Is it true that, worldwide, people took 4 billion flights in 2018?

**Tl;dr:** Yes, 4 billion (or that order of magnitude) seems correct to me, after checking its source with other independent sources. The largeness of this number shocked me but it doesn’t change what has to be done about flying and long-distance travel with respect to climate change: reduce flying and long-distance travel, first by taking this step in our own lives and then by promoting the message to others. This is especially true of leisure/vacation trips.

Beware: this post is about to get into the nitty-gritty of data sources.

The International Air Transport Association (IATA) estimated that [nearly 4 billion “passenger journeys”](https://www.iata.org/contentassets/c81222d96c9a4e0bb4ff6ced0126f0bb/iata-annual-review-2019.pdf) were flown in 2018, where one passenger journey refers to one person taking a one-way trip (multiple flights with layovers count as one passenger journey). It seems that this estimate is based on data given by IATA’s member airlines, which apparently include [“290 airlines or 82% of total air traffic”](https://www.iata.org/en/about/).

Isn’t this a remarkably high number? If we convert that to 2 billion round trips, that’s like saying there was 1 round trip per 3-4 people on Earth in 2018. How does that square with my intuition that probably the majority of people have never stepped on the plane? Probably because certain people fly so often? I was suspicious of the 4 billion number so I checked the IATA report against independent sources.

The main IATA report claim I had the time to check was this: [“the US domestic market [is] where almost 590 million passenger journeys were undertaken in 2018”](https://www.iata.org/contentassets/c81222d96c9a4e0bb4ff6ced0126f0bb/iata-annual-review-2019.pdf). “Domestic market” means that 590 million is counting only journeys that started and ended in the US.

First, I checked this against the major US airlines’ [10-K’s](https://en.wikipedia.org/wiki/Form_10-K). These are the forms that companies send to shareholders each year, required by the Securities and Exchange Commission, which often contain important data on those companies’ revenues. For the airlines, they contain numbers of passengers and flights. [There are 18 airlines in the US with revenue greater than $1 billion annually](https://www.bts.gov/sites/bts.dot.gov/files/docs/explore-topics-and-geography/topics/airlines-and-airports/227546/number-332-air-carrier-groupings-2020.pdf): Southwest, United, American, Delta, Jet Blue, etc. A mostly domestic airline, Southwest reports in [its 10-K](http://otp.investis.com/clients/us/southwest/SEC/sec-show.aspx?Type=html&FilingId=13882031&CIK=0000092380&Index=10000#LUV-12312019X10K_HTM_S33F2C3F2F70F5BFBB619FCBA5C426869) about 150 million passenger flights in 2019 (I took a midpoint between the reported 134,056,000 revenue passengers carried and 162,681,000 enplaned passengers). American Airlines reports 241,252 million revenue passenger miles in [its 10-K](https://americanairlines.gcs-web.com/static-files/d46a00e3-db05-4a91-af7a-fbe0fc2a7f08); by my estimate, this is equivalent to 276 million passenger flights (I scaled Southwest’s “150 million passenger flights” figure by a ratio of AA:Southwest’s revenue passenger miles). And so on for the rest of the airlines. One might start thinking the 590 million number is low because 150 million + 276 million + … for several more airlines quickly adds up to more than 590 million, but (1) note that 590 million counts passenger _journeys_, which can contain multiple passenger flights as legs of a journey with layovers, while 150 million, 276 million, etc. count passenger flights; (2) I started with the largest airlines so I expect the remaining terms in that sum to decrease quickly; (3) 590 million is only meant to count journeys starting and ending in the United States, whereas there airlines’ flights will also include those to and from international destinations; and (4) scaling AA and other airlines’ revenue passenger miles flown by the Southwest ratio might overestimate the non-Southwest passenger flight numbers if those airlines tend to be international carriers because those flights will have more miles per flight.

However, it is possible that the 590 million number is too low: if we extrapolate from [an Ipsos/Gallup survey of 5,048 American adults from their online panel](http://airlines.org/wp-content/uploads/2018/02/A4A-AirTravelSurvey-20Feb2018-FINAL.pdf), in which the average adult flew 2.5 round trips in 2017, we get [253,768,092 American adults](https://www.reference.com/world-view/many-adults-live-usa-b830ecdfb6047660) * 2.5 round trips per adult * 2 passenger journeys per round trip, or approximately 1.3 billion passenger journeys by American adults. The survey found that 2/3 of these trips were domestic; extrapolating that, we get about 850 million domestic passenger journeys, which is the same order of magnitude as 590 million. It may also be useful to note that the 590 million number [is probably a difficult estimate](http://www.intervistas.com/downloads/CAIR/articles/06_jun2008_b.pdf) which guesses at travelers’ true origins and destinations given the data on the legs of their flights; the algorithm reducing passenger flights into passenger journeys might have reduced by too much.

