The 21st-century academic conference: Global, inclusive, low-carbon, semi-virtual

Richard Parncutt 

2021 April


Contrary to persistent images of lone geniuses working late at night among dusty books or in stuffy laboratories, scholarship and research have always been social enterprises. In both humanities and sciences, new knowledge is created, and new insights revealed, when experts with contrasting backgrounds get together to solve big problems. The place where people bounce the most original and promising ideas off each other is the international conference. 

Until a few years ago, we academics thought nothing of flying from anywhere in the world to a single location and enjoying a few days of intense communication.  We returned home with a new sense of purpose. For many years, I enjoyed conferences of that kind. For me and many of my colleagues, conferences were our career highlights.

Corona and climate change have put an end to that. Corona happened more suddenly, but climate change is more important. The global climate crisis is a matter of life and death for a billion people, mostly in developing countries (Parncutt, 2019), and it will almost certainly get gradually worse in coming decades (IPCC, 2018). Although some colleagues still seem to be in denial about that -- perhaps the clearest evidence being their willingness to again consider long international flights -- the naked truth is that the golden age of carefree jet-setting is over.

"Back to normal" is not an option. To save what can still be saved of the world's climate and biodiversity for future generations, emissions in all areas must now be drastically and urgently reduced.
Anyone can convincingly argue that they deserve special treatment, but the truth is that almost nobody does.

Besides, flying may represent the biggest single contribution to climate change of both academia as a whole and individual academics. Quinton (2020) commented that "for many academics, the carbon emissions associated with air travel dominate personal carbon budgets, dwarfing other contributions such as from driving or eating a meat-based diet."
Flying to an international conference in economy class burns roughly a tonne of carbon -- comparable with driving a typical car in a typical way for a year, or eating meat in a typical way for five years.

That being the case, it is surprising how little most universities seem to know about the carbon emissions of their own researchers. At the University of British Columbia, Canada, emissions from flying were estimated as 63-73% of total university emissions (Wynes & Donner, 2018). Ahonen et al. (2021) surveyed Finnish universities and confirmed that flying made up roughly half of total emissions for the average university, but with a lot of variation -- from 10% (University of Helsinki) to 78% (Hanken School of Economics) (see their Figure 14). Aalto University calculated independently that air travel contributed 58% of total university emissions in 2019.

While these realisations initially sound disastrous, they could be blessings in disguise. In forcing us to rethink the purpose and format of academic conferences, they may result in conference formats that improve on all traditional and current formats.

Pros and cons of virtual and hybrid conferences

The most common solutions currently being considered in most academic disciplines are the fully virtual conference and the hybrid conference. While both have their pros and cons, relative to the traditional single-location conference both are deeply problematic.

The virtual conference has the advantage of reaching anyone in the world, regardless of financial means or mobility. That includes colleagues in non-wealthy countries or students who cannot afford a regular international conference. It also includes colleagues with caring commitments or disabilities. For all these groups, a fully virtual conference is like a breath of fresh air. But a virtual conference also means sitting for hours at home or in the office in front of a laptop, often at strange times (depending on international time differences). After a year of corona, people are understandably exhausted ("zoomed out"). Besides, organizers of virtual conferences can expect participants to attend only a small fraction of the talks they would otherwise have attended, and to be constantly distracted by whatever is happening where they physically are, new emails, or social media.

The hybrid conference is a conference at a central location, to which some participants travel and the rest stay home and participate virtually. The advantages are obvious: the conference is open to anyone in the world, regardless of mobility. If the fee for virtual participation is low, almost anyone can afford it. But there are significant problems. The first is that the physically present have a very different experience from the absent ("virtual") participants. The physically present have a great time meeting and discussing all kinds of things, as at a traditional one-location conference. In addition, they have the benefit of virtual presentations from colleagues who are unable to travel to the conference. The virtual participants spend the whole conference looking at their laptops, often at very strange hours. No matter how hard we try to compensate for that with advances in virtual meetings and virtual socializing sessions, the virtual participants still have a massive disadvantage.

