Richard Smith: Medical journals: a gaggle of golden geese

Richard SmithI used to be part of running a course for editors of medical journals, and on each course we’d encounter editors, usually distinguished professors, who worked evenings and weekends on journals for free. They did so as a contribution to their specialty and for a dollop of honour. In their naivety they imagined that the journals made little or no money. In fact some scientific journals are considerably more profitable than oilwells. The editors were being taken for suckers and exploited more cruelly than the people working in fast food joints.

I was reminded of this, a favourite theme of mine, when I read the paper on high reprint orders in medical journals posted on last week. The main message of the paper is that high reprint sales are strongly associated with pharmaceutical funding. The authors have laboured long and hard to produce this result, but it won’t surprise anybody who knows the inside of medical publishing. The big sales of reprints are almost always to pharmaceutical companies, and the publishers are very active in selling them.

What the authors of the article don’t say (and perhaps don’t know) is that the profit margin on reprints is very high, perhaps 80%. The paper shows that one Lancet reprint was sold for £1.55m, which would mean a profit for Elsevier, the owners of the Lancet, of well over a million pounds. That’s one of the reasons why Elsevier makes a profit margin of over 30% by publishing its 1200 or so journals, a far higher margin than in most industries.

The conflict of interest is clearly huge. If Elsevier had to maintain its profit margin by cutting costs rather than publishing that one article then it might mean firing 25 editors. I’ve written about this before in my article arguing rather painfully that “Medical journals are an extension of the marketing arm of pharmaceutical articles.”

I applaud the Lancet and BMJ for providing data on reprints, but the publishers of the New England Journal of Medicine (Massachusetts Medical Society), JAMA (American Medical Association), and the Annals of Internal Medicine (American College of Physicians) all declined. They clearly have something to hide.

Indeed, have you ever seen the financial accounts of a medical journal? Probably you haven’t. Although most medical journals are owned by medical organisations, their finances are much less transparent than any publicly quoted company. So let me share with you the financial statement of what I’ve called “an average medical journal,” of which there are hundreds. This is an imaginary British journal; American numbers would be much bigger.

Income statement of an “average medical journal”
Income (000s)
Subscriptions 600
Display advertising 200
Classified advertising 100
Reprints 250
Copyright fees 50
Total 1200
Salaries 100
Typesetting and printing 100
Paper 200
Website costs 75
Communication costs 25
Editorial expenses 25
Marketing 25
Total 550
Gross margin 650 (54.2%)
Overheads 150
Profit 500 (41.7%)

You can see that this a very profitable activity. Let me now imagine the financial statements of a very unaverage journal—the New England Journal of Medicine. My guess is that the journal has an income of around $100m a year—and perhaps $20m of that may come from reprint sales. It is generously staffed, the editors are no doubt well paid, they are generous with people from the developing world, and the whole thing is a Rolls Royce operation—so expenses are high. Even so, I bet the journal makes over $20m in profits. That’s why the Massachusetts Medical Society has grown fat and has an expenditure per member way above any other state medical society.

But there downsides to all this opulence. The society wants more money and would like, as one editor of the journal put it to me, to start “New England Journal of Medicine Fried Chicken.” There are anxieties that the need to keep up the profits will lead to compromised decisions. But the biggest downside is the journal’s opposition, sometimes vehement, to open access. It’ll be tough, probably impossible, to make the same profits from with open access publishing.

This is the age of transparency and accountability. Whatever is not transparent is assumed to be corrupt, biased, or incompetent until proved otherwise—like it or not and fair or not. The time has come for journals to open up their financial statements. I apologise to the New England Journal of Medicine if my guesses on its finances are wildly wrong. If I’m wrong, please show me. This blog overlaps to some degree with a rapid response sent to the BMJ in response to the paper on reprints.

Competing interest statement: RS was the editor of the BMJ and the Chief Executive of the BMJ Publishing Group. He is a zealot for open access and was from 2004 to 2011 a member of the board of the Public Library of Science. Richard Smith is director of the United Health Group’s chronic disease initiative.

