Last week Canberra was alive with security and Parliament House was surrounded by guards with assault-rifles. Anyone who has gone into Parliament in past years knows it already has extremely tight security, but now we are ready for an invasion. True, there was the attack in Ottawa. so we need to assess if a lone gunman could have walked into our Parliament as easily as the one who walked into Canada's. If we were already sufficiently secure, the right answer was 'do nothing'.

Since 9/11, anyone involved with security seems to be ready to over-react and has an unlimited budget to do so. No one seems willing to say that the world is full of a wide variety of risks, and we can't hope to reduce the possibility of something going wrong to zero. With each type of risk, we have to weigh the probable damage against the cost (and feasibility) of reducing the risk.

We do that, implicitly, in many other instances. Traffic engineers know (and the rest of us intuitively understand) that we could reduce the number of road deaths and injuries if we spent more on straightening out the dangerous curves (Google 'Pacific Highway' and you'll have an example). Experts are reluctant to put out precise cost/benefit calculations on these trade-offs because that would mean putting an explicit value on human life, which they don't want to do. But the point remains: we decide to tolerate some deaths and injuries because we don't want to spend more.

Sometimes the risk-reduction takes the form of constraints on action, rather than expenditure. But the same principle applies. We don't reduce the vehicular speed limit to 15 kph, even though there would be fewer road deaths if this were to be done. How much inconvenience will we put up with at airport security queues before we say we'll take a chance on the hijackers (provided the pilots latch the cockpit door)?

If we were interested in getting the best return on our public safety dollars, the money might be better spent on reducing the risk to the public of a king-hit from a drunken lout in Kings Cross on a Saturday night, rather than on another layer of protection against a terrorist attack.

Stirring up panic is always an easy story for journalists and shock-jocks. As well, behavioral economics tells us that people give too much weight to low-probability risks. Thus it's not surprising that we spend too much on some risks and not enough on others. In addition, perhaps security (and intelligence) matters avoid proper accountability because they can claim some need for secrecy.

In an earlier era, the stoic response was 'Stay calm and carry on'. It might still be the way to go.

 Photo courtesy of Flickr user Rusty Stewart.