New Study Finds Blocking Middle Seats Reduces Risk Of Virus Transmission Nearly 50%

A new medical research preprint is generating buzz claiming that airlines blocking middle seats on planes would cut coronavirus infection risk in half.

Let’s state the obvious, that it runs head on into field research in the real world. There’s only one flight where there are clearly suspected cases of virus spread since the start of the pandemic, Vietnam Airlines London Heathrow – Hanoi on February 1. Flights have continued every day since then, the virus is studied in greater depth than virtually anything in human history.

The paper, though, claims virus spread:

  • 1 in 4,300 without blocked middle seats
  • 1 in 7,700 with blocked middles

They also use a rather high 1% infection fatality rate to suggest passengers currently face a 1 in 500,000 chance of dying from COVID they pick up on a flight.

Over 66 million people have flown in the United States since March 1, the paper would attribute over 13,000 cases of COVID-19 to flights and we haven’t identified a single one of them. The study assumes,

  • Everyone wears masks, cutting down transmission of the virus, though they don’t factor differences in mask quality (whether it’s cloth or medical grade, or whether it’s worn over the eyes).

  • Flights are full (the airlines wish that were true!) – they model 2019 load factors, but not the effect of airline capacity growing faster than demand in July and beyond

  • That all flights have equal risk regardless of flight duration

The study recognizes It could be overstating risk because people don’t talk as much on planes as in other settings, and certainly not now or with masks on.

Authors also just do a back of the envelope adjustment for people flying being healthier and wealthier, and as long as they’re not flying Allegiant, probably taking better precautions generally.

Your risk really comes down to likelihood of being right near someone else who has the virus – much like anywhere else you go except that in most indoor places there isn’t HEPA air filtration. Planes are probably safer, then, than bars, indoor restaurant dining, gyms, and open plan offices.

Moreover your risk really varies by air market. If you’re flying from a non-connecting hub in a hot spot city you’re much more likely to be sitting next to someone with the virus than if you’re departing – say – Hartford, Connecticut.

China, by the way, is limiting load factors on flights to Shanghai at 75%, which means a lot of empty middles. That’s not a blanket middle seat ban but it’s something, targeted market-specific way.

It’s confusing to look at planes, where social distancing isn’t mandated, and hear airlines say it “isn’t possible” while restaurants are being required to do it to their own financial detriment all over the country and to wonder what’s different? And besides HEPA air filtration, to realize the only real difference is that state regulation of airlines is pre-empted by the Airline Deregulation Act and airlines have better lobbyists.

Nonetheless, at the end of the day flying seems to be one of the less risky things we do. It’s not been identified as a vector of significant virus spread. But I will still choose to fly Delta, Southwest, Jetblue or Alaska because in blocking middles they offer far more value than United or American does.

About Gary Leff

Gary Leff is one of the foremost experts in the field of miles, points, and frequent business travel - a topic he has covered since 2002. Co-founder of frequent flyer community, emcee of the Freddie Awards, and named one of the "World's Top Travel Experts" by Conde' Nast Traveler (2010-Present) Gary has been a guest on most major news media, profiled in several top print publications, and published broadly on the topic of consumer loyalty. More About Gary »

More articles by Gary Leff »



  1. Well, duh. It’s called common sense. All of the skeptics will tell you how this is a fake, that it’s all a hoax, but you ain’t getting me on a plane without masks and a blocked middle seat.

  2. Clickbait title …50% reduction
    And then the actual numbers
    1/4300 vs 1/7700

    Yes all the media love to play around with relative risks to get numbers like these. Of course masks and proximity to an actual infected individual is the real issue.

  3. “However, data from late June 2020 imply that approximately 1 in 120 Americans have Covid-19 on a given day” . Since airline passengers may originate anywhere, “Q” should be 1/120 which represents the average probability that any single American at any time has an active case of COVID, as opposed to trying to justify that airline passengers on a specific route, and in general have a socioeconomic status, or any other predisposition that make them less risky than any other group. Keeping all other variables the same makes the risk on a full plane 1 in 1666.

  4. “Your risk really comes down to likelihood of being right near someone else who has the virus – much like anywhere else you go except that in most indoor places there isn’t HEPA air filtration. Planes are probably safer, then, than bars, indoor restaurant dining, gyms, and open plan offices.”

    Study you are quoting did not compare the risks of the infection in airplanes vs. bars/restaurants/gyms/offices, so I am not sure how you came up with this. I think we call this non sequitur.

  5. I’m not sold on HEPA filtration.

    The virus itself is smaller than what the HEPA can filter. I believe the HEPA filters down to 0.3um but the virus is 1/3 of that size (0.1um or if you prefer 100 nanometers). I did see where the virus usually travels on another particle that is larger and maybe then it could be filtered like that.

    Why don’t we do a test with people that believe it works by putting them in a room and in another room put in people who are sick and see if the air going between the 2 rooms causes anyone to get sick?

  6. @Naoyuki – it’s not a non-sequitur it’s important to compare relative risk! point is that middle seat blocking wouldn’t be at the very top of the list, based on this study, of covid interventions.

  7. What’s often not mentioned with seat blocking, maybe because it’s so obvious, is that this means seating caps on flights. The fact is, one’s risk of catching something is materially reduced if there are 60 fellow pax around you instead of 100. Fewer people, less chance of ending up with an infected person amongst you, etc.

