There's something very strange about the lack of passengers on the hijacked planes. As this site puts it:
Flight 11 was a Boeing 757/200. This plane holds 239 passengers. There were 81 passengers and 11 crew. The 11 crew members included 2 pilots and 9 flight attendants.
Is it normal that this flight would not be at capacity based on other Flight 77s that leave from Boston to LA at the same time every day?...
Flight #175 had 54 passengers, the Pentagon plane, Flight #77, had 56 passengers, and the Pennsylvania flight, 38. All less than 25 percent of capacity. Why?
911 Research has another take on this (follow the link for the full analysis, don't just read this segment).
The number of people on the flights and the seating capacities of the jetliners has been a source of some confusion. Boeing's website gives the capacities of the 767-200 and 757-200 as 181 and 200. However the seat charts given by SeatGuru.com for the seating configurations by American Airlines and United Airlines 757-200s and 767-200s... show fewer seats. [158 to 188]
As a result they derive load factors of 51.3% (Flight 11), 31.3% (Flight 175), 29.3% (Flight 77") and 16.5% (Flight 93). Are these figures correct? Do they still represent below-average numbers? Staff Report 3 from the 9/11 Commission has more.
"The percentage of seats occupied on the aircraft -- also known as the "load factor" -- on September 11, 2001, was 51%, compared to an average load factor for Flight 11 of almost 39% on Tuesdays over the three months preceding 9/11. Thus, the load factor on this flight was slightly above the norm".
Staff Report 3 from the 9/11 Commission
And the other flights?
The same report gives us a 33% load factor for Flight 175 (average 49%). That's significantly lower, although apparently there were two other Tuesdays in the past three months where there were even less passengers. Although it's not as low as suggested at 9-11 Research, who for some reason list only 46 passengers.
Flight 77 also had a load factor of 33%, almost exactly on the regular average of 32.8%.
And Flight 93 had the lowest load factor of 20%, well below it's average of 52%, and the lowest in the preceding 3 month period.
Therefore we have one plane with more passengers then expected, one with an average figure, one below but within an expected range, and one exceptionally low (in Flight 93). Subtract the hijackers and the figures fall a little more, but it’s been suggested that traffic does fall post-Labor day, and there's no real pattern here. Without further evidence it's hard to see anything significant in the load factor data.