An aisle passenger and a window passenger look at each other on board a Lufthansa narrowbody, where the middle seat has been blocked as part of the airline's Eurobusiness service

Delving into the economics of middle-seat-free flying


Airlines restarting operations to and from those countries and regions where the COVID-19 pandemic is coming under some semblance of control are in a quandary. On the one hand, they want to get flying again. On the other, they need to do so safely. And then there’s the invisible hand of the market, guided in some cases by an iron fist of regulation. That is, clearly, at least two too many hands involved.

Some airlines have already fallen into the quagmire of being seen to profit off a pandemic. This will likely have a chilling effect on others looking to provide spacing options, with a large amount of inconsistency between carriers. Airline trade association IATA, surprising few, said in early May that it “does not recommend restricting the use of the ‘middle seat’ to create social distancing while onboard aircraft.”

Said the organization, “Calls for social distancing measures on aircraft would fundamentally shift the by slashing the maximum load factor to 62%. That is well below the average industry breakeven load factor of 77%. With fewer seats to sell, unit costs would rise sharply. Compared to 2019, air fares would need to go up dramatically—between 43% and 54% depending on the region—just to break even.”

But is this true? Airline fares are notoriously opaque — another case of the airline industry crying wolf while also itself being a wolf — but it is possible to do some rudimentary calculations.

Today, airlines by and large use one of two models: either selling a full second ticket, or by treating a second seat as a kind of extra-legroom ancillary. The latter — used by airlines including Peach, which calls it the Space Seat Option and charges the equivalent of roughly US$20 for a domestic flight and $30 for an international flight — is perhaps the easiest from an airline economics point of view.

It is, however, less easy from most other points of view, as Frontier Airlines found out when it tried to introduce a “More Room” no-neighbor seating product “starting at US$39” on May 4, quickly walking it back.

So what are the economics — and accounting calculations — for middle-seat-free flying? Let’s look at some back-of-the-envelope figures.

At a full ticket cost replacement model, the fare increase would be 50%

Let’s say that an airline flies a 180-seat jet in a 3-3 configuration, and that the normal fare would be $100 for each passenger. (This is something of a simplification, with no ancillaries or other charges, but let’s run with it.) It therefore takes in $18,000 for that flight. If a third of the seats are not filled, it in theory needs to find the $6,000 from the remaining 120 passengers, so the ticket price rises to $150: a 50% jump, well within the IATA band.

Raising a $100 fare to $150 — essentially, the equivalent of adding a bag and a reserved seats on some airlines — feels is unlikely to provide a massive disincentive to travel, even more so if we’re talking about raising a $30 fare to $45, but that question has to remain outside the scope of this article.

That all roughly matches IATA’s numbers. Yet that 50% number is likely too high given the ways that modern airline fares are constructed. Government and airport fees, charges and taxes are often twice or more the base fare even in normal times, but particularly now that demand remains relatively low.

RGN selected cheapest economy flights on popular routes for five over-weekend days in mid-July, including London-New York and London-Amsterdam, using the Google Flights metasearch, clicking through to the cheapest option for each.

Longhaul, legacy carrier: closer to 33%?

The cheapest airline was charging GBP369.18 for a return LHR-JFK, of which £101 was the fare and the remaining £268.18 came under its “taxes, fees and carrier charges” section. That amount was split between:

  • UK air passenger duty and passenger service charge, UK: £95.18
  • Five US user/service fees and the passenger facility charge: £22
  • Airline “carrier imposed charges”: £151

Given that carrier imposed charges are, in essence, a dynamic pricing mechanism, it seems reasonable to include these and the base fare as what should be charged for an extra seat, making a total of £252.

Split between two passengers, an extra seat should be an additional charge of £126, or about 33% extra on top of each fare.

Shorthaul, LCC: an even lower percentage?

The cheapest return from London (in this case “London” Southend) to Amsterdam (Schiphol) priced out at £42.98, and although this airline doesn’t break out the full list of fees and charges, it does note that APD is GBP13, which makes up more than 50% of the UK departure leg.

There will also notionally be fees to pay to the airport, which a PDF from 2018 suggests should be some £23.20. Southend Airport head of communications Georgina Pavelin declined to answer any questions from RGN, including about how much would be charged per passenger.

APD and the airport fees would, doing the maths, total more than the actual fare for that leg, highlighting that for an LCC in particular the lack of any ancillary product — speedy boarding, seating, bags, meals, add-on car hire or hotels, and so on — means this cannot be a precise analysis.

A UK government spokesperson speaking to RGN on background confirmed that Air Passenger Duty is calculated on a per-passenger basis, and if a passenger books more than one seat the airline is only liable to pay the government the APD for the number of passengers carried — not the number of seats sold. Very few, if any, airlines make this clear, make this calculation automatically, or provide a process for passengers to this incorrectly paid tax.

More widely, for both kinds of airline, fewer passengers and fewer bags means less fuel, fewer flight attendants, fewer meals, and so on. The way this scales is also likely to change depending on a number of factors: route length, cargo demand, et cetera, which should reduce the amount passengers are charged. So could proposals to reduce the rated maximum takeoff weight of aircraft to fit them into lower categories for charges such as landing fees.

