Jenny says that ‘most people’ reckon the average cost to produce a peer-reviewed article in an online-only journal is around $3,000. Which means that the publishing houses now charging around that amount of money to make your article ‘open access’ are actually offering a pretty good deal.
Now, I know we’ve done this to death (example) but bear with me just a minute. I used to live in a village near Cambridge that happened to house Sigma’s oligonucleotide manufacturing facility. We could call them up, say “Ho good yeoman, I’ll have 40 nanomoles of your finest CATATGTTTCCCCGGAAGGAAAGCGC: bill my grant” and they’d deliver within about 48 hours, and often earlier if you asked them to get in a taxi and bring it to you.
Of course, with an oligonucleotide synthesizer and the right chemicals and a technician and a bit of luck and swearing you could make CATATGTTTCCCCGGAAGGAAAGCGC yourself. But these days you’d be mad to. Sigma ran a service.
Surely that’s what we do with journals? Rather than organizing reviewers ourselves, typesetting and marketing and everything else, we pay PNAS or Nature or whoever to do it all for us. We produce knowledge, and we contract out one part of that process.
Absent a Star Wars-style economic model, any business that provides services and materiel to scientists is going to attempt to run at a profit. There is, in fact, a moral obligation to do so.
It is my opinion that we should recognize this as soon as possible. The current model is a mess, because journals tend to charge the end-consumer for the product, which seems to be deeply unsatisfactory for a number of reasons that I’m not going to go into here (and yes, here the analogy with Sigma breaks down. I can’t think of a better one right now — actually, try this: in the UK we pay taxes that cover such things as police, dustbin lorries, schools, hospitals and the armed services. Imagine calling out your local bobby because some hoon has nicked the Sex Pistols plaque from Cromer National Park, and after apprehending the miscreants Sergeant Dixon turns round and says “Right guv, that’ll be £342.56p”. It breaks the model).
Now, Jenny rightly points out that many journals don’t charge the author: but some do, some charge both ends, and some (*cough* Elsevier cough) charge exorbitant subscription fees and rely on the goodwill of the scientific community for editing (and have broken web access to boot).
However, let’s pretend that all libraries simultaneously said “We’re not going to subscribe to any journals anymore. You should all charge the authors” — what would happen? Would it work? If not why not? Who would be upset? Would Nature still have News & Views and Futures (which, after all, are the main reasons I read it)?
Whatever happens, someone’s going to feel hard done by. Perhaps people might argue that scientists from poorer countries, although they’ll suddenly have all the access they can eat, can’t afford to publish. But subsidy mechanisms already exist under the current model so I don’t see that there’d be a huge problem. Maybe we’d have to redefine the “smallest publishable unit”? If it was suddenly expensive to publish would we see more thoughtful and thorough papers (because believe me when I say there is a lot of crap in the literature). Would, perhaps, only the rich labs publish: in which case maybe the entire edifice would fall (is that a bad thing? I don’t know) and we’d see Universities and Institutes starting up their own journals, organizing peer review etc. (after all, those made redundant from the traditional publishers would need to find jobs somewhere).
Maybe it’s time to stop tinkering with the engine and buy a new car; or, alternatively, shove some more gas in the tank, stop whining and just get on with it.
I refer the (dis-)honourable Gentleman to the answer I gave some weeks ago. And more importantly to the discussion that followed.
The only thing I would add is that that this may be too strong:
I think such a business should attempt not to run at a loss, i.e. to be sustainable (which is what PLoS is aiming for), but I think they only have a moral obligation to run at a profit if their owners bought them to make a profit.
Why should scientists get subsidized goods and services? It’s just a profession like any other. A hospital has to pay for bed-pans, even though it too is a ‘worthy’ cause.
In my opinion, profits bring incentives to do better, to always stay one step ahead. The publishing companies I worked at were always striving to keep up with other company’s innovations, to mature their journals and to add more value to them. We worked hard to do this, and skilled staff had to be hired to effect it — we had to travel to more conferences, offer more honoraria for commissioned review articles, etc. At my previous stint, the journals improved as a result (as measured by increased readership, submissions and citation), and the academic editorial boards and scientist readers seemed happy with the results.
BioMed Central, a fully open-access publisher that does not rely on philanthropy, is making a profit right now, is thriving, and is pretty innovative. It used to charge $300 an article, but now the APCs are up to 4 figures. Not to fleece the customer, but to make the business viable. I’m not saying PLoS isn’t innovative, but it might be a lot more so if it was turning a modest profit and had some breathing space to experiment and spread its wings.
I realize my stance is not a popular one. Again, I think it comes from having worked on both sides of the fence, and the perspective that brings. If I’m happy to pay for my shiny new centrifuge, why not for knowledge, which is even more precious?
