[sustran] Re: distance based fuel tax technology test

[Paul Barter]

Dear Lloyd

Thanks for taking time to read and comment.

> I wonder, though, if ultimately the vehicle quota system is 
> dropped and a distance-based scheme is the sole mechanism, 
> then there will be significant purchases of second vehicles.  
> Families could get around a 25,000 km quota by simply 
> purchasing many vehicles.  

Actually, they can't get around it this way. The proposal is to allow
Singapore's vehicle quota to be retained (but to allow the auction price
- known as the COE price - to be variabilised). If demand for vehicles
increases, then the price under the vehicle-quota auction would still
rise just like now. The Singapore authorities would retain the ability
to control vehicle numbers just like now. 

Also please note that the 25,000 km is not a usage quota. Instead,
25,000 km is one suggestion for a usage limit for the vehicle taxes
(including the quota auction price). In other words, it is a tax-paid
allowance. A usage limit is not a quota, it is just a neat way to
variabilise lump-sum vehicle taxes. (I am proposing usage limits to
replace the current 10-year limit on the auction price)

You can keep using your vehicle after the usage limit is used up. You
just need to pay another round of vehicle taxes. The proposal is for
such payments to become routine, whenever you run out of your tax-paid
allowance. The incentive to drive less is in trying to delay your next
round of vehicle taxes. It truly is usage-based charging, albeit
integrated with the vehicle quota scheme. 

> Further, if a person has a 25,000 km quota, then there may be 
> a psychological push to fully use up your credits, even when 
> there may not be an actual demand for the distance.  Or 
> persons may end up leasing out the vehicles which have spare 
> kilometres.  Whenever there are credit systems based on a 
> total amount, then the system works to ensure persons will 
> fully use the available credits. 

I think you may be assuming that there is a time limit on the 25,000 kms
(as with other similar proposals in the literature). Under this proposal
there is no such time limit. If you decide to use your vehicle
sparingly, then you could make your vehicle taxes last for many years.
So there is no incentive to use up any quota.

> By contrast, marginal charges, such as the ERP, are effective 
> in directly charging for each km at the point of use, and 
> thus directly provide an incentive to curb usage.

Agreed. And this proposal also charges for every km by running down your
tax-paid usage allowance and making your next payment loom up faster. 

> The paper begins with the point that the current combination 
> of vehicle quotas and road pricing has "major drawbacks".  
> But it is not quite clear what those major drawbacks are.  
> The only point made in this regard is that the large upfront 
> vehicle costs dwarf the marginal ERP charges, and thus 
> psychologically create a situation where usage seems 
> inexpensive.  If this is the only problem, it seems like you 
> could just change the balance (e.g., raise the ERP charges).  
> This would seem far simpler and more effective than 
> introducing another new scheme altogether.

The main problem is indeed mainly the sunk costs psychology. This HAS
prompted the Singapore authorities to begin to gradually reduce fixed
vehicle costs, as you suggest. So they do indeed plan to rely more on
usage charges like ERP. I certainly agree with you that Singapore could
reduce fixed costs dramatically and just let the ERP mechanism do its
work. 

However, they are unable to reduce the fixed taxes very quickly (for
various good reasons explained in the paper). They also seem unwilling
to comtemplate going very far in this direction. An important reason is
fear about the possibility of sky high ERP prices, with uncertain
political cost. Under the existing mechanisms, reducing fixed costs and
increasing usage costs necessarily involves a large increase in the
vehicle fleet (unless demand is dampened by economic crisis or such
like). There is an understandable reluctance to let go of control over
vehicle numbers. 

(Speculating a bit, but I imagine that nightmare scenarios haunt the
authorities here as they contemplate the day fixed vehicle taxes are
suddenly dropped to zero... Would thousands of people rush out to buy
new vehicles for the first time, not realising that ERP charges are
about to skyrocket. These people then feel extremely annoyed a few
months later when they find they can't afford to actually drive their
new vehicles.)

The main point of the paper is therefore that it allows the authorities
to retain this policy lever, the vehicle quota, while eliminating its
main downside. It would allow them to quickly remove the sunk costs
problem yet to be gradual about loosening their control of vehicle
numbers. 

Have I managed to clarify?

Paul