Monday, October 18, 2010

Counterintuitive findings?

Pool fences
Could Queensland’s new tougher pool fence laws offer an opportunity to study the Peltzman Effect? Will we now feel that pools are no longer a safety hazard for toddlers and drop our supervisory guard? One man, who refuses to comply with the laws, has argued this exact point and is strongly supported in his views (if you can trust the newspaper comments).

In one case, a pool owner living on a canal has had to fence their pool, yet is not obliged to fence off access to the canal.  One does wonder about how far governments can go to protect us from our own behaviour.

Pool fences are only there to protect kids from parents who don't. There are no fences around all the lakes in Brisbane, Southbank's lagoons are not fenced, the Brisbane River is not fenced. Why? Because we are responsible enough to ensure our children don't get into danger in these areas.

What further astounds me is that lack of evidence in the pool fence debate. In one of the more interesting studies I could find, 52% of pools where toddler drowning events had occurred in Western Australia where compliant with the pool fence legislation (compared to 40% for randomly selected pools).  There was no further discussion of this key point – that statistically it appears more likely to drown in a fenced pool that an unfenced one (I would be very interested if anyone can find a more thorough study of the effectiveness of pool fence laws).

While this is just a small sample from one State, and I would question whether general conclusions can be drawn, some more rigorous examination of the effectiveness of pool fence laws is seems appropriate before toughening the laws.  Is the government really going to do the same thing and expect different results?

Cycling by the road rules
The Council is inviting CityCycle subscribers to undertake a Cycling Confidence Course to improve their bicycle skills and brush up on their knowledge of road rules.

Maybe that's a bad idea. Recent research suggests that people obeying road rules are more likely to be killed by trucks than those who disobey the rules by, for example, running red lights. 

Women may be overrepresented in [collisions with goods vehicles] because they are less likely than men to disobey red lights.

By jumping red lights, men are less likely to be caught in a lorry driver’s blind spot. Cyclists may wait at the lights just in front of a lorry, not realising that they are difficult to see.

In more than half the fatal crashes, the lorry was turning left. Cyclists may be deceived by a lorry swinging out to the right to give itself room to make a left turn.

I can’t agree more with these findings.  Every day I see cyclists waiting in the blindspot of a car or truck at traffic lights, and occasionally see a cyclist sneak up the left side of a bus while it is turning left.  I hope Brisbane City Council’s cycling confidence course acknowledges that sometimes it is safer to break the rules.

Congestion (queuing) as an efficient allocation mechanism
I have raised the idea in the past that road congestion is in fact an efficient allocation mechanism provided that there is prior knowledge of expected travel times.  Now, from The Australian we have this:

Sure, if we invested enough in roads, all cars could travel at the speed limit. But the costs of thus expanding road capacity would greatly outweigh the value motorists place on the savings in time and discomfort.

Exactly the same applies to road charging. With charges set sufficiently high, remaining drivers could go at speeds rivalling the Melbourne grand prix. But even Mrs Moneybags, rocketing in her Ferrari, would not value the benefits enough to offset the welfare loss to the peons forced by the high charges to walk to work. Add to their loss the costs of implementing the road charging scheme and the efficiency loss is all the greater.

1 comment:

  1. it is interesting that economists could possibly disagree on this issue. The existence of congestion is an example of inefficiency. Queuing of vehicles is inefficient, and results in considerable large economic costs. The key problem that supports my contention is that roads are inefficiently priced as they are indirectly priced. Without direct pricing that reflects the cost of the infrastructure, resources will attract demand that exceeds supply, resulting in queuing, which is an inefficient outcome. To be efficient, roads would need to be priced according to time of day (this would mostly apply to urban centres, main arterials leading to these centres), road type, location with some cost associated with each vehicle's environmental impact (ie a rating for emissions).
    For a direct charging scheme to work, drivers would need access to real time information re road conditions, associated costs (that would exist during peak/congested periods). That way, when a driver decided to drive their vehicle, they would know the cost that would be associated with that decision. As such, it would be likely that drivers would change their driving habits, such as combining trips, travel more at off-peak periods (where possible), etc.
    It is true the cost of implementing a charging system would be expensive, but that does not suggest we shouldn't consider it.