In this post, Stijn van Ewijk writes about the environmental impact of academic conference travel, options for reducing emissions, and priorities for post-COVID conferencing.
Our lunch at the Corner Bakery opposite the Art Institute of Chicago was one of the most enjoyable and productive moments of the conference. It was 2017, my colleague Paul and I had both flown across the Atlantic, and indoor dining was perfectly normal. Over soup and a sandwich, we wondered about the climate impact of the 625 conference attendees’ journeys.
The conference we attended was the 9th instalment of the biennial conference series of the International Society for Industrial Ecology (ISIE). The industrial ecology community is concerned foremost with the environmental impacts of human activity. If anyone, we figured, our community should be able to estimate and reduce its emissions, including from air travel.
The subject was of personal interest to us, partly because we don’t live in our home country, which tends to increase travel. We are both Dutch, but Paul Hoekman had travelled from Cape Town, South Africa, where he heads the global network Metabolism of Cities. I had travelled from London, United Kingdom, where I was doing my PhD at University College London (UCL).
The Chicago lunch planted the seed for our recently published study on reducing conference travel emissions. Though the work always remained an unfunded side-project, the subject matter proved rich enough for a peer-reviewed research article. It was a rewarding effort because many of our colleagues were genuinely interested. It was, after all, about them.
The questions we tried to answer were simple. How much greenhouse gas emissions are caused by travel to the ISIE biennial conferences? How can we reduce these emissions? The literature often suggested many options, including taking the train or having multiple video-linked events, but none of the studies directly compared all the options.
We started our research by collecting attendance data for the Chicago conference, and for the two preceding conferences in 2015 in Surrey, United Kingdom, and in 2013 in Ulsan, South Korea. Based on the attendees’ affiliations, we estimated the travel distances, likely modes of transport (plane or land transport), and greenhouse gas emissions.
We found that travel emissions ranged from near-zero for local attendees to about 5 tons of carbon dioxide equivalent (tCO2e) for travellers crossing half the globe. The average travel emissions ranged from 1.3 to 1.8 tCO2e, depending on the conference. These figures are very substantial compared to the average annual global emissions per capita of 7 tCO2e.
Trains and taxes
We looked into a range of options for reducing conference travel emissions (see figure below). To our surprise, the most obvious solution was the least effective. Greater use of land transport instead of air travel could reduce travel emissions by at best 5%. Even worse, it could increase emissions when attendees rely more on cars than trains.
Train travel can strongly reduce emissions for an individual traveller. However, most attendees of the global conferences we studied travelled much too far to take the train. For conferences with an audience from a single continent, land transport is feasible for a greater fraction of the attendees; previous studies suggest an emission reduction potential ranging from 10% to 33%.
Next, we studied a carbon tax, an instrument often touted essential to address climate change. Yet, a tax of up to 100 $/tCO2e may reduce emissions by only at most 14%, and perhaps as little as 4%. The uncertainty range is large because travel behaviour depends on more than just cost. If anything, a tax will mostly deter low-income travellers, including students.
The most promising options
The most effective solutions cut out the longest flights, which has an outsized effect on total travel emissions. An ‘augmented’ conference that convinces the top 10% of emitters to attend virtually or not at all, would reduce emissions by 20-30%. For a conference series that alternates locations, the top 10% would consist of different people each time.
A multi-site conference with simultaneous video-linked events achieves even higher reductions (see also Joshua King’s post). In our estimate, a two-site conference cuts emissions by 25-50%, a three-site conference by 46-75%. We estimated this from the travel data of the Chicago, Surrey, and Ulsan conferences, assuming the multi-site alternatives would use two or three of the same locations.
The multi-site conference works because it strongly reduces the average travel distance. Besides, unlike the augmented conference, it does not bar anyone from attending in-person. Shorter travel distances also enable car or train travel: a three-site conference combined with greater use of land transport can reduce emissions by up to 82%.
Benefits and barriers
It will take more than abstract environmental concerns to change conventional conferences. Fortunately, there are many potential co-benefits to the alternatives. Fully or partially virtual events can be more accessible to researchers with limited budgets or family responsibilities, and recordings can avoid the ordeal of choosing between parallel sessions.
There are barriers too, such as false ideas about the effectiveness of interventions that do not target air travel. Vegetarian or vegan lunches would be great indeed, and we should all walk or cycle between the hotel and the venue, but neither will make a dent in the total conference emissions when most attendees burn hundreds of litres of kerosene.
Technological fixes for aviation are not a safe bet either. The mass production of aviation-grade fuels from renewables is an unresolved technological challenge, and sourcing sustainable biomass is very problematic. Electric planes will not be widely available before 2050, the year emissions should be near-zero according to climate change mitigation targets.
The greatest barrier to change is the potential loss of in-person contact. Our study would not have happened without lunch in Chicago and a poster presentation two years later at the ISIE conference in Beijing, China. Because of COVID-19, I have now attended several virtual events, but even the best ones kept me longing for traditional conferences.
Our bet is on multi-site conferencing because it achieves very substantial emission reductions and maintains much of the in-person contact. There are other opportunities that we did not research but consider promising, such as holding fewer conferences, alternating virtual and in-person events, and aligning the time and location of events with overlapping audiences.
The next ISIE conference should have been in the summer of 2021 in the Netherlands, and I looked forward to combining it with visiting friends and family. Like many other events, it was postponed, and the organisation of a virtual replacement is in progress. If COVID-19 brought us anything good, it may be a greater understanding of the alternatives to conventional conferencing.