Great post! How much of our job should we outcontract and if so, to whom? While this is the current model, I don’t think it’s entirely clear that publishing should be outsourced. Many institutes share expensive equipment for common use. Publishing could easily be done that way. Haven’t made up my mind, yet.
But Richard is definitely thinking in the right direction: the current system is completely broken and we need a new one. Which one should it be? We currently have ~24,000 journals. If some hypothetical bomb eliminated all of them simultaneously and we were faced with starting from scratch, would we re-create all of them? Would we even create 2?
‘Why should scientists get subsidised services?’
To be fair I think we are a subsidised service. The question surely is one of where the subsidies should be applied and how much? And indeed whether outsourcing different parts of that service is cost effective and delivering value to the public (who pay for it after all).
Is the government obligated to provide a profit stream for commercial entities as part of the research they fund? I would say yes, as long as said commercial entities provide a service that is more cost effective than could be provided in house. If the market demonstrably brings better value for funders then that is great. If a moribund market leads to immense government subsidies for publishing houses then we need to worry. I think there are elements of both happening at the moment.
Richard, this is indeed a discussion we had before. With all the obvious advantages, there are at least two problems with the author-pays model:
Author-pays means that the paper author is the customer. Journal editors will therefore have an incentive to publish as many papers as possible, rather than publis papers that are read by as many people as posible (reader-pays). This can sometimes create problems.
If authors pay for accepted (and not submitted) articles as is currently the case, journals with high rejection rates will have a hard time to be self-sustainable.
Richard, you brought up an important point which has (perhaps strangely) gone unnoticed so far.
In economic terms, journals get a free ride, as they have to outsource a major part of their publishing process to the scientific community, but don’t have to pay for it.
Peer review is necessary. We all require it if we want the scientific process to retain its integrity. But if we were to start charging the journals for this service, where would we end up? Would we in turn get editors who were qualified enough to judge whether a review was suitable and pay accordingly – I’m sure some of us have thought that a reviewer has been talking out of their elbow before, not knowing the difference between that and their @rse. Knowledgeable editors would be a good thing in publishing. Paying even more to get articles published might not be.
I’m mainly addressing Bob and Martin here. Yes we have indeed talked about this before. I think it’s wrong to say we got an ‘answer’.
I maintain that the ‘author pays’ (at whatever stage – post-acceptance only or submission charges as well is moot) model of OA is going to be the only way forward in this economic climate, despite the new problems of its own that it will create.
For goodness’ sake gents, we’re scientists. We can work it out.
Cameron hits it on the head with his talk of market forces. Capitalism seems to be working for most of the world (and previous communist strongholds are just that: ‘previous’), so why not the doing of science. Yes, poor labs might be squeezed out, but is that a bad thing? Physics and biology are expensive endeavours in 2008, so maybe we don’t want poor labs to play — unless they have something the big boys don’t?
Another analogy: in the high street of the village I mentioned there was a greengrocer and a butcher, who could not compete with Tesco on price and volume, but who thrived because people appreciated the higher quality (seriously) and the service. Now it’s a little difficult at first to see how that might translate into science, but when I was in Cambridge we competed really quite successfully with the big American labs. We were small, were always well below budget and yet published major papers. It was all about leverage: we found and used certain advantages.
There’s another thing I mentioned that might be of interest — someone queried it in a private email — but this margin is too small to hold my response.
Mike, interesting point:
But if we were to start charging the journals for this service, where would we end up?
Seriously? Paying more for access or submission. Journals are a business, and if they have to spend money they will try to recoup it.
@Martin: Author-pays means that the paper author is the customer. Journal editors will therefore have an incentive to publish as many papers as possible, rather than publish papers that are read by as many people as possible (reader-pays). This can sometimes create problems.
I used to work at BioMed Central, so I can speak about at least one author-pays journal model. The most important factor for viability of a journal these days is impact factor: there actually isn’t a lot of incentive to publish crap papers, because you will lose your customers if they don’t want to publish in a crap journal. When I left the company, the rejection rate of the the lowest-tier BMC journals, which had the theoretical capacity for unlimited numbers of papers per year, was in the neighborhood of 50-60% – which was about the same as the society journals I worked with in my next stint, that had a fixed number of papers per year and weren’t author-pays. So your assumption isn’t universally true.
@Mike. In economic terms, journals get a free ride, as they have to outsource a major part of their publishing process to the scientific community, but don’t have to pay for it.
I disagree this is a major component in economic terms. In the life cycle of a journal article, peer review takes about 4-6 hours. This is a very tiny fraction of the amount of time, effort and FTEs that goes into producing a quality article, so I’d hardly call this ‘a free ride’. (For example, it can take more that 4-6 hours just to find and convince three referees to review a paper, and to chase and cajole them to get their reports in on time.) Editorial efforts take much more time and, I would argue, a completely different skill set. Paying peer reviewers would be great, but you’d need to factor that into your business model as well.
