On pre-flight security and which bits matter

So much for passenger safety announcements. Now for searching passengers.

The overall summary of what follows is that we seem to do a good job of searching folks – certainly far better than the much- and seemingly-justifiably-maligned TSA in the US. (See also this, rather damning recent report). But there could be more information made available about its effectiveness, and there must be real questions about whether most of what goes on is really worthwhile at all in terms of reducing the risk of air crashes.

The law

First, I find, it is always good to start with the law in our remarkably law-abiding democracy.

By a notice issued under Section 77B(1) of the Civil Aviation Act 1990, the Aviation Security Service (part of the Civil Aviation Authority) is required to screen all passengers and undertake searches of passengers where reasonable, “if necessary”, for planes that carry more than 90 people. There are also requirements on airport and airline operators to maintain security procedures under Part 108 of the Civil Aviation Authority (CAA) rules.

This law is required of New Zealand because it a contracting state of the International Civil Aviation Organisation (ICAO – a part of the UN) and a signatory to Annex 17 of the Chicago Convention 1944, which requires contracting states to meet various security measures to protect the travelling public as well as flight crew and ground personnel.

The long list of no-nos

The list of things that the Aviation Security Service (AvSec) is required to search for under the screening notice (see pages 2-4) is long and divided into categories:

  • Firearms, guns and weapons (including catapults, flare guns, and toy guns of all types)
  • Pointed and edged weapons and sharp objects (including axes – don’t worry, there is already likely to be one on board for safety reasons if you need it – ice skates, pointy walking sticks, swords, and throwing stars)
  • Blunt instruments (baseball bats, but also fishing rods and skateboards)
  • Explosives and flammable substances (including explosives, grenades, mines, ammunition)
  • Chemical and toxic substances (including batteries, pepper spray, anything that could spontaneously combust, radioactive material, poisons, or fire-extinguishers – don’t worry about that last one either, the crew have one).

For international flights, passengers are also searched to ensure they are not carrying more than the prescribed amounts of any liquids, aerosols and gels, and are doing so in the prescribed way (i.e., in a transparent bag with nothing else in it).

This search is required by another notice from 22 March 2007. “Liquids, aerosols and gels” includes, of course, water, but also shaving foam, soup, mascara, toothpaste, and hair gel. Interestingly, the order does not contain the 100 ml limit, nor the rule about putting things in a see-through bag – although it does require the screening to be “reasonable”.

This list applies from 1 August 2011 replacing a list from March 2011 which in turn replaced one from August 2002. Unfortunately I can’t find on the CAA website the previous directives, so I do not know how they have changed over time, but it looks like they are altered relatively seldom. I am not sure either why the restrictions on liquids, aerosols and gels are not included in this one notice (since they were put in place before this notice was re-issued, so it would seem tidier to have just one screening notice). However.

Lots of screening going on …

The numbers involved in all this screening and searching are pretty substantial:

According to AvSec’s 2010 Annual Report (the one to June 2011 does not appear to be online yet), in the year to June 2010, it screened:

  • 4.5 million international passengers
  • 5.7 million domestic passengers (the rest of the 9.5 million total domestic passengers were passengers on planes with fewer than 90 seats)
  • 15 million pieces of hand-luggage, and 5.5 million pieces of hold luggage (meaning on average each screened passenger had 1.5 bits of hand-baggage and only 1/2 of a piece of checked luggage).

… and therefore a high cost

All of this work means substantial costs:

  • AvSec employs 800 people. (Interesting 60% of them are men, and two-thirds are aged between 40 and 60. Perhaps ex-police or military folks?)
  • It charges $10 per international passenger (down from $15 in 2010) and $4 per domestic passenger (down from $5), a charge that you pay as part of your ticket.*
  • The total cost of screening was $70 million in 2010 year (note that that also includes air-side security – i.e., checking that crew and groundstaff meet security requirements).

And it holds passengers up a bit: for just over a minute internationally, and just under a minute domestically, they say. So if AvSec screens 10 million people a year, that is 10 million minutes a year (or approximately 19 person years) spent queueing, which is quite a lot of time, to say nothing of the hassles of getting undressed and dressed again, and taking one’s computer out of one’s bag.