And that is not all. From a sustainability viewpoint, hybrid conferences motivate colleagues to fly at a time when we should be encouraging people to stay grounded. They are like an experiment in behavioral psychology that aims to train people to fly by providing salient emotional rewards. From the viewpoint of inclusion, hybrid conferences systematically discriminate against those with less money, those with caring commitments, and those with disabilities. Whereas they are a step forward if they include those colleagues for the first time, the second-class experience of virtual participants at hybrid conferences is reminiscent of institutionalized discrimination or even colonialism.

The semi-virtual conference

This article is about a third,
"semi-virtual" option: a conference that is widely distributed across several global locations (Coroama et al., 2012; Parncutt, 2020). This model has considerable potential that, so far, has barely been appreciated -- perhaps because it challenges both organizers and participants to think rather differently about conference logistics.

A conference can be "semi-virtual" in two senses. First, it can involve face-to-face communication with some participants and virtual communication with others. Second, participants can repeatedly choose between parallel real and virtual presentations. Unlike a hybrid conference, where remote participants may feel like second-class citizens (given that, for them, face-to-face interaction is not an option), all participants at a semi-virtual conference are treated equally.

The semi-virtual model solves the main problems of virtual and hybrid conferences as follows:
The main advantages of a semi-virtual conference are:
Klöwer et al. (2020) considered distributing a conference across three specific locations -- Chicago, Tokyo and Paris -- for future international conferences of the European Geosciences Union, the Japan Geoscience Union, or the American Geophysical Union. The locations were chosen to minimize emissions for expected participants. But there are still problems with such a model, as the authors themselves noted:

To create an optimal format for conferences in the 21st century, it helps to think again about why we have conferences, what they are for, and what we expect from them. What criteria should an international academic conference fulfill? Most academics will agree about the following points:
Creating a practical solution

At first glance, the above criteria seem to contradict. How can a conference with physical face-to-face contact be open to anyone in the world regardless of financial means? How can emissions be reduced while at the same time improving the conference experience?

The solution is to appropriately mix physical and virtual interaction, using modern audiovisual communication technologies:

Meeting these criteria may be a non-trivial logistic problem, but a clear conclusion can be drawn immediately: the time difference between adjacent hubs should be 8 hours or less. Therefore, there should be at least three hubs, roughly equally spaced around the globe. Beyond that, there is no limit to the number of hubs. Three primary hubs, 8 hours apart

A promising approach is to set up the hubs in two stages. First, establish three nominally equal primary hubs that are equally spaced around the globe relative to time zones, exactly 8 hours apart. Second, invite colleagues to establish additional secondary hubs, anywhere at all. (Note that the terms primary and secondary are not value judgements. From the point of view of the conference program or the conference experience, all hubs are equal.)

The placement of the primary hubs is largely determined by the
many time zones spanned by the Pacific Ocean. For practical reasons, two of the three primary hubs must be located near the Pacific rim: in Japan or Eastern Australia on one side, and on the West coast of North America on the other. That in turn limits options for the third main hub, which can only be in Europe or Africa. To check that the three chosen hubs are exactly 8 hours apart at the time when the conference is taking place, use an international meeting planner such as

Many academic disciplines have separate societies in North America, Europe, and Asia. The three geoscientific societies mentioned above are good examples. In my discipline, music cognition, we have APSCOM in the Asia-Pacific region, SMPC in North America, and ESCOM in Europe. A conference with three primary hubs is ideal for such disciplines: each regional society can be responsible for one of them.

The exact timing at the primary hubs might be as follows.
To ensure that adjacent primary hubs can work together in real time, evening sessions must begin exactly 8 hours after morning sessions. Daily working hours might therefore be 8h-12h and 16h-20h, local time. Breaks and social contacts are important, so the first half hour of each two-hour slot is for virtual socializing. The actual sessions, then, are limited to 8:30-10, 10:30-12, 16:30-18, and 18:30-20. If people are encouraged to participate in virtual socializing, they will be seated in time for the first talk of each 90-minute timeslot. That might comprise three regular talks of 30 minutes or a keynote of 60 minutes followed by a discussion of 30 minutes. Each conference will divide the 90-minute timeslots in different ways, depending on aims, content, and tradition.

