of their priority list, they should not that in the medium to long term, changes to this policy are in their own best interests, too.
The vast majority of funding for scholarly journals at present - percentages range from 68 - 90% (see my draft dissertation for details) comes from academic library budgets. The UK is a major sponsor of research, but even so only 6% of the world's scholarly research comes from the UK. If the UK goes ahead with this obviously unsustainable approach to supporting OA publishing, OA publishers should be aware that this is highly likely to result in a drop in support from academic libraries around the world. For example, price inflation to fit this exceptional UK market will likely result in a drop in support for article processing fees by libraries around the world - a relatively new trend that has the potential to grow, but is likely to be nipped in the bud if this policy is not fixed. Also, if funding is diverted from research budgets to open access article processing fees, OA publishers should expect well-deserved backlash from scholars and universities. I'll be on their side; my draft thesis is called Freedom for Scholarship in the Internet Age, not Give Money to OA Publishers. Cash from the RCUK for article processing fees might seem like a really good thing right now, but a portion of 6% of the revenue from the world's scholarly publishing is not a good reason to jeopardize transitioning the 68-90% from subscriptions to OA publishing.
To conclude: I recommend that open access publishers working with the article processing fee approach join the rest of the open access movement in calling for the RCUK to fix the flaws in their open access policy before implementation, to remove the blank cheque that forces scholars and universities to pay for OA. Perhaps, as a long-time open access advocacy leader, not-for-profit publisher and open access advocacy organization, Public Library of Science should take the lead in calling for this change. An open letter to this effect posted prominently on the PLoS website would be a welcome development.