- Get advice on "cheap" cities; Minneapolis in winter (site of many IETF meetings) was chosen for this reason. Nobody can accuse the attendees of a junket in that case, either. In my limited experience, the cheapest cities are 2nd tier cities (i.e., *not* London, NY, Paris, Beijing, Delhi, Tokyo) served by multiple air carriers, in the off-season. They usually offer a range of hotel options within walking or mass-transit distance, while still being easy to reach without three or four air-hops or extensive bus and train trips (or car rentals). Site proposals should provide an indication whether <$100 hotel options are within easy reach.
- Conferences should announce their rough fees (say, +/- $50) at the time of submission, not after papers have been accepted. That way, I can tell my students not to submit a paper to a $1000/attendee conference.
- Have conferences publish their budget, at least in outline form.
- Ask attendees whether they prefer the social event to be 'a la carte', i.e., as an option. (You'll divide attendees into first-class and steerage, but, as Joe keeps pointing out, you can't ask for all the goodies and then complain about the price.)
- Establish a rough conference fee guideline (something like "should be no more than $200/day"), and make it part of the TC approval process.
- Consider whether combining multiple events in one venue can help reduce travel expenses. SIGCOMM and Infocom are doing this to some extent, for example.
- Consider two-day instead of three-day events, e.g., with the help of poster sessions. (You could imagine a bidding process: if you speak as an author, you have to spend points, and you get some number of points from the audience; if you don't have enough points, your next presentation is a poster -- I'm [somewhat] kidding here.)
- Establish clear guidelines on volunteer (chairs, steering committee, staff) and keynote speaker reimbursements/honoraria.
- Where feasible, consider universities or research labs for hosting events. (Labs often have good facilities, but the details can be tricky. We tried to host a mid-sized events at a Seattle-area lab, but the organization charges for the larger lecture rooms, and hotels in the immediate vicinity were scarce and would have required attendees to rent a car. All the major labs in the US that I can think of are in the suburbs, from Bell Labs, AT&T and IBM to Microsoft.)
I suspect that the biggest expense is not each conference, but the number of events that people are supposed to attend. My gut feeling is that the standard "custom" for academics a decade or two ago was that you went to one discipline-wide event a year, where you met all your colleagues, and one specialty event (a workshop in your narrow research specialty).
As I mentioned in an earlier message, all of these are tweaking. Personally, I think we need to reconsider how we run things on a larger scale, rather than worrying about the number of coffee breaks at our conferences.
Henning
On Dec 6, 2009, at 2:49 PM, Bijan Jabbari wrote:
> Such a review of conference fees is very timely, given the majority of
> attendees in our events are from academia and the economic hardship
> universities in many states in the US and internationally are facing, at
> present. In particular, many universities have increased tuition and
> other fees and the increments have to come from faculty research
> budgets, reducing available funding for other areas such as conferences.
>
> I believe the subject is very important and it would be good if an ad
> hoc committee at the IEEE COMSOC management level could be created to
> look into how to make the IEEE conferences more financially accessible
> to its membership.
>
> - Bijan
>
>
> Joe Touch wrote:
>> -----BEGIN PGP SIGNED MESSAGE-----
>> Hash: SHA1
>>
>>
>>
>> Roch Guerin wrote:
>>> Joe,
>>>
>>> Points well-taken. On the other hand "not for profit" and "fiscal
>>> responsibility" are two different concepts. I would argue that many
>>> conferences could run a tighter budget and pass the savings on to
>>> attendees.
>>
>> As I noted, I have quite a bit of direct experience with budgets of
>> specific meetings in both the IEEE and ACM.
>>
>> The basic costs are 75% food (or more), and the rest is *very* hard to
>> minimize.
>>
>> The bottom line is that we could reduce the cost IF we delete *food*.
>>
>>> Let me make just one punctual and I believe representative comparison:
>>> - ACM CoNEXT 2009 early registration for (ACM and SIGCOMM) members:
>>> $525 (it was $525 last year - an increase of 0%, and btw so was it in 2007)
>>> - IEEE INFOCOM 2010 early registration for (ComSoc) members: $750 (it
>>> was $680 last year - an increase of over 10%, and it was $695 in 2008??)
>>
>> CoNEXT is 2.5 days.
>>
>> Infocom is 3 days, or 20% longer, just to start.
>>
>> Here's a food difference, using my previous estimates:
>>
>> Infocom = 410 food (est)
>> - 1 welcome reception 40
>> - 6 coffee breaks 6*25=150
>> - 3 lunches 3*40=120
>> - 1 banquet 1*100
>>
>> CoNEXT = 305 (est)
>> - 5 coffee breaks
>> - 2 lunches
>> - 1 banquet
>>
>> Finally, the IEEE model of overhead and contingency isn't exactly the
>> same as the ACM model.
>>
>>> Looking just at the latest numbers, this amounts to a $225 difference
>>> (about 40%) for essentially similar "services," and that cannot be
>>> simply brushed aside.
>>
>> Accounting for the 25% difference in "services", we could either say that:
>>
>> Infocom should cost 650, or is basically overpriced 15%
>> CoNEXT should cost 600, or is basically discounted 12%
>>
>> The difference can be the result of a number of different issues - e.g.,
>> the need to 'restore' buffer accounts in each organization, or other
>> overheads as you note below...
>>
>>> In particular,
>>> some conferences have a tradition of freebies for volunteers that I
>>> personally find distasteful, and that play no small role in raising
>>> their costs (e.g., CoNEXT has no free registrations or travel subsidies
>>> for anyone except local student helpers who receive a free student
>>> registration).
>>
>> Well, none given out from the conference, but there are subsidies given
>> out by the SIG - and that money comes from conference surpluses. For the
>> SIGCOMM conference, this includes the SIGCOMM award winner (who is
>> typically the keynote), the student award winner. There are also funds
>> for the GeoDiversity grants.
>>
>>> I believe it is time to take a hard look at what this is
>>> costing all of us.
>>
>> Keep also in mind the difference in the size of the meetings. CoNext had
>> 100-150 attendees in past years (the recent numbers aren't posted).
>> Infocom has 600. That means, effectively, that every 'free' registration
>> costs Infocom attendees only 1/4 what it costs CoNext - e.g., even if
>> Infocom gave out 20 free registrations, that's only about a $30
>> difference in the overall fees.
>>
>> As chair of TCCC, I'll be glad to look into this issue further and
>> discuss how your concerns can be addressed, though.
>>
>> Joe
>> -----BEGIN PGP SIGNATURE-----
>> Version: GnuPG v1.4.9 (MingW32)
>>
>> iEYEARECAAYFAkscAEoACgkQE5f5cImnZrvSfgCg7b07NXkjO1C5fXyEVTe7pn9W
>> lEQAn2/klFiZrZ9VfbTYlTgPnEhpxDoD
>> =9fEb
>> -----END PGP SIGNATURE-----
>> _______________________________________________
>> Tccc mailing list
>> Tccc@lists.cs.columbia.edu
>> https://lists.cs.columbia.edu/cucslists/listinfo/tccc
>>
> _______________________________________________
> Tccc mailing list
> Tccc@lists.cs.columbia.edu
> https://lists.cs.columbia.edu/cucslists/listinfo/tccc
>
_______________________________________________
Tccc mailing list
Tccc@lists.cs.columbia.edu
https://lists.cs.columbia.edu/cucslists/listinfo/tccc
Posts
No comments:
Post a Comment