  • Sylvester Chima, MD, LL.M

    If medical journals are making so much money from the publishing enterprise. Shouldnt authors be paid for their contributions?

    Sylvester C. Chima, MD, LL.M
    Associate Professor and Head,
    Programme of Bio &Research Ethics and Medical Law
    College of Health Sciences
    University of KwaZulu-Natal

  • Peter Ashman

    Richard Smith is a, highly regarded and respected figure. He is also an experienced medical editor and publisher, so I was surprised to see that the “gaggle of golden geese” blog contains so many factual errors and so much emotive language criticising publishers.

    I have never published a journal on which the Editor works for free. At BMJ Group, editors are either employees of the company or are clinicians/academics who are appointed by competitive interview to the role of Editor in Chief. They are paid an honoraria which reflects the amount of work expected to be done. We value greatly the intellectual input from our Editors and we appreciate the time, care and attention they put into making their journals the best that they can be. Editors are certainly not suckers and are not cruelly exploited. I’m not sure what the evidence is to show that oilwells are less profitable than some scientific journals but Exxon Mobile posted profits of $9.45bn for the first 3 months of this year – I’d certainly like to work on a journal that makes THAT much money!

    I’m sure that Elsevier can stand up for themselves but it should be noted that a quick search of their site shows that they publish over 2600 journals – not “1200 or so”.  If Elsevier had to cut its costs to replace the £1m profit from the article mentioned, there are very many ways that they could do that – it’s wrong to speculate that they might fire “25 editors” [who, it’s worth noting would have to be earning £40k each…..probably more than someone working in a fast food joint!].

    I think it’s also wrong to speculate that the reason that some journals chose not to disclose their sales data means that they have something to hide. There could be any number of valid reasons not to have participated – secrecy shouldn’t be assumed as the reason.

    Many (not sure about “most”) medical journals are indeed owned by medical organisations – these societies, colleges and associations often rely on the surplus provided by their journals to fund the organisation’s other activities. Is it wrong for such  organisations to make money from their journals?

    The numbers shown as being from “an average medical journal” are indeed “imaginary” – they certainly bear little resemblance to any journal I’ve seen and there’s no evidence to show that these imaginary numbers would be “much bigger” if they were from an American journal. If the numbers are shown in order to prove a point, then surely they should bear some resemblance to the real world. For example, very few scholarly journals enjoy classified advertising revenues any more (the BMJ is unique in having a classified section amongst BMJ Group-owned journals); the imaginary reprints revenue of £250k shows no corresponding costs in the expenditure line (reprints aren’t free to sell, typeset, print and distribute); it would be impossible to run a journal such as this with salary costs of just £100k and one of the major costs (which are out of the control of publishers) is distribution – this imaginary journal seems to have no distribution costs at all. It’s a completely inaccurate picture of how a journal is run and I would urge caution for anyone who might think otherwise.

    Finally, whilst I agree that, in many cases, transparency in medical publishing is a good thing – especially in terms of editorial rigour,  peer review and trial reporting – there may be any number of reasons why a publisher or medical organisation might choose not to open up their finances. It shouldn’t be assumed that medical publishers are biased, corrupt or incompetent.  Most of us are doing a good honest job and producing the best possible publications.

    Peter Ashman
    Publishing Director
    BMJ Group

  • Richard Smith

    Thank you for your comments, Peter.

    You encourage me to spell out why it is that scientific journals make so much money. I’ll then respond to some of your particular points.

    But first I have to make the obvious point that your best response to my blog would be to publish the accounts of all of your journals. I challenge you to do so.

    A slide that unfortunately the technology doesn’t allow me to post shows the profit margin of 16 companies. At the top is  Springer’s science publishing, which has a profit margin of 37%. Near the bottom is Walmart with a margin of 5.8%. For Elsevier science publishing it’s 31.7%. Only Microsoft had a higher profit margin than Elsevier. For Exxon Mobile it’s about 17% and for Apple 18%. Hence my comment on oilwells.