The restrictions on liquids seem a particular source of difficulty. In one month (May 2010) AvSec intercepted 10,000 passengers (about 3% of international travellers) who were not meeting these rules and therefore, presumably, had had to have some of their prized liquid possessions confiscated, thrown away or consumed before their time.

And does it work

The interesting question, of course, is whether all this is worthwhile. Does it work, i.e., does it find anything and, more complicatedly, is the reduction in risk generated by this search work really worth the costs of achieving it – both the $70m costs to passengers of running AvSec, and the hassles for passengers of taking off clothes, unpacking bags and waiting in lines.

Unfortunately there is little information in AvSec’s Annual Report that would let you figure this out. There is nothing on how many prohibited items are found, nor on how many passengers are found carrying these things.

There are reported 21 instances where a restricted item was found after the screening point, i.e., where AvSec failed to find something that it should have. Unfortunately it is not explained how these were found, nor how many items in total were found at the screening point, so it isn’t possible to figure out how effective AvSec is. We don’t know how the process works that found the 21 items (how regular it is, how reliable it is as a measure of the effectiveness of the screening), so it is not possible to draw any real conclusions. That said, 21 items seems like bugger all in the context of 10 million screened passengers.**

We do know that, of the 21 items, 19 were “unauthorised and prohibited” whereas 2 are described as “unauthorised dangerous” items (page 50), although it isn’t clear what that means. And the service also reports a 1% failure rate of agents under what sounds like standardised testing to see if they find stuff. The CAA conducted 11 compliance audits in the 2010 year, including at least one on passenger and baggage screening performance. It seems odd that the CAA would do these audits – AvSec is part of the CAA, so it looks like the agency is auditing itself. But perhaps we can be reassured to know that the American TSA also conducted a review and pronounced itself satisfied that AvSec met its screening requirements.

AvSec reports 1 formal complaint for every 423k passengers, or around 25 “justified complaints” against officers in the course of the year, which doesn’t seem like very much, suggesting they acquit themselves well in interactions with individual passengers. AvSec also reports a curious 1.25 complaints from airlines (perhaps the quarter of a complaint was a barely decipherable mumble).

In summary, there are few obvious reported problems with the service. It clearly achieves its major goals. As the Service reports on page 51, in 2010 there were “No in-flight security incidents”, and “No airside security incidents”. It also says, more heroically, that there were “No dangerous goods introduced into aircraft”. I say that this is more heroic only because I can not see how AvSec would know what got on to aircraft.

A better approach

So as far as we can tell from the limited reporting, we don’t seem to be screening too little, i.e., there don’t appear to be big risks to the flying public from the current level of screening, and AvSec seem to do a potentially unpleasant job with a reasonable degree of applomb. The trickier question is whether the approach to screening could be changed or the level of screening reduced, and some of these substantial resources saved without an increase in danger to the travelling public.

Because there is one thing that stands out obviously from the long list of prohibited things that AvSec looks for. Strangely, AvSec is not being asked to screen for people who are likely to want to hijack or blow-up airplanes. Aside from the items that are dangerous to airplanes in themselves (like items that might self-combust), all the things that AvSec looks for require another ingredient to be dangerous, i.e., they require someone who wants to cause harm. And there is nothing in the screening approach that is intended to find these people. It is as is if the police were to hunt for burglars by searching everyone’s house for stolen items, or stopping every car to look through the boot, rather than focusing on more directly on finding those who steal.

AvSec seems to understand the importance of searching for potentially bad people. It talks about in its Annual Report (page 47) the importance of “reliable intelligence gathering and dissemination” to counter terrorist threats to aircraft, while also recognising that “the current threat to New Zealand aviation is relatively low”.

The consequences of our approach to airport security were obvious in 2007 with New Zealand’s second recorded hi-jacking, an incident where a woman endeavoured to hijack a flight from Blenheim to Christchurch with a knife without great success (the plane landed safely at its destination), but with injuries to the two pilots and a passenger who intervened. The hijacker was subsequently sentenced to nine years in prison  for her crimes. Because she was flying on such a small plane, neither she nor her bags would have been screened prior to boarding.

