Monday, May 07, 2012

Paying to Publish

You proved a nice theorem, wrote up the paper and submitted it to a major computer science conference. Your paper was accepted! Congratulations. Now pay up.

An author of every paper accepted at a CS conferences is expected to present that paper at the conference. To do so requires at the least paying the registration fees, travel and lodging to go the the meeting. That can easily run one to three thousand dollars (or more) depending mostly on how far you need to travel.

You can use grant money for these expenses. Some conference offer support to those who need it, particularly students. CS departments will often help out if needed. Sometimes people pay out of their own pockets and, in any case, the funds come from limited pots that could have been used for other purposes.

In the "old days" this was less of a problem. There were only one or two conferences relevant to one's field and you were probably going to those conferences anyway. Now as the field has grown and it has been harder to get your papers published in the strongest conferences, you may find yourself traveling just to give the talk. Even many major conferences don't draw many attendees who don't have papers in the conference.

We haven't seen an outcry of these expenses, say compared to the outcry over the cost of digital library subscriptions. Perhaps we consider attending the conference a "reward" for getting published.

We could just eliminate the conferences and publish the proceedings and post videos of talks, made at the home institutions, saving the field huge amounts of money. Then we could actually choose to attend conferences to meet people in our field instead of just talking at them.

Note: Elchanan Mossel had a similar observation in a comment on a recent post.


  1. A great point, Lance. It's something I've wondered myself for a long time. Logically, there ought to be an outcry over these expenses and it is puzzling to me why there isn't. Thank you for opening some common ground with this anti-copyright researcher. :)

  2. I find that sending a student to speak about his/her work is great for getting them started in the community. For your "meeting" purposes there are invitation workshops such as Dagstuhl, which may work better at some points in one's career (though the "premium" FOCS/STOC also work well).

    To your last point, going YouTube would be giving up the venue for nothing. There will be no connecting, no post-talk discussions, no "captive audience" effect, just yet another obscure video.

  3. I was once asked why journals that charge you to print your work are any worse than conferences than force you to pay to present your own work. I had no answer. CS should ditch the conference publication model.

  4. I find it far more annoying that we are asked to pay to serve on the program committees.
    If there is a physical meeting, we are asked to pay to travel to it. We are then asked
    to chair sessions at the conference, and we are asked again to pay for the privilege.
    In short, the program committee is asked to subsidize the cost of the conference.

  5. Journals that charge you to print your work are not common in
    our field and more importantly there ARE alternatives.

    Conferences ALL charge for registration, some more than others,
    and Travel of course costs money. An alternative might be a lower
    cost conference, but still SOME cost AND Hotel AND Travel.

  6. Some people care - here's a comment I posted on this blog on Mar 30. I am also consistently presenting this view for a couple of years now - Elchanan.

    On Mar 30:

    Conference proceedings limit access in other ways too.

    In order to publish in conferences you must be able to attend the conference.

    This results in the rich getting richer effect: you can only publish in conferences if you live in a country where there is sufficient grant funding. In the US you must be well funded.

    Moreover conference publications are important in evaluation of computer scientists - so even in the US it is the rich getting richer effect (and rich here means in money - not in intellect).

    Similarly attending conferences is easier for healthier people with good family support.

    Why should the main venues for publishing results in computer science should be so biased in the 21th century?

    Elchanan Mossel

  7. Good point. I think a more effective model is to have one large annual meeting (like many other disciplines have), and transform conferences like FOCS/STOC/SODA/... to refereed tracks on arXiv, with the possibility of presenting the accepted papers in each of these tracks at the annual meeting.

  8. Completely agree with Lance!
    I have to spend too much time and money on futile expensive traveling, just to randomly chat with people for about 10 minutes at the lobby?!
    This is much worse than expensive journals. They don't cost you the time at least.

    If I want to have professional meetings, I could do it in an organized manner: plan and prepare a trip to some university/institute and work with the person.
    There is no real justification to pay expensive hotels (this STOC costs >200$ per night) and fee (>600$), and travel cost (absurdly high). Not to mention all the time spent in planning this trip.

    And then, after this, I still have to prepare the journal version.
    Really a waste of time and money.
    (p.s., "waste" here means that the revenue is far less than the cost. Not that there are no pros to conferences.)

  9. One thing I've wondered about is that CS conferences are often held in somewhat high end hotels ($300 - $400 a night). This affects not only the hotel costs but the registration costs (which are partly used to pay for the hotel ballrooms rooms).

    Is there really no way to find cheaper venues so that registration fee is closer to $100 than $1000 for a 5 day conference?

    With 500 attendees you would have $50,000 which should be enough to rent 3 large rooms, A/V equipment and refreshments for 5 days I think?