Would we in turn get editors who were qualified enough to judge whether a review was suitable and pay accordingly
This, in my opinion, is the crux of the matter. What makes a good scientist does not overlap much with what makes a good editor. I’ve interacted both with trained editors and well as untrained academic editorial board members, and though some of the latter were truly excellent and had good native instincts, not a few were pretty appalling. If scientists are to do it all themselves, they need to be trained, and not all scientists will have the knack, so there would have to be some sort of screening process.
Considering only a few people in the world actually are interested in reading any of my articles (because we are all so specialized) I could easily put my articles on my website.
costs: zero.
I use my website for other stuff and have plenty of bandwidth left. The few people who will download the article over the years won’t make a dent in it.
How expensive would it really be to publish your own work? It would be dirt cheap.
Include peer review from your peers at your university and host it on the university server, and you are still looking at zero costs (assuming your peers are being paid for other reasons, such as doing research and teaching).
You want a hardcopy? If I remember correctly I received 150 professionally printed PhD dissertations for around 1000€. It would mean you could start writing articles of a few hundred pages long again, instead of 1-10 pages.
If we would want to we could easily go for cheap.
If impact factor wouldn’t roughly determine funding possibilities.
what about peer review?
Peer review doesn’t need to be anonymous. It could also happen after the publication of an article. By means of publication of counter arguments/evidence/papers.
I don’t buy it, Mark. There are already enough nutters on the internet.
Impact factor of the journal research is published in is a lousy way to evaluate the value of that research and an even worse way to evaluate whether other research proposed by the author should be funded or not. Everyone knows this.
The reason that people do this to evaluate the value of research and whether proposed research should be funded or not is because they are either to lazy to evaluate it themselves, do not have sufficient time to evaluate it, or are incapable of evaluating it on its merits and so they choose an indirect surrogate measure which sometimes more or less corresponds to the measurement they are trying to achieve. If a scientist attempted to publish conclusions based on such indirect surrogate measurements the paper would be (or should be) rejected.
The largest cost by far, by at least a couple of orders of magnitude in the publication of research is in the funding to do the research in the first place. It makes little sense to me for the various funding agencies to spend 99%+ to get the research done and then depend on charity or the whims of the marketplace to get the findings published. If the research was worth funding, surely 1% of that could be allocated to publishing the results.
Richard, I think the main features that you read Nature for, news and views and Futures are written by people who are actually paid for what they write.
Richard, I think the main features that you read Nature for, news and views and Futures are written by people who are actually paid for what they write.
Um, duh?
Yes, they are. Failing to see your point.
@Richard
The quantity of nutters on the internet has probably no causal correlation with the quality of your or my research.
What I mean Mark is that it’s going to be difficult to distinguish the signal from the noise. Nutters publish nutty stuff on the internet. How do I know, a priori, that your excellent paper isn’t nutty? Peer research? But if your peers are your friends and they’re all nutters too…
Which is why we trust journal editors to find non-nutters, in our roles as producers and consumers of scientific knowledge. There has to be some arbiter of quality.
That’s why I suggest the alternative that your work is published on university servers and reviewed by university peers.
The university will have a vested interested, in the sense that they have a reputation to protect. I would imagine some procedures that will guarantee a certain quality level will be put in place.
The main point I am trying to make is that publishing is basically for free in this information era. Reviewers already work for free. Many editors work for free. Why not accept the work of editors and reviewers as part of the academic job and integrate it into the actual research job. Make the university the hub of science publication.
It would probably lead to a healthier situation even regarding publication because it will become easier to publish controversial or negative results.
That’s odd. My comment was swallowed whole like a 7″ trout.
I’ll try again.
You’re nearly right about the University hub idea. You still need to convince me that your university peer reviewers are not your friends. And if you really think that you can do it for free, they will be your friends and no one will ever believe anything you publish.
Publishing is not free, and never will be in this economic world, not if you want some semblance of quality. Do you really think that you can edit a full-on journal ‘unpaid’ and carry on your academic life?
I think there’s a number of people here, not just Jennifer Rohn, who would laugh at you (well, probably not: they’re too polite).
it will become easier to publish controversial or negative results.
Yeah, that’s what’s worrying me.
Richard, I almost start to believe that you think that all departments at the university are run like a regular old boys’ club, and you can’t trust anyone.
Of course, this may well be the case, but I can assure you that the qualitative characteristics of the international social peer network isn’t much different at all.
It’s all subdivided into groups that could be labeled as ‘friendly’, ‘neutral’ and outright ‘enemy’.