The hijacker is a great demonstration of my point because she was on bail for threatening to kill and possession of a weapon when she got on the plane. It turned out at the trial that she had a very long and tragic criminal history, with 27 previous convictions. We didn’t search everyone getting on board the flight from Blenheim to Christchurch, but could we not have at least searched her, given the history?

After a May 2009 review, the Minister of Transport (correctly in my view) decided not to change anything about domestic security screening following the incident, but instead to look at ways to isolate the cabin from the passengers in smaller planes. The costs (at around $160m) were said to be prohibitive to screen all passengers.

A better approach, I think, would be to alter the screening approach. Instead of stopping everyone and screening everyone, we should search only crazies and dangerous people. I suspect the reason that we do screen everyone for dangerous items is because it is easier than finding dangerous people, and it makes it look like we are doing something, even if it is unpleasant for the vast bulk of people who are not of any interest from an aviation security point of view. The “screen everyone” approach is what emerges as the definition of security when the problem is given to a bureaucracy that doesn’t want to be blamed if something goes wrong.***

Conclusions and stable doors

I have read with interest the security blogs of Bruce Schneir in the course of looking at this issue. He is a very firm critic of the American TSA and its approach to security, describing most of what they do as “security theatre” designed to make travellers feel more comfortable but not actually making them safer.

He is also very critical of the escalation of requirements over time, pointing out that changes in the security screening are mostly responses to particular terrorist plots. Liquids, aerosols and gels are restricted thanks to these guys. And, at least in America, thanks to this cat, you are required to take off your shoes (which thankfully isn’t required in New Zealand).

Mr Schneir says it well:

A short history of airport security: We screen for guns and bombs, so the terrorists use box cutters. We confiscate box cutters and corkscrews, so they put explosives in their sneakers. We screen footwear, so they try to use liquids. We confiscate liquids, so they put PETN bombs in their underwear. We roll out full-body scanners, even though they wouldn’t have caught the Underwear Bomber, so they put a bomb in a printer cartridge. We ban printer cartridges over 16 ounces — the level of magical thinking here is amazing — and they’re going to do something else.

His point is that we are shutting the stable door after the horse has bolted. In my view, New Zealand does not present sufficient risk to have a stable in the first place.

We also used to screen every bag for biosecurity risk items (mostly fruit) on passenger arrival. It is interesting to see, therefore, MAF recently moving towards the approach that Customs takes to looking for contraband, i.e., it will try to stop only passengers of interest and stop holding the rest of us up.

Perhaps this is a consequence of falling budgets. But at least it means I can hope that the CAA will learn from its fellow agencies, and either get rid of bag screening entirely for domestic flights, or move to a more sensible system based on individual customer risk rather than searching every bag.


* To me there are real questions about why passengers are paying for security requirements decided on by the CAA. If the government wants to scan passengers, it should pay for it from taxes and prioritise that spending against all its other priorities. Otherwise the temptation is to over-screen because the passengers who are paying for it are not in a position to disagree. But that is a post for another day.

** I note that the TSO says that “Since January 2010, advanced imaging technology [their x-ray scanners] has detected more than 300 dangerous or illegal items on passengers in U.S. airports nationwide.” In that time the TSA has scanned hundreds of million of travellers passengers. Which does rather suggest there isn’t much to find – calling into question the whole need for this security in the first place.

*** Interestingly, some also say that the “screen everyone” philosophy is actually harmful rather than wasteful – because x-ray scanning can cause cancer. Which is why some types of scanning have now been banned by the Europeans.

PS My favourite question about aviation security:

Q. Can I take my wedding cake on my flight?

A. You can take a wedding cake as long as it is solid, and does not have a liquid, cream or jam filling.


2 thoughts on “On pre-flight security and which bits matter

  1. Pingback: Free to roam | whereishayden

  2. Pingback: Cut and run | whereishayden

Comments are closed.