    It would probably be a bit less convenient and a bit less comfortable but it would certainly save a lot of money for all involved and allow more local students and researchers to attend. At a $100 locals could even potentially pay the registration out of their own pockets if they really wanted to attend..

    But I've never organized the logistics of a major conference so I'm probably missing something.

    1. CS conferences are often held in somewhat high end hotels ($300 - $400 a night).

      This is not the case for theoretical CS. I don't know about other subfields but in TCS, more than $200 a night is a rarity.

  10. I don't have a great answer, but regarding optional attendance, I'd be concerned about increasing the divide between the academic haves and have-nots. Students from well-funded and typically larger and more prestigious research groups would get to go to conferences where they could begin the sort of networking that is absolutely critical for their careers, while less fortunate students would be shuttled to YouTube and miss out on that completely.

  11. I agree with Lance too (as many commentators apparently). As a student, I had the chance to participate to a math conference. The conference had no proceedings, the idea was simply to present some of your work. And it was consequently a much larger conference than the usual CS ones since accepting a lot of talk bring no risk of "devaluation of the conference" (as we hear a lot when people propose that FOCS/STOC/etc... conferences accept more papers). From what I heard there, some people chose to present some on-going unpolished work, while others (including me) presented results from several already published papers in a single talk. I've been to several CS conferences too, and I clearly found the system of math conferences much better. I was interested in many more talks than a usual CS conference (even though it was math focused) just because with a dozen of parallel sessions (I guess), there was almost always a talk of interest to me. The only drawback was that I had to miss some interesting talks because of overlap...

    Another point is the following: Suppose you publish papers in conferences A, B or C (because you were accepted in those conferences), but you are more interested in papers published in conferences D and E (where you were not accepted say). Then you have to go to conferences A, B and C, and you likely do not have any money left for conferences D and E. That is a shame...

    I think all the discussions we have in the community about the drawback of this or that system have the same simple conclusion. To cite Lance, "it's time for CS to grow up" and publish our papers mainly in journals, just like any other serious field!

    1. Translation: It's time for CS to follow established practices in other fields whether it makes sense or not, just because they are "serious" (whatever that means).

  12. The cost of travel is certainly getting more and more prohibitive over time. For those with young children, especially women, there are other implicit "costs" that don't get enough attention or support from the community. It is difficult to travel, especially to far off places or for long trips, leaving the kids at home. It is even more challenging to bring the kids along -- one is continually distracted and childcare costs can be immense. (As a side note, I wish NSF would support childcare costs during essential travel!). This puts people with young children at a considerable disadvantage in terms of lower visibility in the community compared to peers without family obligations. Being one half of an academic couple (with a very supportive spouse) I can say from personal experience that women are more disadvantaged than men in this regard. I wish we were living in an era where visibility was not so highly correlated with travel. However, admittedly, I don't have a good solution to this issue.

    1. "(As a side note, I wish NSF would support childcare costs during essential travel!)."

      As a researcher who struggles both to find conference support and simply employment, I find the above comment offensive. As part of a dual-career (dual salary) couple, surely you can afford to pay for your childcare! The NSF money could be better spent on a student/researcher who otherwise would not be able to attend.

      On the one hand we say that conferences are too expensive and on the other hand, we want to "expense" things that should not be paid for by research funds.

  13. Why are conferences always held in seemingly the most ludicrously expensive locations? We could just as well hold them in flyover states at locations far away from downtown (but correspondingly closer to the airport). Sure, some people see conferences as vacations and this would defeat that purpose, but let's be honest, if the money's coming from a grant or an institution, the vacation model is blatantly unethical.

    And why not host more conferences at universities? During, say, the academic breaks, we have all these empty classrooms just sitting around, but no, the conference has to be held at poshy downtown convention center for $5000 a seat!

  14. OK, so who is willing to boycott conferences?

  15. I see that some very well known people in the field (for example, Dan Spielman, Elchanan Mossel and Shuchi Chawla, and the blog owners too) believe at the very least that conferences are either sub-optimal or are not being organized in the right way (in expensive locations, for example). This leads me to wonder: who exactly are the people calling the shots and ensuring that the status quo is maintained?

  16. @Anonymous9:35 PM, May 08, 2012:


  17. You have valid points but they don't imply getting rid of conference model.

    We can have a meeting that everyone would attend of the kind you are suggesting without changing anything about other conferences. People are still free to publish in journals in place of conferences. If you and others think it is a waste of time and money to attending and publishing in conferences why are you doing it?

