bioinformatics, genomes, biology etc. "I don't mean to sound angry and cynical, but I am, so that's how it comes across"

The UK’s funding for open-access publications is not enough

The clear answer is “no”, and of course it was always “no” from the very beginning.  But let’s explore a few ideas below, just to show you how far off the funding actually is.

I want to start off by saying, categorically, that this is not the fault of my employer, the University of Edinburgh.  They are doing all that they can.  The fault lies with the UK government and their refusal to provide sufficient funds.

The basics

The public pay for the vast majority of academic research, through taxes that are distributed by the government through its research councils.  It stands to reason, then, that the public should be able to read the outputs of that research free of charge.  I’ve never heard a cogent argument against this, and I don’t believe I ever will.  It’s blindingly obvious.

How it used to work

The research councils have a great system, called Je-S, which stands for “Joint Electronic Submissions”, and it means there is a single grant application submission system for all of the UK’s research councils.  I have had a few BBSRC grants, so here is a screenshot from one awarded in 2011:

jesI have removed the costs for the other items as they are irrelevant, but as you can see, we were able to request funds for open-access publications, and I asked for £3k to fund 4 open access charges.  Whether this is enough or not is irrelevant – I could have asked for £10k – the point is I was able to ask.  And we got it.

I ask that you remember this single point – in the previous system, we were allowed to request funds for open-access publications, the applicant set the rate and the money was awarded.

How it works now

Applicants are now not allowed to apply for OA funds as part of a grant application.  Instead, research councils UK (RCUK) will provide centralised funds to Universities and then Universities will provide these to researchers wishing to publish OA.

So how much have they provided?  According to this article:

RCUK has earmarked £17 million to create open-access publication funds for more than 100 higher education institutes in 2103-14

So.  £17m across 100 higher-education institutes.  £17M sounds a lot, but actually, simple maths tells us this amounts to an average of £170,000 per institute.

So we’ve gone from a system whereby any PI applying for RCUK funds can request, and have awarded, any amount of money for open-access charges, to a system where each HE institute has a set budget.

And this is progress?

The maths

The University of Edinburgh, one of the 100 HE institutes the £17m is for, my employer, whom I’m not criticising, has 22 schools spread across 3 colleges.  I know most about science, so perhaps we want to limit this to the 14 or so schools that are “scientific”.  Let’s take the school I’m employed by, the Royal (Dick) School of Veterinary Studies.  What we can do is go to to Edinburgh’s Research Explorer and search for research outputs.   We can filter by school and by year and we can see that, in 2013, my school published 547 peer-reviewed papers.

The cost of an open-access publication has been much debated, but let’s take two figures: £1.5k which is the lower bound from the Finch report, and £839 which is approximately how much if costs to publish in PLOS ONE.

The estimated cost of making all 547 papers from the R(D)SVS open access is therefore between £460,000 and £820,000.

Now, let’s be pessimistic.  Let’s assume that the R(D)SVS is special, and that other schools only produce half of what R(D)SVS does in a year.  So the cost per school per year is between £230,000 and £410,000.

That means across the 14 schools, the cost is between £3.2M and £5.7M.  Remember, RCUK has committed £17M to 100 institutions.  Remember also that I have assumed that the normal output per school is 50% of the R(D)SVS, so these are conservative estimates.

Let’s make another conservative assumption.  Let’s assume that the average output of the other 99 institutions is 50% that of the University of Edinburgh.  So let’s assume the average cost per institution is in fact between £1.6M and £2.8M.

Across 100 institutions, then the ultra-conservative estimate of the cost of OA is between £160M and £280M per annum.

What we have is £17M, which is about 5-10% of the estimated cost.  This agrees very well with the Russell Group’s estimate that the government money will cover about 10% of publications.  A figure which the government disagreed with.  I know who I believe.

Announcing a budget cut

Let’s call the £17M what it is – it’s a budget cut.  Recall that PI’s used to be able to apply for any amount of funding for OA publication charges as part of their grant application.  This therefore meant there was effectively an unlimited fund for OA.  Now there is a limit, and that limit is £17M.  That’s a budget cut.  The UK government didn’t invest or provide funds for OA publications, it cut the available budget.  We should call it what it is.

Note that RCUK funds £3bn worth of research at UK universities annually.  The £17M they have promised for OA is 0.5% of their budget.  The required amount, our estimate of £280M, is about 9% of their budget.

Admin, admin, oh God the admin!

When using JeS to apply for OA funds, there were no additional admin costs – the grant gets peer-reviewed and funded or not-funded.  The same is true whether or not OA charges are included in the application.

However, now we have 100 institutions all administering their own OA fund.  Let’s assume each institution needs a person to do this, and that that person gets paid £25,000 per annum.  That’s £2.5M on admin.  That’s 14% of the entire budget spent on administering the fund.

Arggghhh!!!!!  So now we have a budget cut coupled with an additional administrative burden!

Gold vs green

Of course what I have been talking about is Gold open access, and there is another way – Green open access.  In Gold OA, we pay the journal to make the article free to everyone.  In Green OA, we stick a copy of the paper in our institutional archive and make it available that way.

