Statement from Paul Beame (SIGACT chair)

This is a guest post from Paul Beame, the chair of the SIGACT executive committee.

First of all, I would like to express the desire of the SIGACT leadership to work for the benefit of the whole TCS community including SoCG. We are trying to bring the TCS community together more, as in the proposed 2016 co-location of SoCG with STOC but also more generally. At the STOC business meeting there was also strong support for a Federation of TCS conferences in 2017 where we would try to co-locate many more theory conferences with STOC in the latter half of June. We’d love it if SoCG wanted to join in 2017 also but also understand that SoCG is expected to be outside North America in 2017. Obviously, with this perspective, we hope that SoCG stays with SIGACT and ACM.

I should explain more about finances. ACM provides a substantial net financial benefit to the TCS community. SIGACT gets money from ACM for its share of the digital library revenues based on downloads of SIGACT material. Over the years, from a combination largely of STOC surpluses and ACM DL revenues, SIGACT has built up quite a large operating balance as well as a separate endowment fund for awards. We do expect DL revenues to decline markedly as things become more open access, which has led to conservative budgeting of SIGACT funds. (2/3 of revenues are from papers older than 5 years, often from conferences that are now sponsored by other SIGs, and a move to delayed open access might wipe out all of that revenue stream overnight.) However, in the last year, SIGACT got $39K more from ACM from the DL than it paid to ACM in allocation fees for its own operations and passed on overheads of its conferences. This isn’t all available: SIGACT does lose money on every member because it costs more to print and mail SIGACT News than an average membership costs, a net loss of $14K overall, though there is now an electronic-only option that may make this break-even on average. STOC also made a total of over $110K in surpluses the last two years (we don’t have a 50% surplus return agreement for STOC as we do with SoCG). For several years SIGACT had an operating fund balance of around $800K but these two items have caused a very significant jump to well over $900K.) The SIGACT leadership are trying to return much of that extra money to the community both to STOC and other conferences. This year’s money to conferences, of which $4500 went to SoCG, was a start. We don’t want conferences to count on this, or make promises that we can’t keep in future which is why there are no commitments at this point. Leaving ACM would cut off SoCG from this money.

A bit more detail about in cooperation requests. SoCG would have to ask each year, but the bar for SIGACT support for this will be low for conferences outside North America. (After all SIGACT has committed to pay for extra costs in this case so it is reasonable that SIGACT get asked each time.) The bar for approval of this inside North America, especially in the US, would be very high. (Having seen how things worked with ACM when I ran local arrangements for STOC 2006 in Seattle, it really couldn’t have been easier. All I had to do was tell an ACM person to pay some bill and it got done; I didn’t have to handle a penny and their advice saved the conference a bunch of money. They negotiated some really good rates for us and some good stipulations in the contract that would have been a mess for us if they hadn’t been included.)

There are benefits of ACM policies for sponsored conferences that SoCG would be foregoing when it elects to operate in cooperation rather than with sponsored status. Just one example is the following. ACM has an arrangement with RegOnline (the usual registration service for many conferences) that halves the costs (I can attest to this having paid for them at regular price for FOCS conferences) and sends the money direct to an ACM sub-account for the conference. Expenses are paid by ACM staff as authorized by the organizers. This means that organizers do not set up bank accounts and reporting is easy (no final auditing or anything for the local organizers, something that would be necessary for a non-profit to do for tax purposes).

Contingencies required in budgets for ACM should not be an issue here. Even without changing the number of attendees, it is easy to meet contingencies by adjusting the assumed distribution of registrants to include a higher percentage of non-members or late registrants, or a smaller percentage of students to compensate for much or all of the contingency. It is also possible to approve budgets with lower contingency percentages for conferences that consistently have budgeted well. Contingency is sometimes necessary, though. I did have one case when I ran IEEE FOCS 2010 where with a fudged contingency we would have had a deficit except that SIGACT stepped in to help. Cutting things close works fine if you have an organization with deeper pockets backing you up. An independent conference could not cut it so close with the contingency and would need to build in a real contingency that is quite a bit larger than what actually one can budget for ACM.

Finally, the situation with SoCG and ACM is very different from that of CCC and IEEE, particularly in terms of the cost versus support for the conference.  Moreover, unlike with CCC, SoCG leaving ACM would eliminate the possibility of future “in cooperation” status with ACM, something that the members of the CCC steering committee are very keen on maintaining. I hope that SoCG wants to maintain our relationship, too.

About Jeff Erickson

I'm a professor of computer science at the University of Illinois, Urbana-Champaign and the chair of the steering committee for the International Symposium on Computational Geometry.
This entry was posted in Uncategorized and tagged , . Bookmark the permalink.

11 Responses to Statement from Paul Beame (SIGACT chair)

  1. someone says:

    Thanks to everybody for the explanations here and the thoughtful comments.
    My opinion is that ACM is too large and dysfunctional for the interests of SoCG. On the other hand SIGACT has a reasonable size and shares a large portion of the interests of SoCG. What annoys me is that our relation with ACM very much depends on who is in the SIGACT board. If we get somebody taking care of things, then things will work out. I think that if the result turns out to be “staying with ACM”, we should also try to increase the involvement of the SoCG community into the SIGACT board. In particular, for the last elections I do not recall seeing any CGer in the list. (Not blaming SIGACT, but blaming our lack of interest to be involved.)