    Let me answer: you want is to shift the weight from publishing in conferences towards publishing in journals and you want it to be a revolution. It is not going to happen that way. My suggestion to those who want to shift the system towards publishing in journals is to stop complaining about the current system and put their actions where their mouth is: give more weight to publications in journals when you are evaluating other researchers (e.g. for hiring), express publicly that you are doing it, and publish your results in journals in place of conferences (it won't hurt tenured researchers), and over time we may see the shift they want.

    There is clear differences between criticism against publisher like Elsevier and your criticism against what you consider a waste. I don't think people would complain about Elsevier as much if their costs justified what they are doing. Not only it doesn't, it is an outrageous situation where they are making huge amounts of profit on our backs. There would be a much harsher reaction if conferences made similar profit out of conference fees.

  18. Traveling and meeting each other is an essential part of research, and well worth its (high) cost -- even in today's world one does need this semi-structured face-to-face time for various human reasons. A completely different question is whether these meetings should be competitive and "prestigious".

    While CS is rather unique among academic fields in terms of the latter, I do not think that we are unusual in terms of the amount of travel we do or the amount of money we put into it. So the question is really whether we (the CS community) are using our travel time and money wisely in comparison to other forms of conferences and travel common in other academic fields. Looking at other fields, my answer is "yes".

  19. Is the "TCS confereces problem" not somehow related to the puberty problem? Here and there one needs recognition. Not a silent publishing of results. "Adults" (like Math) have easier. But we ("teenagers") must pay for this recognition. Normal teenagers pay with their health (drug orgies), we pay with travel, accomodation, registration costs. And what if one doesn't like traveling at all (like me)? And what if one cannot travel by personal reasons? Are all they automatically outside the competition?

  20. Frankly, it's a wasted opportunity that computer scientists aren't demonstrating the value networks can bring to academia. At the very, very least host one big experimental conference, while retaining the traditional ones.

    The thing too is, it doesn't have to be Youtube video with 8 views or everybody fly to Hawaii. Why not have a distributed conference? Get 20 or so universities to volunteer to host the conference. One or two schools could easily support all of New England, with the participants able to drive home at the end of the day if they wanted. It's still a social event, just a regional one now. And if you like the vacation model, or want to talk to somebody across the Atlantic, you can still travel there. You've just got options now.

    Each university connects a television to the internet in a couple classrooms and puts a webcam in another. Now everyone can watch the talks (with other humans) and present their talk from wherever they happen to be. And it becomes that much easier to involve more of the community: For the price of one student, you can now send the whole grad class. You can even bring undergrads to see what research is really like.

  21. I have the following questions:

    1. Why is that the most prestigious conferences like FOCS/STOC/CCC/LICS/ICALP charge so much? This ultimately leads to less participation of students and early career researchers.

    2. We just go to conferences to present our paper and thus ensure that our paper gets published. Other than that, we effectively do nothing or hardly discuss some bland stuffs. There is no intellectual gain at all.

    3. It may be possible that we all resort to publishing our works digitally and thus, there may be much reductions in the paper publication costs. All our papers can be circulated digitally and paper publications should be only subscribed by an institution and not by an individual. This may drastically reduce the costs. Is it that, the publishing companies earn a larger profit when publishing in paper form, rather than in digital form?

    4. What may be the possible solutions, so that more number of both graduate and undergraduate students can participate in these conferences and symposiums ?

  22. Noam: Traveling and meeting each other is an essential part of research, and well worth its (high) cost -- even in today's world one does need this semi-structured face-to-face time for various human reasons.

    I disagree. For some researchers traveling is essential. For many others it is not. It depends on one's personality. There is no universal law here. And since the cost is becoming too high, and the alternatives for different kinds of communication are improving fast, I believe there will be a substantial change on this matter in the near future.


  24. This is a particularly big problem with the weaker conferences. People spend a ridiculous amount of money to attend a conference where the only talks will be on results that were not strong enough to get into a better conference. This is a huge waste of time and money just to get a "conference publication" to list on a CV.

    There's nothing wrong with writing papers that are at not at the top level; I often do. But they should be published in lower tier journals rather than conferences.

  25. Dear Lance Fortnow

    I am an early bird researcher from India and I recently attended the GRC conference in Italy. I had applied for a travel grant from the Indian government which was approximately 1.2 Lacs INR. The registration fees was the same amount. But the conference chair waived off the fees entirely and I could attend the conference without paying a single amount. I believe that reputed researchers get a lot of funds from govt organizations and other philanthropists and not all money is spent in the conference which they have to give back. By providing such waivers, I think they are helping the people to communicate face to face and actually absorb how scientists actually work. They can get updates in their respective fields. I would suggest that based on a student's merit, there should be no registration fees, boarding and lodging fees to encourage more collaborations.