That’s why you suggest certain reviewers to be excluded from reviewing your paper to start with, and you give a list of ‘friends’. Friends in the sense: people that are not hostile towards your research.
What’s the difference?
I can’t see it. There are plenty of ‘friends’ and ‘enemies’ within a university as well.
As for the editor problem. I’m sure the universities could hire one or two if it would take up too much time. Many journals actually do work with unpaid editors. So most of the time nothing would change whatsoever. And who says one person has to do all the editing. Spread it out; the experience and the responsibility.
What I mostly see in your reactions is actually incredulity. The first scientific journals were published by academic institutions. The only difference was that they needed a commercial publisher to publish their work. In the modern era anyone can publish anything easily for the whole world to see on the internet. And that includes actual research institutes.
I’m sure that Universities also could opt for reviewers from outside the university. And why would peers refuse to do that? Especially if it became common practice? What would be the difference except a technical one?
Would you find a commercial institute which needs to make profit more reliable regarding the publication of scientific data, or Universities which rely on scientific reputation?
I can only see one rational outcome to this question. I would opt for University publication.
Were it not for the fact that funding depends to a large degree on impact factor (and the social network).
Tradition has pushed scientific publication into the realm of commercial enterprises and we are a bit stuck now, because nobody wants to stick their neck out, and nobody questions tradition in a rational manner. Not even scientists.
Question: who will publish the work of scientists in companies? Non-university academic institutions? Small colleges?
Of course we have ‘friends’ and ‘enemies’ amongst our peers: but the point is we have an editor who — at least some of the time — is neutral.
In the modern era anyone can publish anything easily for the whole world to see on the internet.
Yes, and the internet is crap.
I almost start to believe that you think that all departments at the university are run like a regular old boys’ club, and you can’t trust anyone.
well, if you don’t agree, maybe you’ve benefitted from it more than I have.
Nah, I haven’t really benefitted. Quite the opposite probably. I just noticed being negative about it doesn’t really do much, so I switched tactics and just ignore it, or pretend it is part of the game and be neutral about it. I’m quite sure already that I won’t have any form of career in the far future here. And nobody will deny that.
That said, are there really differences between departmental networks international ones? (besides that you don’t see them on a daily basis)
I haven’t seen any qualitative differences and the latter does determine peer review.
Mark, to be fair lots of people are asking questions about the commercial enterprise of STM publishing. There are very good questions to be asked. I just don’t think we have that many answers at the moment – plenty of ideas but not many really solid answers. That’s what these discussions are about.
But never think that any form of publication is free. Institutional websites cost hundreds of thousands to millions of pounds to run. 100s of millions have gone in to the development of repository systems and many people think they are deeply flawed as well. Even if you think the marginal cost of putting up your paper now is minimal bear in mind the following rather disturbing statistics someone told me (for which I have no citation, but he, this is the internet) that the UK’s carbon footprint due to IT is as big as that due to air travel. Just keeping disks spinning costs millions.
Someone somewhere has to bear the cost of this. I think its good to argue about how best to do that. And new ideas are going to be very helpful.
That’s why I suggest the alternative that your work is published on university servers and reviewed by university peers
It’s a pretty basic editorial rule that an editor should exclude peer reviewers housed in the same institute as the authors. It is not possible to be unbiased when (for example, in the UK) the publication record of your peers affects your own standing, as is the case with the Research Assesment Exercise. Also, proximity tends to foster cronyism.
Maybe universities could have agreements to review a designated sister university’s papers, and vice versa.
_Knowledgeable editors would be a good thing in publishing. _
I agree. It would.
If you’re not part of the solution Henry, you’re part of the precipitate.
Hang on, that’s not right…
Why nobody have talked bout rating?
one can design al algorithm to rate papers which depends in your strength as scientist. The rating cannot be anonymous so if you cheat with a paper people can see it.
As I’ve previously said, if you have wiki papers the peer-revision process can be done in the fly.
Somewhere i read the argument that there are too many papers. Also there are a lot scientists, if each one rate one paper a week in some months you’ll have some nice results of the important work in some field.
Each time a scientist read a paper he makes a lot of connections that would be really useful to save as part of human knowledge.
I don’t know if I’m talking about the same, but I think that we have to change the car, sell it by parts, buy a bicycle and with the extra money plant some trees (because a wikibicycle is a lot cheaper than the journalcar).
@myself: obviously if you edit a paper you share a percentage of the authorship.
The bycicle is no good to go from santiago to sydney so you also need some planes with cheap tickets. Lets say Nature airlines cost US$0.99 per paper, you just need to sell papers to the 10% of world population and you are rich (ok, I think I stole this idea from somewhere…). In the same way last.fm connects with itunes, mendeley conects to iNature to seed your reading so your friends can see in what you are.