  2. I can believe that relations between STOC and ACM are good, I don’t now why SoCG had such a different experience.
    I can see that STOC was always held in the USA, except for a few times in Canada, and once in Greece ( Not sure whether it is enough to explain.

    It may look rude, but to me, these 4500$ look like what we call “electoral tarring” in French (“goudronnage électoral”): a few months before elections, some mayors who have not done much for their city during their mandate suddenly order visible improvements in the streets, in order to get re-elected. Elections in France are often depressing…
    At least here we don’t get electoral promises, so there is no disappointment to expect.

  3. Jeff Erickson says:

    ACM provides a substantial net financial benefit to the TCS community. SIGACT gets money from ACM for its share of the digital library revenues based on downloads of SIGACT material.

    This claim strikes me as a bit backwards. Surely “the TCS community” also includes the people who currently pay ACM to download SIGACT-sponsored digital library papers!

    SIGACT is trying to do the right thing with their unexpected surplus, and they should be applauded for their efforts. (To counter Monique’s suggestion of goudronnage électoral, they are also offering similar support to other SIGACT conferences.) But I would have been much happier if Paul had argued, with evidence, that SIGACT and ACM are trying to make everything less expensive in the first place. If STOC is running that large a surplus, it should be cheaper. If Digital Library downloads are bringing SIGACT that much income, they should be cheaper. Similarly, if sending paper copies of SIGACT News is so expensive, SIGACT should make it electronic-only (and open access while they’re at it)!

    After all SIGACT has committed to pay for extra costs in this case so it is reasonable that SIGACT get asked each time

    If SOCG declined SIGACT’s offer to pay those extra costs, could SIGACT offer a more definitive guarantee of in-cooperation status instead, or would it still require us to ask every time?

    ACM has an arrangement with RegOnline…that halves the costs

    RegOnline’s normal price ( is $4 per registrant, so any savings from this 50% discount are insignificant, especially in comparison to the overhead ACM charges for sponsored conferences.

    SoCG leaving ACM would eliminate the possibility of future “in cooperation” status with ACM

    The ACM conference guidelines say “Such cooperation is to be primarily for publicity and to encourage Association members to participate by contributing papers and attending technical sessions.” Would SIGACT not encourage its members to participate in an independent flagship computational geometry conference? Would announcements about SOCG++ no longer be welcome in SIGACT News? Or is ACM merely ruling out the future possibility of publishing SOCG++ proceedings?

  4. I think that like other people I am tired of the topic. In the end, ACM failed to provide simple clear answers, and there is no guarentee that a few years down the line the ACM would not revert back to its old policy. While I like the ACM DL, and their journals, I feel this is tangential to the topic under discussion.

    BTW, I find the ACM open access scheme appalling (i.e., authors pay to make their papers available), and completely unnecessary, as the arxiv does the same for free. The ACM seems to be in their own parallel universe. Too bad.

  5. Pingback: For Me, It’s About Open Access | Making SOCG

  6. some body says:

    I ended my ACM membership a few years earlier because of their copyright policies. ACM is a big organization and despite how Paul might feel is mainly ruled by breaucrats whose main objective is preservation of ACM independent of whether it’s beneficial for its members. Convincing them to part with their golden hen seems impossible.

    Paul, do you really think it’s beneficial for TCS community and science to continue with ACM’s closed DL? I for one would like to see a break down of ACM incomes and expenses. My guess: most of it is out of DL which WE pay indirectly through our libraries. Now should we thank ACM for handing out a bucks back to us?

    • Jeff Erickson says:

      You can find more than a decade of ACM’s annual reports at ; these include high-level breakdowns of income and expenses for each year.

      In FY 2013, for example, ACM’s largest sources of revenue were conferences ($25M), publications ($20M), and membership dues ($8.5M), and its largest expenses were conferences ($23M), publications ($11M), and general administration ($11M). During that year, 150,000 bibliographic entries, 30,000 full-text articles, 465 proceeding volumes, and 1260 books were added to the Digital Library.

      • some body says:

        $11M for publication expenses is ridiculous at this age. Even so by their numbers they earn $2M from conferences, $9M from DL and publications and $8.5M from members and they spend $11M on bureaucracy. We should create a new AMS-like society for CS.

  7. some body says:

    Also for comparison, I think AMS does a much better job than ACM in serving its community.

  8. To be fair to ACM, publicity is worth something. I’m thinking of, but won’t name, some conference series in another discipline, where successive conferences are largely independently organized w/o coordination with either the loosely sponsoring society or the previous conferences. The result is that they don’t announce the next conference even to attendees or presenters at the previous one. Also, each conference’s website is independently organized and not guaranteed to remain online.

Comments are closed.