Once the primary hubs are established, colleagues are invited to propose secondary hubs at any global location. Secondary hubs are equal to primary hubs in every way except, in most cases, for convenience of working hours. The program at each secondary hub is simultaneous with the program at the temporally closest primary hub. Participants at most secondary hubs shift their daily schedule, getting up and retiring either earlier or later than usual during the few days of the conference -- a kind of mini-jetlag.

With three equally spaced primary hubs plus any number of secondary hubs, almost any qualified colleague, anywhere in the world, can participate fully. Those who can travel to the nearest hub do so; they are treated equally and enjoy the same conference experience, depending only on the organization of their local hub and the number of people attending at that hub.
Those who stay at home due to caring commitments or immobility have virtual access to the entire program  in real time -- similar to a hybrid conference. They do not enjoy the same conference experience as those attending a hub physically, but the conference does offer them the maximum possible benefit given the practical limitations.

New hubs are hubs in locations that have not previously hosted an international conference in the academic discipline in question. They may be either primary or secondary. New hubs are opportunities to create new local and regional academic societies and to promote the discipline of the conference at regional universities. This can have interesting implications for other academic disciplines. The international contacts that are created during the conference can contribute positively to international development and be seen as a form of international aid. Neither virtual nor hybrid conferences have this potential. If a single-location conference is moved to an "exotic" location, the advantage for colleagues in that location is offset by the environmental disadvantage: the carbon footprint of such a conference is typically even higher than that of a traditional conference.


The discussion has so far been rather theoretical. To evaluate and implement this proposal, specific hub combinations need to be considered in more detail. How would the 24-hour global program work in specific cases? How would each local program work?

Unfortunately, the amount that we can learn from past conferences is limited, because to my knowledge, no past conference has ever satisfied the above criteria. At the conference that I organized in 2018 (ICMPC15/ESCOM10), the time difference between the Sydney and Montreal hubs was more than 8 hours, with the result that the program in both locations was less than satisfactory. Each of these hubs was isolated from other hubs for a few hours each day. My attempts to convince my colleagues that we needed a hub on the West coast of North America,
while planning the conference in 2016 and 2017, had unfortunately failed.

Consider a conference taking place in July 2021. (We could also consider July 2022, but there is some uncertainty about whether clocks will still be changed in March 2022, as they were in March 2021.) The following options for primary hubs are promising, being exactly 8 hours apart. To check the exact time zone differences, visit

Tokyo, San Francisco, London
(or other locations in the same time zones). What if colleagues organized the three primary hubs in those cities? Their time zones are exactly 8 hours apart in July 2021. Let's say the program in those locations is exactly 8:30-12 and 16:30-20 daily. So the morning session at one hub always exactly coincides with the evening session at another. So far, so good, but what about secondary hubs? New York is 3 hours ahead of San Francisco, so the program there would be 11:30-15 and 19:30-23. Those hours are rather inconvenient, so the East-coast US hub would be better located in Chicago, which is only 2 hours ahead of San Francisco (and relatively well served by train and bus connections).
Beijing would be 1 hour earlier than Tokyo - no problem. There would be an additional central European hub with the program happening 1 hour later than London, and perhaps also Tel Aviv, 2 hours later. There would be plenty of other good opportunities for hubs, but also several locations that do not work so well. Delhi would be 3.5 hours earlier than Tokyo, or 4.5 hours later than London, both of which are very inconvenient. But Indian colleagues may be so happy to be included in the international conference for the first time that they deal with it anyway. Alternatively, they might travel to a hub in Beijing. Rio de Janeiro would similarly be very inconvenient, exactly 4 hours later than San Francisco or 4 hours earlier than London. Bogota would be a better location for a South American hub -- 2 hours later than San Francisco.

Sydney, Calgary, Warsaw. In this case, an additional hub in New York would be more feasible -- only 2 hours ahead of Calgary.  A hub in Delhi would also be more comfortable - only 2.5 hours ahead of Warsaw. Rio would still be 3 hours ahead of Calgary, but Bogota would be only 1 hour ahead. Beijing would be 2 hours earlier than Sydney, making for some early rising (or travel to a Japanese hub).