    You’ll note that the profit margins of these scientific publishers are close to the profit margin of my imaginary journal, and you’ll know that the publishers’ lists include many journals that have low margins. So some must have much higher margins than the overall margin, and you’ll know that many do.

    The reason that scientific publishers make so much money is that all the value is added by others, mostly without payment from the publishers.

    Imagine opening a journal and reading a big randomised trial. I suggest that 99% of the value is in the research itself. Big trials are expensive and difficult to do. They are funded either by public money or pharmaceutical companies. What value does the publisher add? Peer review is one answer, but even if we accept that this is a value added process (which I don’t) then the peer review is usually done without payment from the publisher. If accepted the paper is then copy edited, but many publishers as you know offer minimal editing, which anyway is not a very value added process. The publishers then print and distribute the material—either on paper or electronically, but these are not value added processes. They are cheap commodities.

    The publishers, you will say, market the journals, but many journals are “must have” and require minimal marketing. You will know that some journals cost over $15 000 for a year’s subscription.

    You might tell me that publishers have built the brand, but the brand comes primarily from the research published in journals over the years—research that is not funded by the publisher.

    I remember vividly when I first encountered the business aspects of scientific publishing. My predecessor as editor of the BMJ was having a row with rheumatologists over the Annals of Rheumatic Diseases at a time when the BMJ owned the whole journal. The rhematologists wanted half. We sat around a table, and one of the rheumatologists said “We do the research, we edit the journal, we peer review for it, we buy it, we read it, and we store it. What exactly do you do?”

    “Tell them,” my predecessor said to me, and I struggled.

    The final insult after all this is when publishers charge for access to the research so restricting access. You’ll know that this is why so many scientists are angry about publishers and are pushing hard for open access, which will come.

    “All publishing is theft,” said the BMA Librarian to me when I was first at the BMJ.
    That seemed extreme at the time, but I’ve come to understand what he meant. (Amusingly that libarian now works—or at least did work—for Elsevier.)
    Let me now respond to some of your points.

    I know that the BMJ Group pays its editors, but there are still I suggest editors who work for free. The editor I was thinking of was the editor of Social Science and Medicine. Not only was she not paid, but there was no payment for the work of her secretary. I accept that many Elsevier editors probably are now paid.

    I’ve spent some time searching for the number of science journals that Elsevier publishes, but I can’t find it easily—even in their annual reports. You will know, however, that Elsevier publishes many non-science journals.

    I do think that it’s wrong for scientific societies to make money from restricting access to the research that is fundamental to their mission. If you are the British Society for Hypocrisy then your mission is to promote hypocrisy. How can restricting access to research on hypocrisy advance that mission? I fear that these societies have forgotten their mission.

    I mostly stand by my figures for the imaginary average journal and repeat my challenge to you to publish your accounts.

    You don’t dispute the 80% margin figure for reprints, and that explains why the printing costs are low and included in the overall figure. The companies distribute the reprints. I should have added £50k for distribution costs, although you will know that increasingly distribution is electronic—and the marginal cost of adding one more subscriber is effectively zero.

  • Mike Taylor

    “But first I have to make the obvious point that your best response to my blog would be to publish the accounts of all of your journals. I challenge you to do so.”

    Exactly. That would end all speculation instantly, and we could discuss reality rather than “average” figures. I don’t understand what reason there can be for not doing this.

  • At our recent “Avoiding Avoidable Care” conference there was a panel devoted to the role of journals in the culture of over treatment. One of the journal editors asserted that there is a “Chinese wall” between the editorial side and the advertising side. I think it is fair to say that many in the audience were skeptical about that answer being sufficient.

  • A very interesting blog, I didn’t realise quite how much profit some medical publishers make from reprints so thank you for highlighting this fact.  I work for a medical publisher and would be very happy to publish our accounts considering we have little or no revenue from our journal.  I know this is not replicable for larger publishers but it shows that running a sustainable business publishing medical articles without raking in huge profits and remaining completely unbiased is possible.  We do not charge scientists to either publish or read the results of their own research and they are free to use their articls in any way they wish. And we pay our editors very well!