I read the WSJ Middle Seat Column article on buffer times (Why a Six-Hour Flight Now Takes Seven)

Scott McCartney did some analysis on how block time (or scheduled flight duration) has changed since 1996.

He noted that in a sample size of 50 flights he looked up, block time has gone up, indicating airlines are adding more buffer time into flights.

This creates different customer expectations and allows a better on-time performance.

Since I'm a stickler for comprehensive analysis, I decided to run the analysis on ALL flights for March 1996 and compare it to March 2009 (I wanted to compare actuals also, so I used 2009 instead of 2010 scheduled). Yes -- you can get all of this data publicly. You don't need to hold onto old timetables as Scott suggests. That way I can run the analysis on all flights simultaneously rather than looking up a sample size.

So what are the results?

(note: Scope is March 1996 compared to March 2009. Like flights are the same carrier, origin, destination, and hour of departure. Only major airlines that operated in both months are included: Delta, Northwest, US Airways, American, Southwest, Alaska, United, and Continental)

1) For all like-flights, average scheduled time has gone up by a whopping 8 minutes on average. Actual flight time (gate to gate) also went up, by an average of 6 minutes. That increase in 6 minutes is split mostly between air time and taxi-out time (with taxi-in time going up on average only by 1 minute).

2) By airline, Delta increased their schedule times the most (up by 13 minutes on average!), Northwest and American the lowest (5 minutes increase on average). But Continental and United experienced the biggest increase in actual time -- 9 and 8 minutes respectively.  For Continental, that was mostly taxi-out time (Congrats Newark!), for United it was air time.

3) By Departure airport, JFK had the highest increase in scheduled times -- an average increase of 27 minutes across all airlines. This is the phenomenon Scott was referring to -- airlines buffering with more time to account for long airport delays. However, the actual increase in time for flights from JFK was only 10 minutes, and 2/3rds of that was the result of air time, NOT taxi out time. The next highest increase at major airports came from LGA and IAD at 16 and 14 minutes respectively. Lots of smaller airports also had increases in block time -- most notably the Hawaiian airports. Anyone have any ideas why? Remember this would be only Hawaii - Mainland flights operated by the major carriers.

4) By Arrival Airport: EWR takes the cake for major airports (15 minutes average increase in time), followed by LGA, and ATL at 14 and 13 minutes respectively.

5) For Flight: The winner is Continental's Chicago - Newark flights leaving at 5 and 6pm, both getting an additional 45 minutes of time. It's notable that that actual increase in time was only 19 and 22 minutes for those two flights! So Continental is certainly over-compensating.  Those time increases are mostly from taxi out time (due to ground delay programs at Newark, no doubt). AA's JFK-LAX flights also got a bunch more time  -- near 40 minutes despite actual flight time increases of only 13 and 15 minutes. (Note: Scott refers to Delta's JFK-LAX flight as adding an hour. This is true, but the departure time also changed. In 1996 it was a 9:40pm flight with a 6 hour duration, and in 2010 it is a 9pm flight with a 7 hour duration. In my 2009 analysis, it was an 8:30pm flight with a 6 hour 33 min duration. Certainly some of that time is related to the micro-managing of the departure slot given how finicky JFK is in the evenings, although I don't disagree with Scott's assertion that part of it is also related to adding more buffer time).

What about flights that were actually much more late -- did airlines add more time?

The flights whose flight time grew the most from 1996 to 2009 were Continental's Newark-Boston  and Cleveland-Boston (between major airports with big enough sample size). These flights took on average for the month 69 and 54 more minutes in 2009 than 1996. But Continental only added 2 minutes and subtracted 1 minute to those two flights! Continental appears to have missed the ball on adding buffer time to those particular flights.

Overall -- 87% of flights had time added to their scheduled between 1996 and 2009, while only 80% experienced longer actual elapsed times. Meanwhile, 10% had time subtracted from their schedules, but 16% of flights were faster in actuality. So airlines were certainly over-compensating in 2009.

Motives? Like Scott says, they are many fold: Better operations overall, better on-time performance, better ability to plan.

It's a game airlines play to balance their operational needs and customer service. Sometimes they win, sometimes they lose. But predictability of delays is the biggest lever to help them play this game. Over time, we hope to use FlightCaster data to help with these kinds of decisions as we gather more data and analyze it in different ways.

Evan