Auckland, New York, Yerevan. This option is not promising for San Francisco or London, each being 3 hours earlier than the closest hub. Colleagues on the West coast of the USA could travel to the midwest (Mountain Time), which is 2 hours earlier than New York. Colleagues from the UK could travel to a central European hub that is 2 hours earlier than Yerevan. Delhi would be comfortable, being 1.5 hours later than Yerevan, but Beijing would be a poor hub location, being 4 hours later; Chinese colleagues might consider attending the Indian hub. Sydney would be 2 hours earlier than Auckland, which is manageable.

The above solutions are for summer in the Northern Hemisphere. They are also for current daylight saving regimes, which could change at any time. At other times of year, and with changes in daylight saving, other options will emerge. To take advantage of different options and treat different locations equally, the date of a recurring international event could be changed back and forth from winter to summer.

Four primary hubs, 6 hours apart

The working hours in the 3-primary-hub approach can be strenuous, especially if shifted forwards or backwards by 1, 2, 3, or even 4 hours. What about a solution with 4 primary hubs separated by 6 hours? One possibility in the month of July is Denver, Accra, Dhaka, and Auckland. Another is Honolulu, Boston, Frankfurt, and Beijing.

Having chosen four primary hubs, the daily schedule at each can be shortened to last from 9h-12h and 3h-6h local time. Each primary hub then communicates with the two neighboring primary hubs in real time, one in the morning and the other in the afternoon -- but never with the primary hub on the other side of the world. In this solution, each hub communicates with about 3/4 of the other hubs in real time, whereas in the 3-hub solution above, every hub communicates with every other hub.

The logistic advantage of the 4-primary-hub approach is shorter working hours, which make it easier to shift working hours at each hub forward or backward, leading to more flexible solutions. The disadvantage is the inability to communicate with all hubs in real time, which from my perspective is the most important point, and the reason why I favor the 3-primary-hub approach.
Practical tips
Room fees. Some universities are charging high fees for room rent for large lecture theatres -- those which are suitable for large international conferences. In this semi-virtual conference model, the hubs are smaller, so hub organizers will need to cater for a smaller number of participants. For that purpose, regular smaller teaching rooms may suffice, for which little or no rental is payable.

Technical support. Each hub needs a technician (for example, a master's student in audio engineering). The head technician at the main organizing hub prepares detailed guidelines and works together with the other hubs individually to ensure that procedures are understood and followed. That includes responding quickly to possible problems (trouble shooting) without delaying the program, which must in general run exactly on the time, to the nearest minute. In addition, every hub rehearses independently with every other hub.

Budget. The budget at each hub is similar to the budget for a regular conference at that location, based on the number of physically present participants. Extra costs are mainly for technical assistants, and can be covered in part by charging a small fee to virtual participants (those who do not attend any hub). The main expense is wages for the hub's head technician, who works for about a month (starting, say, three weeks before the conference) to ensure that all hardware and software operates properly. One or more additional technicians are needed for the duration of the conference -- one for each conference room in which live or virtual talks are held. These costs are relatively small compared to those of
the organizing hub, which organizes abstract submission and peer review on behalf of all hubs. (The organizing hub will need to employ a conference manager for several months before and during the conference.) The extra costs at each hub are also moderate compared to room rent, food and refreshments (including reception and banquet, if applicable), printing, conference bags and other giveaways such as cups, and bank fees. Equipment costs are low if the hub is hosted by a university that makes available the necessary computers, data projectors, loudspeakers, and cable and wireless internet connections.

Technical reliability. At a conference organized by the author in 2018, we optimized audiovisual quality and reliability by simultaneously running one-way and two-way communication software in each room throughout the conference. One-way software (e.g. YouTube live streams) offers high audiovisual quality with a time delay of a few seconds to a minute, whereas two-way software (e.g., Skype or Zoom) offers quasi-instantaneous communication. Each room has a technician who, at the required moment, switches between one-way communication for lectures and two-way for discussion, changing only what images are projected and what sound is heard. Each system provides a backup for the other, should something go wrong (Parncutt et al., 2019). Since the technology is constantly changing, this may no longer be the best solution, but it is an interesting benchmark with which to compare current alternatives. There is currently great potential for imaginative software engineers with experience of different conference formats to create a new software tool that offers conference organizers a flexible platform for organizing and presenting semi-virtual conferences, bringing together all currently promising software and allowing new developments to be included as they become available.

