Hacker News Reader

Comment by endtwist | Hacker News Reader

While these numbers will never be perfect, per se, RITA[1] shows that in 2011, Southwest's loadfactor was 80.8% while in 2012, loadfactor was 80.4%. So, the flights remained around 80% full on average.

I took this into account in my calculation and assumed roughly 4.8 more seats were filled on average per flight, not a full 6.

[1] http://www.transtats.bts.gov/carriers.asp?pn=1

View on HN

I think the key question then is why it was that 4.8 more seats were filled on average per flight. If we had data the showed that the distribution of plane occupancy were the same year over year save for a few occasions where last year the plane had been at capacity but this year those extra 6 seats had been put to use then the conclusion could be valid. However with the data I've seen so far it seems spurious to conclude that the extra 4.8 filled seats were caused by the extra 6 physical seats.

Since not all seats are sold for an equal price, the extra seats may have indeed increased revenue potential by allowing for more "premium location" seats on the plane.

There are a lot of factors and it would be silly to think that the airline didn't take more than we can come up with into account in their decision.

Assume the pricing for seats is as follows:

$100 for first 50% of capacity

$120 for next 20%

$170 for next 20%

$300 for next 5%

$600 for final 5%

What is the revenue impact of expanding capacity? The solution is not obvious, but the author's approximation is more accuare than you suggest. You obviously can't solve this without a stochastic ticket demand model, but 4.8 is much closer to the right answer than 0.

Also if the airline wants to maintain an 80% capacity, having an extra x% of seats means they add to revenue if they can maintain that occupancy percentage. In this case we don't even need to factor in 100% occupancy, which if anything would help the numbers as I'm sure last seats on a fully booked flight aren't cheap.

They target around 100% full. That being said, having more seats allows them to adjust their pricing model so that they have more seats to attempt to maximize revenue.

Depending on the demand structure of routes, those 6 seats may have wildly different values and could potentially be worth more than the average ticket price, on average for the airline.

Since this is counter-intuitive, I'll explain why briefly:

Airlines attempt to price discriminate by selling low priced tickets to leisure travelers and high priced tickets to business customers.

The problem airlines face is that business travelers book at the last minute.

Since business travelers book at the last minute, they need to estimate the number of business travelers who will be flying.

If the average leisure passenger ticket sells for $100, and the average business ticket sells for $500, just having 1 extra seat can add 5 leisure tickets worth of revenue assuming they sell the same number of leisure tickets.

Realistically, their model will readjust the optimal pricing changes overtime to capture the most value. Because the additional seats give the model more flexibility, I would not be surprised if the additional seats added more than an average of 80% of 6 seats multiplied by the average ticket price.

The article states that reducing leg room from 32 to 31 inches allowed for one additional seat row. Which means that there were already no less than 32 rows on those planes. Six seats per row give 192 passengers. RITA numbers clearly show that the load factor have never exceeded 90% since 2010. That means that at least 10% of seats were empty, that is, 19 seats. There is no sense of adding another 6 seats just to have them empty. If there were a demand for six seats, the people demanding them would just have occupied those of 19 free ones. Adding another row only makes sense if the plane is full and there's additional demand for seats. We should have looked at the load factor distribution, not only at it's mean value. Most attentively we should have looked at the standard deviation of the load factor around it's mean, but RITA didn't provide such an information.

There are a number of factors in play. But without knowing how the priced seats, and the number of capacity flights, this could just as easily be normal growth as it could be the impact of extra seats.

Not that I'm saying you are wrong, just that it's not the only explanation of the known facts.

This is still a gross mis-representation.

The extra seats are only utilized when the plane is beyond the capacity of the old configuration. You'd need to know how often this event occurs in order to produce a meaningful number.

Until you can separate cause from effect, you have no way of knowing if the change in loading was due to reconfiguration of the seats, or other policy changes that occurred around the same time.

Reply on news.ycombinator.com