How are TCS conferences adapting after COVID-19?
Before 2019, the question of conference travel and its sustainability had been on the radar of many for several years, but with little concrete progress. Then, in 2020, the COVID-19 pandemic started: it brought international travel to a standstill, left people stuck at home during lockdowns, and academic conferences adapted on very short notice. A widespread solution was to move events online, replicating their on-site format as much as possible. The pandemic demonstrated that large-scale online conferences were in fact possible, though they raised many questions, e.g., about timezones or the feasibility of online socializing.
The COVID-19 pandemic is not over, but in many countries the situation is slowly coming back to normal, so that in-person conferences are again a possibility. We are now at a unique time for the academic conference ecosystem: Should we go back to conferences the way they were before the pandemic? Or should we adapt them with some of the lessons learned during that time? Arguably, we should aim for solutions that are scientifically effective (including networking and socialization), while addressing key questions such as environmental sustainability (e.g., plane travel) and inclusivity (e.g., visa requirements, lack of funding, caretaking obligations, etc.).
To understand the present trends, I survey here the current practices of conferences in theoretical computer science. There are too many conferences to investigate them all, so I made an initial list by picking those that currently having a representative in the recently-started open-access theoretical CS journal TheoretiCS. (This is just an arbitrary list, and not meant to imply endorsement by TheoretiCS or endorsement of these specific conferences.) I consider for each conference the upcoming edition at the time of writing, and/or the latest edition.
Our key criterion is the following: can you submit and formally publish at the conference in the usual way, but without traveling to the conference (e.g., with a remote video presentation, possibly prerecorded); and is this possibility made prominently clear before submission (e.g., in the call for papers). This was typically disallowed by pre-COVID conferences, which required on-site presentations except for last-minute emergencies beyond the control of the intended speaker. In my opinion, this criterion is a core issue, because formal publications are an essential requirement of career progression, especially for early-career researchers; and because academic conferences are the preferred publication venue in many communities, thanks to their prestige and predictable reviewing time. Thus, mandatory in-person presentation means that people who cannot travel or refuse to travel are excluded from the mainstream publication system. This criterion also echoes Moshe Vardi’s call in CACM for optional travel, shortly before lockdowns.
I do not examine the question of registration fees: it is fine for now if remote attendance costs the same as on-site attendance. Specifically, I agree that registration fees can exclude prospective participants, but the issue of how to bear the costs of conferences has always existed and I believe it is distinct. I am also interested in whether travel-free attendance is possible, i.e., are the talks streamed, again with or without registration: though again I believe that this issue to be less central than that of publication.
A disclaimer: this is an initial list, compiled at the time of this writing, and based on my best understanding. It may misrepresent the policies of some conferences, or the reason justifying some policies. It only gives information about individual events, which may be at odds with long-term plans for a conference, e.g., as decided by steering committees. More generally, at this early stage after the pandemic, I understand that many conferences are still experimenting and figuring out a new model, so the idea is to document, not necessarily to argue about what is right or wrong. If I did not accurately reflect the policy of some conferences, or if you have other feedback, I would be glad to know. My eventual hope in the longer run is that such a list can be kept up to date within the TCS4F initiative.
Purely online conferences
These conferences took place entirely online, so they allowed remote presentation and attendance:
Conferences allowing remote presentation
These conferences are fully hybrid, i.e., they allow remote presentation, and also allow remote attendance (even though they sometimes discourage it):
- SoCG 2022, in Berlin and online, in June 2022.
- ICALP 2022, in Paris and online, in July 2022.
- COLT 2022, in London and online, in July 2022, which is hybrid though remote participation is discouraged: “The conference is hybrid in the minimal sense that we will enable authors to give their presentation-at-a-distance, and we will stream the conference and enable questions to be asked from a distance; but that is all. So on-line participation is possible, but support will be minimal: after over two years of pandemic, we want to very strongly encourage you to come in person if you can!” (source).
- MFCS 2022, in Vienna and online, in August 2022.
- CONCUR 2022, in Warsaw and online, in September 2022.
- FSTTCS 2022, in Madras and online, in December 2022, but “In-person presentation is encouraged” (source).
- ICDT 2023, in Ioannina and online, in March 2023.
- FoSSaCS 2023 (part of ETAPS), in Paris and online, in April 2023. “We plan ETAPS 2023 as an on-site conference; nonetheless, remote attendance and presentation will be made possible” (source).
There are also conferences which will allow remote presentation, but for which it is not explicit yet whether they will allow remote attendance:
Conferences with no explicit policy
These conferences will feature remote attendance, but the call for papers and website do not seem to indicate explicitly whether remote presentation is possible:
- ITCS 2022, in Berkeley and online, in February 2022. The policy for submissions to ITCS 2023 is also not explicit.
- FoSSaCS 2022, in Munich, in April 2022.
- PODC 2022, in Italy and online, in July 2022. The policy of PODC 2023 is also not explicit.
- SODA 2023, in Italy and online, in January 2023.
Conferences with exceptional remote presentation
These conferences only allow remote presentation for situations outside of the control of the speaker:
- STOC 2022, in Rome and online, in June 2022: “One author per paper is expected to attend the conference in person to present it unless there are strong reasons not to, including international travel restrictions due to COVID-19 or impossibility to travel for all the authors of the paper” (source).
- FCSD 2022, in Haifa, in August 2022: “One author of an accepted paper is expected to present it at the (physical) conference, unless Covid restrictions prevent travel” (source).
- TCC 2022, in Chicago, in November 2022: “TCC 2022 is an in-person conference. Speakers may be permitted to present live talks via Zoom only under extenuating circumstances, with permission of the program chairs” (source).
- LICS 2023, in Boston, in June 2023: “Remote presentations can be organized for authors who are not to be able to attend the meeting” (source).
- ICALP 2023, in Paderborn, in July 2023: “At least one author of each accepted paper is required to register for the conference, and all talks are in-person. In exceptional cases, there may be support for remotely presenting a talk.” (source).
Purely on-site conferences
These conferences have apparently resumed in a purely on-site fashion, with no online presentations or online attendance:
- CCC 2022, in Pennsylvania, in July 2022.
- LICS 2022, in Haifa, in August 2022.
- FOCS 2022, in Denver, in November 2022.
I am grateful to other TCS4F members, in particular Thomas Colcombet and Thomas Schwentick, for feedback and contributions to this post.