THE BOOK--Playing The Percentages In Baseball

Oh, I agree with that person who was disappointed with Nate’s blog at NY Times (with regards to the comments).  I just posted a comment, and it’s being moderated.

Sorry, but that is a horrible system, and guarantees I won’t be commenting there again.  JC had the same insane policy about moderation.  You either require registration or not, but you don’t moderate every comment.  NY Times even requires registration, so they’re position is even more absurd.

interesting that DCA reagan’s premium is nearly equal to dulles. reagan is so much more convenient unless you live in northern virginia near dulles. it’s worth so much more to be able to metro back and forth. and there are fewer runways and terminals. however, flight distances are limited at reagan, so that might affect things. and i think the airport is run by congress. but i still always find that the prices at reagan seem lower than they could be.

Nate’s data is interesting.  But the interpretation is very problematic.  He says “These two factors — distance to destination and size of market — are the ones I would propose can be fairly reflected in ticket prices.” Everything else that affects price is implicitly “unfair,” and Nate goes on to label airports as “overpriced” if fares exceed his model’s prediction.  But many other valid factors impact fares.  Tango’s example is a good one: why shouldn’t people pay more for the conveniences offered by Newark (or if you prefer, pay less if signing up for the hassles of going to LaGuardia)? 

More importantly, his model ignores a huge factor in pricing which is flexibility:  how far in advance was ticket purchased, and was it refundable or not?  There may be 15 or 20 different coach fares charged on a single flight, depending on when the ticket was purchased and its refundable status.  The “overpriced” airports are mainly places with a lot of business travel, while cheap airports are mainly leisure destinations (as he acknowledges).  This almost certainly means that the fares to the “overpriced” airports were on tickets bought with less average lead time, and more often were refundable.  And flexibility has value.  You can’t say those fares are too high, or fares to Orlando are a “bargain,” if you don’t take account of that flexibility factor.  I don’t think you can even come close to evaluating the premiums charged by these airports without accounting for those factors.

I’ve lived in the city and now north of the city, and I have always avoided Newark at all costs.  Unless I’m saving hundred of dollars, I’ll fly out of LGA or JFK.

I imagine that anyone who lives on Long Island would avoid Newark as well, so given where the population concentrations in the New York area are, I’m surprised (and quite pleased given my own personal preferences) that there’s a Newark premium.

I’m not buying the convenience factor you cite.

Newark is more convenient for people in North Jersey, but LaG and JFK are more convenient for people in Brooklyn, Queens and Long Island, which is more people.

Here’s an idea:

What if we assume people are behaving mostly rationally and shop around for tickets across multiple airports, which involves a simple click on most ticket search engines.
For example, I almost always use ORD but I do look at MDW in the event it’s about $80 cheaper, which approximately equals my additional cost in time/$.
When people buy tickets, they incorporate all of their preferences. Thus, what we have is every airport being precisely fairly priced. I trust that the market is efficient here (the man-hours put into those pricing algorithms is staggering) more than this from-the-hip attempt to quantify based on a tiny data set.

Good point about the pop of NJ v NYC.  (Not to mention that in south NJ, they probably go to Philly.)

Newark is my airport of choice for going to Manhattan.  Last time I had to go to the city, I stayed at a great hotel directly across from Penn Station, I took the train from Newark and never needed a car or a cab.  Plus, there’s a convenient nonstop from Seattle.

Just surfing the internet and came across this rebuttal to Nate’s post by an SME, somewhat ironically

http://boardingarea.com/blogs/thewanderingaramean/2011/04/how-the-ny-times-got-it-so-wrong-on-airline-pricing/

Also I agree with Guy above—it’s not too easy to pick holes in Nate’s analysis.

The moral of the story is to consult SMEs before publishing your research. We plead to economists with little baseball expertise to consult with this group before publishing ... Nate should have done the same with the right guys on air travel.