Wednesday, 28 October 2009

Poking economists can be risky

When I got annoyed about BERL's work on the "social costs" of alcohol and wasn't impressed with their response to critiques, things got interesting.

Now it seems that John Small is a motorcycle rider. And he's started digging into ACC's justification for massive increases in insurance levies on motorcycle registrations. And he doesn't like what he's finding. I hope for ACC's sake that they're able to quickly show him what they were up to, or that Small doesn't get a lot of consumption value out of annihilating shonky analyses that justify imposing costs on him. 'Cause as mildly irritating as I was for BERL, I'm sure Small would be far more irritating for ACC.

It makes sense that bikes would have to pay a bigger fixed cost than cars even if risks per mile driven were identical across the two vehicle classes: bikes use less petrol per kilometer, and ACC charges a petrol levy of 9.9 cents per litre to reflect the added risk per additional amount driven. They can't charge differential petrol taxes per vehicle, so it has to get loaded into the fixed charge. Given that, though, it would take remarkable differences in accident risk/damage by type of motorcycle to generate the fixed charge difference between small and large bikes that ACC proposes: $257 for a <125cc bike and $745 for >600cc. The larger bikes use more petrol per kilometer than the smaller bikes.

Small is definitely right that there's a bias against motorcyclists built into the system, but the optimal tariff structure given an untaxed segment is pretty difficult. First best, we eliminate the untaxed segment: we can and should put an ACC levy on cyclists. It would be fairly easy to require a rego for cyclists and to charge an appropriate ACC premium for them (risk rated for type of bicycle, whether there's a kid's seat, etc). Doing it for pedestrians would be pretty tough. The only way of doing it, best I can reckon, would be through a poll tax added onto your regular income tax, perhaps risk rated by number of pedestrian accidents in your neighbourhood. In that state of the world, premiums could reasonably reflect the two-sided nature of accident costs (a car-bicycle accident needs both the car and bicycle present).

Outside of that state of the world, whenever a bicyclist gets on the road, the ACC premium for any vehicle that can injure a cyclist, even if it's one of those horrible cyclists that plug up Riccarton Road going through Hagley Park despite the existence of perfectly good cycle paths going through the park paralleling the road (hates them I do), goes up. So the bias is really on any segment that is taxed compared to the untaxed segment. We'd then expect, though, that trucks and buses would have a heavier charge attached to them -- they do more damage to the untaxed segment in any encounter between the two. Small says trucks are getting a pretty sweet deal. Given that diesel trucks don't pay a petrol levy, he's almost certainly right; would need more details on accident rates to be sure though.
That's enough for now - I'm off for a ride.
If I were ACC, I'd be rather nervous about the "for now" part. Sounds like there's a fair bit he could take a sledgehammer to, were he so inclined.


  1. But surely the cost should lie with the cars. Cyclists (and pedestrians) are not very dangerous to other road users, or themselves.

    2 tonne plus pieces of steel travelling at high speeds are the main source of the hazard.

    I realise that pedestrians and cyclists can be just as 'at fault' as other vehicles. But it is the fact that the hazard exists that allows their human error to get them into trouble.

  2. @Matt. Remember your Coase. It takes two. Externalities are reciprocal. If the car weren't there, no accident; if the pedestrian weren't there, no accident. In a no-fault world, the pedestrian deciding to take a walk imposes costs on all other road users; it's just that his side isn't levied a charge.

    In a with-fault world, insurance doesn't pay out if you're grossly negligent and the side that's at fault pays out for both sides' damages, so a pedestrian doesn't impose any particular costs by deciding to take a walk and both sides have incentive to take due care. Of course, transactions costs may be somewhat higher in that world.