The Ipsos/Gallup data are consistent with my own experience. The 5,048-adult panel were asked how many round trips they took in 2017, and answered as follows, which matches what I know of my friends’ and family’s flying behaviors:

Screen Shot 2020-03-21 at 12.24.39 PM

How does 590 million compare with data from the Federal Aviation Administration (FAA)? The [“Annexes to the 2020 Inventory of US Greenhouse Gas Emissions and Sinks”](https://www.epa.gov/sites/production/files/2020-02/documents/us-ghg-inventory-2020-annexes.pdf) [1], in Table A-125, reports that the FAA reported 6.3-6.4 billion nautical miles of domestic commercial aviation in 2018, presumably because they have access to all flight data in the US[2]; this number is consistent with US Energy Information Administration (EIA) data on jet fuel consumption[3]. This implies 7.3 billion domestic miles (1 nautical mile = 1.15 miles). If we take the Southwest 10-K data on average “passenger haul” (which I am interpreting as passenger journey), average seats on a flight and load factor (fraction of those seats that are filled), the FAA number suggests 7.3 billion miles / 980 miles per journey * 150.9 seats on a flight * 0.835 load factor = 940 million domestic passenger journeys. That is not so far off from 850 million passenger journeys extrapolated from the Ipsos/Gallup survey, and again is the same order of magnitude as the IATA’s 590 million. The 940 million might also be inflated because it includes cargo and chartered flights.

Given the consistent order of magnitude from the IATA, airline 10-K, Ipsos/Gallup and FAA numbers, which themselves seem consistent with personal experience and EIA numbers, it would seem that the number of domestic one-way passenger journeys taken by Americans is in the high hundreds of millions, or that the number of domestic round trips is in the mid-hundreds of millions. Although I won’t spend the time to do such a check for other large domestic markets or international flights, the fact the IATA 590 million number is the right order of magnitude (potentially a slight underestimate) makes me trust the IATA 4 billion number more, although I recognize that I have not evaluated the data quality on the remaining 3.4 billion passenger journeys, which could be quite different. For me, this large number underscores the urgency of reducing flying now.

Outside the US, China and perhaps India are likely increasing their flying volume substantially: [“China’s domestic market added the most passenger journeys in 2018… China[‘s domestic market] comes second [to the US’s], with 515 million, followed by India some distance back, at 116 million.”](https://www.iata.org/contentassets/c81222d96c9a4e0bb4ff6ced0126f0bb/iata-annual-review-2019.pdf) “Flying/traveling less” campaigns should start and accelerate there as well.

Footnotes:

  1. [Annex 3.3 of the Greenhouse Gas Inventory](https://www.epa.gov/sites/production/files/2020-02/documents/us-ghg-inventory-2020-annexes.pdf) also includes jet fuel burn and carbon dioxide (CO2) emissions, as well as international aviation and military aviation. This data may be of interest late in tracking greenhouse gas emissions.
  2. I infer that FAA has access to all flight data in the US based on their ability to use such data to infer jet fuel burn and CO2 emissions from such data: [“Commercial aircraft jet fuel burn and carbon dioxide (CO2) emissions estimates were developed by the U.S. Federal Aviation Administration (FAA) using radar-informed data from the FAA Enhanced Traffic Management System (ETMS) for 2000 through 2018 as modeled with the Aviation Environmental Design Tool (AEDT). This bottom-up approach is built from modeling dynamic aircraft performance for each flight occurring within an individual calendar year. The analysis incorporates data on the aircraft type, date, flight identifier, departure time, arrival time, departure airport, arrival airport, ground delay at each airport, and real-world flight trajectories.”](https://www.epa.gov/sites/production/files/2020-02/documents/us-ghg-inventory-2020-annexes.pdf)
  3. The FAA data informing this estimate also seem consistent with EIA and military data on jet fuel burned. Table A-96 of [“Annexes to the 2020 Inventory of US Greenhouse Gas Emissions and Sinks”](https://www.epa.gov/sites/production/files/2020-02/documents/us-ghg-inventory-2020-annexes.pdf) reports that the FAA reported 13,650 million gallons of jet fuel were consumed by commercial aircraft domestically in 2018. The EIA reported 17,674 million gallons of jet fuel consumed domestically in the same year, with the remaining ~4,000 million gallons accounted for by military and other aviation jet fuel burn.