  8. You point out in the article that they assume equal risk for all flight lengths. There’s something else important to mention…it assumes all passengers are at equal risk of serious complications. The study is useless unless they can determine that. As far as the ‘well it’s about spreading the virus’, aren’t we at the point where we can have different rules for different risk groups? For example, don’t fly if you’re over 75 (or in another high risk group). Oh wait, but that’s not fair to someone who is at a proven greater risk, so let’s just tell everyone that they can have no life so we can have equality in destroying everyone’s life. How about give us a study that says who can fly safely and let them.

  9. I have to take issue with your assertion Gary, that airplanes are no different than restaurants, etc. Far be it from me to side with airline executives, however, in a restaurant, precovid, there was effectively 2-3 feet separation already and many seats could be rearranged as necessary, allowing many, maybe even most, restaurants to maintain 6 feet between tables and still reach about 50% capacity, more in some places. And those that can’t make it with 50% or can’t get to 50% capacity are just not reopening right now.
    Airplanes, on the other hand, have fixed seating and, precovid, offered 0 separation. To truly gain 6 feet of separation, most aircraft would have to reduce capacity by 83 to 89%, which means that, on a practical basis, it really “isn’t possible”. Effectively, all airlines would have to stay closed.

  10. @Gary Leff: Only way to compare relative risks of different activities is to estimate the risks for different activities using the same statistical model, then compare the numbers. In this paper, there is no mentioning of bars, restaurants, gyms, etc. That was not done in the paper, so you can’t really use the number from the paper to say what you said. The only comparison that was made in the paper is flying with middle seats occupied vs not occupied.

  11. @farnorthtrader I think you are misreading me, but airplanes don’t really have fixed seating, rows of seats can be swapped in and out pretty easily (subject to faa rules)

  12. @FFFlier1976 – no, I think the study has several flaws but it does not “assumes all passengers are at equal risk of serious complications” they don’t get into this. it’s a study about infection, not about the consequences of that infection (other than to postulate a 1% IFR which is high).

  13. But Gary, at this point the important issue should be the result of those infections. There seems to be agreement that the actual infection rate is 10x diagnosed cases. We’re all still too preoccupied with infection rates. If instead of locking everyone down we got all the young healthy people on planes (and cruise ships) coughing on each other so they could get infected quickly we would be much better off now. All these travel restrictions based strictly on the fear of infections spreading when it’s entirely possible that you can limit the consequences of those infection by focusing the restrictions. Let people use their common sense. If you’re 80 and have emphysema, don’t get on a plane. If you’re an obese diabetic smoker, don’t travel. If you’re a 20 year old college student, knock yourself out, you have nothing to worry about.

  14. As a risk manager you should be ashamed of giving this study any time whatsoever. Such a shame as you normally reduce things down into understandable terms quite well. This is nothing other than an exercise in numbers, percentages, etc. and not an actual study of transmission on aircraft. Sad for you.

    No company should hire you for their services if this is something you would present for consideration to make policy.

  15. @rich, the size of the virus is irrelevant, as it needs a vector to move around in. Viruses do not move on their own. It’s all about the particulates the viruses are contained in, the spit flying from your mouth. These vary from 0.1 micron, which masks struggle to block but have relatively low viral loads (critically important to how easy it is to infect someone) to 100 microns or larger, which fall out of the air in seconds. The dangerous particles are the ones in the 5-20 micron range, which masks have varying levels of success blocking but all do a better than zero job blocking. These particles have significant viral loads and can stay in the air for a minimum of 30 minutes, with air currents and A/C extending that time indefinitely. Even cloth masks many particles of this size, so masks are helpful. The size of the virus is entirely irrelevant when it comes to mask efficacy, it only matters in how many viruses can fit into particulate vectors.

    This why debunked studies from dangerous sources like the AAPS and others need to be called out. They lump in all particulates over 5 microns as “of little consequence”, lumping them in with the larger, 100 micron ones that fall out of the air quickly. They then say masks are bad at blocking <5 micron particles. Then they conclude masks are ineffective. THIS IS WRONG! This is misleading, as 5-20 microns can float for a long time and are blocked by masks.

  16. No problem – I’m sure everyone will be happy paying 50% more for every ticket since all middle seats will be blocked. Easy math – if an airline got $100 a seat in revenue for the 3 coach seats ($300 total) and the middle seat was blocked they still would want $300 for those combination of 3 seats so the window and aisle are now $150 each (or typically much higher but used $100/seat to make it easier to understand).

    As for me I don’t care – I have never really been price driven and can easily afford whatever price is charged if I want to fly but I can already hear people whining about the cost while demanding seats be kept blocked – you can’t have everything you want people so grow up and live with it!!!

  17. The study mentioned that with blocked middle seats there are fewer infected people walking the aisles and using the toilets. Contrary to airline disinformation fewer vectors means less disease.
    American and United are talking out their rears.

  18. This “study” did not “find” a risk reduction. This is NOT an observational study. This is all hypothetical, and this is a hypothesis. I see no evidence that what this claims is true.

  19. I too disagree that airlines and restaurants can be easily compared. Restaurants only have to cap their capacity INSIDE the building. Many (in my area) are offering seating on sidewalks and in tents in their parking lots. It is impossible for airlines to offer seating outside the aircraft. Also, restaurants offer carryout and delivery service – generating revenue without utilizing their seating. Again, airlines cannot transport someone from point A to point B on a carryout basis. The underlying nature of the businesses are different… transport versus a food manufacturing facility.

Comments are closed.