Monday, November 29, 2010

What The T.S.A. Isn't Telling Us - NYTimes.com

FiveThirtyEight - Nate Silver\'s Political Calculus

November 29, 2010, 8:08 pm — Updated: 8:08 pm -->

What The T.S.A. Hasn’t Told Us

Despite widespread anticipation of chaos at airports over the Thanksgiving holiday weekend — some of it prompted by concerns that passengers would “opt out” of new full-body scans in favor of more time-consuming pat-downs — most media accounts suggested that travel went rather smoothly instead, with few passengers declining such screenings, and short waits at most airport security stations.

Throughout the day on Wednesday — traditionally among the busiest travel days of the year — the Transportation Security Administration updated its blog with the happy statistics. “Minneapolis: Wait times are currently 5-10 mins. No incidents,” went a typical report. “Detroit: 25,000 passengers screened today, and 57 AIT opt-outs. All were screened and continued to their flights.”

I have no reason to doubt the two specific claims that the T.S.A. has made: first, that security lines at most airports were manageable (if not, I’m sure we would have seen plenty of evidence to the contrary, between tens of thousands of passengers with cellphone cameras), and second, that a relatively small number of passengers opted out of the new screening procedures.

Nevertheless, there are several things that the T.S.A. isn’t telling us — pieces of information that would seem to be critical to any comprehensive assessment of the efficacy of the new procedures.

A typical report on the T.S.A. blog read something like this:

Los Angeles: 113 AIT opt outs across LAX’s 8 terminals, which is less than 1 percent of the approximately 50,000 travelers screened at LAX today.

Here, we are told the number of opt-outs (113), as well a the overall number of passengers at Los Angeles International Airport (50,000). What we aren’t told, however, is how many of those 50,000 passengers were asked to pass through the full-body scanners — what the T.S.A. calls “advanced imaging technology” or AIT — in the first place.

At most airports that have the AIT systems in place, the news scanners are used at some terminals but not others, and are used on some passengers but not others (depending, for instance, on whether the crews on duty at the time are sufficiently trained in them). If, for instance, the scanners were only in use at one of LAX’s terminals, and then only for part of the day, perhaps only some small fraction of LAX outbound passengers passed through them, which means that the opt-out rate would have been considerably higher than what T.S.A. is implying.

This is particularly relevant given that there were anecdotal reports that the new scanners were not used over the holiday weekend at some checkpoints where they normally are (an allegation the T.S.A. has denied.) Now, these reports need to be interpreted cautiously, since some of them may have come from infrequent fliers who don’t have a good handle on whether or not the new machines are used under ordinary circumstances. Still, the T.S.A.’s data is not really worth very much without knowing how many passengers had the option of opting out — meaning, that they were asked to pass through full-body scanners than metal detectors — and how this compares to a normal day.

The other thing we don’t know is what passenger volume was like. Wait time at security check-in points is a function of essentially three things: how much capacity the T.S.A. has to handle passengers, how long it takes to screen the average passenger, and how many people are passing through the checkpoint to begin with.

Kudos to the T.S.A. if it anticipated the rush and had more staff on duty (why can’t it always be that way?). But it could also have also been that air travel volumes were lighter than anticipated — perhaps because passengers were perturbed by the new procedures and were traveling by other means (or staying at home.) It would be hard to regard the new procedures as a success if that were the case, particularly given that more people bypassing air travel for road travel means more fatalities on American highways.

We eventually will get some idea about this, since the Department of Transportation keeps relatively detailed statistics about passenger volumes. But it only comes after a lag of several months. In the meantime, we’ll have to keep in mind that if airports were less busy than expected, it may simply have meant that fewer people were flying.

I have e-mailed the T.S.A. and asked them what, if any, additional statistics they are willing to provide. While they undoubtedly did a good job in keeping traffic moving over the holiday weekend — and certainly did a good job in managing public relations over the new procedures — it would mean something much different if this was accomplished because new scanners were turned off, or if overall passenger volume was down from a typical Thanksgiving weekend. We simply don’t have any good way to assess this based on the numbers they have provided thus far.

DCSIMG

No comments:

Post a Comment