Avoiding acoustic feedback. All speakers wear headsets (cheap mobile phone headphones may be good enough) and turn the sound off when not speaking to avoid acoustic feedback. Rehearsals happen in the same rooms with exactly the same equipment as the conference itself.

Optimizing internet speed. Each hub does advance research on internet speeds in different rooms at different times of day, comparing wifi with cable connections, to avoid unpleasant surprises.

Avoiding legal problems. Even if presentations are available only to registered participants, internet security is never 100%, so all presentations are quasi-public, carefully avoiding possible legal problems in areas such as copyright or defamation.

Real-time versus recorded presentation. Pre-recorded videos (followed by real-time discussion) have the advantage of technical reliability and the disadvantage of lack of presence and spontaneity. A mixture of live and pre-recorded presentations is possible.

Additional information. Cloud space is made for every presenter to make additional electronic materials available to othe conference participants about her or his talk.

Improving conference documentation. Traditionally, conference documentation is confined to the program, abstracts, and proceedings. Today, researchers are increasingly relying on recorded videos to learn about the research of others and interact with each other. Conference videos are also becoming an increasingly important resource for students. After the conference, participants put their videos in the internet or link them to their homepages, increasing outreach. 

Virtual socializing

Every two-hour slot can begin with 30 minutes of virtual socializing, comparable with conventional coffee breaks. During that time, any conference participant can contact any other participant at any location toward the East in the morning and toward the West in the evening.

Each hub can have a room devoted to virtual socializing with several computers, each with three sets of headphones and one microphone. Different virtual socializing events can be organized in advance and included in the program:

One student assistant per hub can be put in charge of virtual socializing, and these virtual socializing experts can meet several times virtually before the conference. All participants may be asked which social events they would like to attend and whether they are happy to provide confidential information that will help them to be matched up with colleagues of similar interests. 

Virtual poster sessions are another opportunity to socialize while at the same time getting to know new research, and they can be organized in various ways that need advance planning.


Aalto University (2020). Sustainability Report of Aalto University 2019.

Ahonen, V., Siljander, M., Pellikka, P., Johansson, T., & Rask, M. (2021). The sustainability of academic air mobility in Finnish universities. Sustainability, 13, 2948.

Coroama, V. C., Hilty, L. M., & Birtel, M. (2012). Effects of Internet-based multiple-site conferences on greenhouse gas emissions. Telematics and Informatics, 29(4), 362-374.

IPCC (2018). Special report: Global warming of 1.5°C.

Klöwer, M., Hopkins, D., Allen, M., & Higham, J. (2020). An analysis of ways to decarbonize conference travel after COVID-19 (comment). Nature, 583, 356-359.

Parncutt, R., & Seither-Preisler, A. (2021). Live streaming at international academic conferences: Doing rather than talking. Elementa: Science of the Anthropocene, 9(1), 00435. 

Parncutt, R. (2020). Grounding the ICMC. Computer Music Journal, 43(4), 7-11. 

Parncutt, R., Meyer-Kahlen, N., & Sattmann, S. (2019). Live-streaming at international academic conferences: Technical and organizational options for single- and multiple-location formats. Elementa: Science of the Anthropocene, 7, 51. 

Parncutt, R., & Seither-Preisler, A. (2019). Live streaming at international academic conferences: Ethical considerations. Elementa: Science of the Anthropocene, 7, 55. 

Parncutt, R. (2019). The human cost of anthropogenic global warming: Semi-quantitative prediction and the 1000-tonne rule. Frontiers in Psychology, 10, 2323.

Quinton, J. N. (2020). Cutting the carbon cost of academic travel. Nature Reviews Earth & Environment, 1(1), 13-13.

Wynes, S., & Donner, S. D. (2018). Addressing greenhouse gas emissions from business-related air travel at public institutions: A case study of the University of British Columbia. Pacific Institute for Climate Solutions.

The opinions expressed on this page are the author's personal opinions. Readers who know and care about this topic are asked to contact the author with suggestions for improving or extending the content: parncutt at uni-graz dot at. Back to Richard Parncutt's homepage