International route networks are currently very limited, often only flights that can be supported by cargo with passengers a plus are operating. That’s both because of limited travel demand, with business travel non-existent and a raging virus keeping people at home, and because of travel restrictions and local restrictions too (there have to be activities at a destination happening to make the travel worthwhile).

But which routes are going to come back, and which ones are gone for good? British Airways announced they’re dropping 17 destinations, many likely for good. Two U.S. destinations, where they share revenue with American Airlines, are going away: Charleston and Pittsburgh.

There are some principles for what flights to expect to come back, and which ones are likely gone for the foreseeable future, as the Airlines Confidential podcast noted this week, though I’d add to their list.

• Last in, first out The most recent cities added are likely to be ones that don’t return, because in general they’re probably the most marginal. The reason in many cases they’re the most recently added is because an airline was already operating the routes that are most robust.

  Sometimes there are surprise successes (British Airways only started Austin – London Heathrow because of the Boeing 787, but alternated operating the route in pre-Covid times with larger Boeing 777 and 747 aircraft even after Norwegian introduced Austin – London Gatwick service). And sometimes there are changing circumstances, or new corporate deals, that make a flight possible. But in most cases the recent additions are the ones that were on the bubble to begin with, and on the bubble pre-Covid isn’t likely to make sense again for some time.

• But it might come back if it’s a shorter international flight capable of being operated by a smaller plane, for instance if it’s within the range of a Boeing 737 MAX or Airbus A321XLR.

  Just as the Boeing 787 made some smaller and mid-sized city pairs possible, because of lower costs and the need for fewer passengers to earn a profit, so does the new Airbus A321XLR. We had mostly been thinking about these planes as adding routes that didn’t previously exist (or replacing Boeing 757s on short transatlantics). Now these planes might be used on routes that used to exist with larger aircraft (a Boeing 767 or 787).

• Ability of countries to re-open quickly Countries that didn’t see much Covid-19 spread likely have populations that are most vulnerable to the virus. Where there’s overlap with limited hospital capacity, and where vaccines are likely to come last (the mRNA vaccines require cold storage that many countries in the world cannot support, for instance), we may see hesitancy in opening up – even for arriving passengers with proof of vaccination, because we do not yet know the extent to which each vaccine will limit a person’s ability to spread the virus (even if it protects them from symptomatic Covid-19).

  There will be more data soon, but uncertainty prevails here and could influence the timing and determination of whether certain flights return.

There are some flights in the schedule now that may not operate. For instance it’s not clear when Australia will re-open to foreigners, or allow their own citizens to travel, or even lift the cap on passengers they’ll accept in a day (due to quarantine measures). There are likely Sydney flights that will be pulled off the schedule if restrictions aren’t lifted.

And we should have pretty good advance notice of whether or not a route will be added back into the schedule, because leisure travelers book international trips with longer lead time than domestic trips. For instance it’s not uncommon for people booking travel to take part in European river cruises to book a year out (as soon as they’re able to lock in airfare after the cruise is booked). Of course until cruises come back in earnest, likely not until there’s widespread vaccine dissemination, flights bringing passengers to cruise markets aren’t likely to return either.