Here are the steps for Joe Public to find and read a paper under Gold and Green OA:

  1. Gold OA. Read news article.  Find research article.  Read it, because it’s free
  2. Green OA. Read news article.  Find research article.  Realise it costs money.  Probably give up.  However, for the tiny percentage who don’t give up – figure out lead university.  Find lead university’s repository of articles.  Find article.  Request article.  Read article.

I think it’s fairly obvious which we should aim for, isn’t it?  Isn’t it?

Obfuscating access to research

Cameron Neylon has done a better job of it, but the drive to make research outputs available in libraries is a ruse, and should be called “obfuscating access to research”.   It looks to me like someone in Aberdeen would have to travel to Newcastle (over 250 miles) to find a library where they could read a paper.

Need I say more?

In summary

The £17M provided by the UK government for OA charges is nowhere near enough, and is in fact a budget cut.  Gold has to be the way forward as green will simply discourage the public from reading about science, and we don’t want that to happen.  Attempts such as Access to Research are pretty poor, and nowhere near what’s required.


  1. Interestingly, I ran a similar calculation for the university I work for the other day, so I am pleased to see that I am not the only one looking into this. However, while I agree with taking an interest in this is good I am not sure I agree with your methodology.

    The RCUK policy and budget only applies to RCUK-funded research. At the institution I work for it is about 40%. Secondly RCUK do neither require nor do they expect all publications to be made OA – currently we are talking about roughly 50%. They are also not saying that these articles all have to be gold OA. They do have a preference for gold, so let’s be generous and say about 50% of the 50% should be gold. That means that instead of calculating with 100% of the research output you should be looking at something more like 10% of all peer-reviewed articles – so turns out your numbers are not conservative at all but rather really high.

    I also don’t think you can take the output of a research intensive university like Edinburgh and generalise those across the UK, even with a 50% reduction – the funders know this, so some of the big research intensive universities get well over a million per year from the overall budget. If you go with the average cost of £1,700 per article then the £17M you quote would pay for 10,000 articles – and with my percentage of 40% of research having some funding from RCUK being correct that would mean we are looking at 10K out of 100K peer-reviewed articles in one year across the UK. That’s not too far off the lower figure you cite.

    You should also take into account that other funders, such as the EU and Wellcome, also make money available for OA publishing. And you should take into account that the discussions on reducing library subscriptions in line with increasing gold OA costs have started.

    So what I am saying is you need to look at all these numbers, across the board, and then calculate the current cost vs the one in the new system; that also means taking into account management costs for the current system including library subscriptions. It would also be nice to consider whether with OA publishing academic productivity will go up (faster, easier access to papers) and what this will do to the economy overall.

    Now, I am not saying that your overall conclusion that there is not enough money available for gold is wrong – but that we need solid numbers if we want to make that claim. Either way it is good to hear that academics are considering these questions now and that the debate broadens.

  2. Thanks Torsten, I wouldn’t be surprised if my numbers are wrong, but I guess we are coming from different angles.

    You’re starting from what RCUK are trying to achieve, and I am starting from what I think they should aim to achieve. So the fact that RCUK are aiming for 50% OA articles is, well, irrelevant – I don’t think they should aim for 50%, they should aim for 100%. I also think it should be 100% gold, not 50% gold. Green OA reminds me of The Hitchhiker’s Guide to the Galaxy where the Vogon plans are on display at Alpha Centauri – sure, you can get access, but it ain’t easy.

    If Edinburgh is a research-intensive university, I can tell you categorically that we got less than £1m per year 🙂 So from your account, there must be universities that are more research-intensive – therefore perhaps my 50% is not so far off?

    And finally, my numbers may not be rigorous, but I find it comforting that they agreed with the 10% figure that the Russell group came up with. So I’m pretty happy with my conclusions!

  3. Well, I work for a seriously research intensive university and I ran the numbers and concluded that we can meet the requirements of RCUK, at least for this year and probably also for next. The staff effort to do so, well, that is a different story, but in terms of what they want to achieve it seems feasible for the immediate future. Long term however, it can only work if the subscription budget comes down – that is what you should figure into your numbers. So ideally you’d get the overall number of peer reviewed articles published by Edinburgh (we had 8K last year), multiply by average APC and then compare that to the RCUK budget and the library’s subscription budget. If all comes to about zero then we know if is feasible – if the publishers reduce the subscription budget accordingly. That, well, may be a different story.

    Anyway, it is great to see that more people are now thinking about the consequences to the whole system, not just their personal workflow!

  4. At Edinburgh, certainly in the meeting I was in, there was an acceptance the fund would run out each year and then it was up to PIs to find alternatives. And yes this was just RCUK outputs.

    I don’t think this is the way we want to run OA funding is it?

  5. The situation will be different at different universities, but you are right to say that we should have a model where every piece of research that has been peer-reviewed can be published where it is most suitable. The current system is very unbalanced.
    This is from 2012 but may be of interest: http://www.ariadne.ac.uk/issue70/andrew

Leave a Reply

© 2018 Opiniomics

Theme by Anders NorenUp ↑