But it seems clear that the marginal charge doesn’t look much like the IATA 50% figure here either. The industry may well end up rueing the day it decided on opaque fares as a general policy.

A key problem for airlines is the cornerstone theory of economics that people make rational decisions, with “rational” having a specific meaning in the economic context, but also — in certain countries’ response to sensible public health measures — a more general one here. In reality, passengers’ relationships with airlines and ticketing is more complex than being purely rational, in both the economic and the wider contexts.

That’s true even when there is not a global pandemic of an invisible, poorly-understood disease, the reaction to which has been heavily politicized into a kind of us-versus-them matter of tribal belief rather than being based on any sort of scientific reality.

None of this particularly helps airlines, and the result is the scattergun approach where airline X is selling X percent of seats, airline Y is blocking all middles, airline Z is blocking all neighbor seats, and airline A is doing none of the above. It feels very much like it is time for regulators to step in and bring clarity and consistency to what is currently a morass of ad hoc, unclear inconsistencies.

As this article went to print, Europe’s EASA and ECDC had just published a new COVID-19 Aviation Health Safety Protocol that looked set to do just that.

Related Articles:


  1. Thomas D. Pellegrin

    Hi John, I followed your analysis up to the longhaul / legacy carrier, but see a disconnect after that. It seems to me that you have kept the same underlying assumption of 33% of the seats being neutralized throughout. IATA’s analysis actually uses a global weighted average of 38% of seat removals in its analysis, based on current fleet composition. That figure breaks down into 33% seat removal for narrow-body (e.g. A320, B737) due to the 3-3 abreast configuration, 40% for for wide-body (e.g. A350, B777) due to the 3-4-3 abreast configuration, and 50% for regional jets (e.g. E-190) and turboprops (e.g. Dash-8) due to the 2-2 abreast configuration. I believe that by lowering the load factor cap accordingly, the rest of your analysis will reconcile with IATA’s.

    • John Walton

      Hi Thomas, and thanks for your comment! Since this is very much, as I said a couple of times in the piece, back-of-envelope, pretty rudimentary calculation stuff, I figured that since the 33% works for the 737, 757, some 777s and almost every 787, together with every A330 and most A350s, it’s reasonable to use. I’m very cognisant that distancing has a different effect on, say, your average A330 than your average 787, simply for the 2-4-2 vs 3-3-3 layouts, not to mention for regional jets.

      The weighting you mention is itself questionable, of course. Are we weighting for number of aircraft? Number of flights? Passengers? Flight-kms? Passenger-flight-kms? IATA’s data didn’t seem to say, and I’m sure there are reasons for whichever they chose, but without knowing one can’t really start thinking about what they were, or consider whether they tilt the balance in one direction or another.

  2. It might be a flawed approach to use the ‘cheapest’ price when carrying out an assessment.

    This is because the ‘cheapest’ price usually represents a situation which cannot be adopted by all businesses in a competitive marketplace.

    Many times the ‘cheapest price’ represents a ‘loss leader’ in pricing.

    Using the industry’s average price would yield a much better assessment I’d suggest.

    Just one person’s opinion 😎

    • John Walton

      Hi Dave — you’re right, it might be! But it’s the only data we have at the moment, so I made sure to choose historically popular routes, where there’s current competition, and I did a bit of a sniff test for “are these prices crazy”. Sure, they’re low right now, and perhaps we’ll have another crack at them if we have any kind of price stabilisation sometime, and of course I’ve provided the numbers so you can have a go at them with your own preferred price if you like 🙂

  3. RaflW

    I think market segmentation plays a huge role here. What does Delta (or United, BA, etc) need to do to woo back business travelers, v. what Frontier, Ryanair or Norwegian needs to do for a more leisure (but certainly some company) travel. Price sensitivity for the latter is higher.

    It seems that an airline like Delta — I admit to being both DL Platinum medallion, and a very small shareholder — is aiming to reassure the base of their business who spend more and come closer to *having* to travel. They have said pretty firmly that they are blocking all middle seats thru June. July is rumored.

    I can imagine an outfit like that deciding to limit capacity in all or most categories above Main cabin for longer than that. Int’l J with the suites probably doesn’t need limiting. (If they relaunch 757 transcons with paired lie-flats? I’d say max 75% capacity, with pairs only sold on a single rez). I would pay extra, or fly more willingly, if as a Plat I had a C+ seat with the middle empty. Or PE at reduced density via seatblocking. Perhaps let Main be more full if loads start to rise.

  4. I saw AirNZ mentioning adjusting their booking software to optionally eliminate distancing within a single booking (‘bubble’ in NZ COVID terminology). This enables more than 67% occupancy in their 3-3 A320s if you have bookings of 3-6 while having empty middles otherwise.

  5. Sergio

    It would be interesting to include to the calculation the savings made for having a lighter aircraft due to a reduction of passengers and see how much could that compensate the increase of prices of the tickets.

Leave a Reply

Your email address will not be published. Required fields are marked *