Monday, June 23, 2008

Fuel taxes and subsidies

Whilst those in New Zealand complain about fuel tax you can be grateful of three things:

1. Petrol tax in New Zealand is not high compared to Western European standards (UK fuel tax on both petrol and diesel is just over 50p, that is NZ$1.30 a litre before VAT). European politicians are bigger pillagers of motorists than NZ ones.

2. New Zealand has no specific tax on diesel besides a derisory 0.33c/l local authority diesel tax, almost every country in the world has diesel tax, and the Greens actively support introducing one.

3. All fuel tax money is hypothecated into the National Land Transport Fund. Whilst around 15% goes on public transport, walking/cycling projects and the like, the rest is on road related activities. However don't forget Don Brash was the one that pushed for this at the last election. By contrast NONE of the UK fuel tax is dedicated to transport at all, and if the fuel tax is matched against spending on roads, only 9% would be required. UK motorists have every right to protest, New Zealand motorists have got it comparatively easy.

Now having said that, cutting or abolishing GST would make a positive difference. It would also be positive to charge motorists for using roads directly instead of through a tax on an input to road use. However I've written much on that before, having a fuel tax is so tempting for politicians. Treasuries like it because it costs little to collect and the elasticity of demand for fuel is low, so it is a cheap and effective way to pillage the population. On top of that environmentalists typically support taxes on fuel because "fossil fuel use bad".

Fuel taxes are an appalling way to pay for road use, they bear next to no relationship to the costs of maintaining or building roads, the costs imposed by road use on other road users, and it is only getting worse. Moreover because they are easy for politicians to manipulate they should be avoided like a plague. As a transition they should only be used as a source of dedicated funds for road maintenance, but this should be phased out over time.

Meanwhile, another simple reason why demand for oil is high are the large number of countries that subsidise the price for consumers. That, thankfully is eroding away, with China, Malaysia, Zambia, India and Indonesia all cutting fuel subsidies significantly. Of course these subsidies mean demand is inflated, according to the Christian Science Monitor the IMF reports half the world's population live in countries with subsidies, but the top 20% of income earners in those countries receive around 42% of the value of the subsidies.

Sadly the leftwing government in Chile is increasing fuel subsidies, as is South Korea, and Fiji is looking into it.

So when you take into account the overdemand created by subsidies (yes offset by under demand created by excessive taxes elsewhere), the underproduction created by prohibitions/taxes on exploration, what should be the price of petrol? The truth is that no politician or bureaucrat will ever know, but we can know the distortions that interfere with that price. That should be where people focus attention.


Eric Crampton said...

Fuel taxes might not be a perfect way to pay for roads, but it's tough to think of a better one. Fuel usage ought to be proportionate to the wear and tear a vehicle imposes on roads: heavier vehicles use more fuel than do others, per kilometer, and pay more tax. I rather suspect that alternative mechanisms would be at least as manipulable as fuel taxes and would be more intrusive on privacy.

Libertyscott said...

Crampton, it simply is not true.

It is hardly tough to think of a better one, New Zealand has had one for thirty years, it's called road user charges. It was based on the mountain of evidence that fuel consumption does not rise in proportion to the wear and tear imposed by heavier vehicles on the road surface. Take this quote from an Oregon study:

"Road damage increases much more than proportionately with vehicle weight on a road surface. For example, if Vehicle A weighs 48,000 pounds and Vehicle B weighs 4,000 pounds, the damage done to the road by Vehicle A is far more than 12 times the damage done by Vehicle B. Relative damage is frequently discussed in terms of the Equivalent Standard Axle Loads (ESALs) created by the passage of a vehicle. A standard axle load is defined as 18,000 pounds. Hence, the number of ESALs assigned to a vehicle is a measure of the approximate number of standard axle loads that would have to pass over the road to do as much damage as that vehicle would do on one pass. While ESALs go up much more than proportionately with vehicle weight, fuel consumption does not vary much with weight. Hence, a heavier truck would be expected to pay slightly higher fuel taxes than a lighter truck but would do much more road damage. Thus, a strong reliance on fuel taxes for apportioning the costs of road maintenance and construction leads to the result that heavier trucks do not pay for their cost responsibility."

The relationship is commonly acknowledged to be a power of four. For every additional tonne on an axle, you multiply its impact by a power of four. Diesel engines do not increase consumption by that magnitude.

Road User Charges have long worked well as a way of recovering heavy vehicle costs in New Zealand, and hardly intrude on privacy. In Switzerland a more modern equivalent has the user take out a smartcard that records the total kms and weight and inserts it into a machine which calculates the charge and bills the user.

The privacy concerns are easily fixed, sadly they have been hyped up by media that thinks GPS is some sort of spy satellite system, when it is nothing of the sort.

No road charge system need know the name and exact road you used, it simply needs to know you used a type of road with a type of vehicle at a particular time, and calculate the charge.

Eric Crampton said...

I don't particularly disagree with anything in your prior piece on the advantages of road charging over petrol taxes, but I just have a hard time seeing the gains as being all that large. The biggest variance in vehicle weight is with diesels, and the road user charging there already takes care of it. Motorcyclists get a raw deal on petrol taxes, but I'm not sure that there's otherwise enough variance in weight for petrol vehicles to make it worthwhile.

Libertyscott said...

Crampton. Yes on weights below 3.5 tonnes you're right. Although the maintenance impact on unsealed vs sealed roads is quite significant for cars, that is it.

There are two reasons to justify moving from petrol tax from an economic point of view, for petrol cars. First is congestion charging. By far the biggest car externality is congestion, which once you strip out economically efficient congestion (which can be defined), should be charged for in cities. That can incentivise mode shift, time shift and trip consolidation. Singapore is the most sophisticated example working using prepaid anonymous cards with on board units passing under gantries. Prices vary by location and time (and adjusted every 6 months). This could be implemented selectively in Auckland and Wellington.

The second reason is diminishing returns. Fuel efficient cars generate less revenue while fixed roading costs continue to rise, as does the cost of road improvements. Raising fuel tax to make up that penalises poorer fuel consumption (which outside cities doesn't generate a negative externality).

I don't think motorcyclists get a raw deal at all. Half of all road maintenance costs are fixed, motorcyclists demand higher standards than cars, benefit from route efficiency improvement projects (but not capacity) and consumer much less fuel than cars. Their ACC levies are insufficient to recoup their share of costs as well.

Looking beyond charging, by no means can it be said that government allocates funding for capital investment on roads as efficiently as it could. Ideally electronic charging could provide detailed information on revenues per route, identifying whether users are prepared to pay for route improvements. Only with some form of universal distance or bulk purchased charging that can vary by time and location can road use be managed as effectively as possible. On board units can calculate charges and spit them out to be managed by account. Competing companies can sell road services on the same corridor. There is enormous potential, and the more you move to road pricing the less scope there is for politicians to fleece it.

Eric Crampton said...

Congestion charging takes a bit more complicated a mechanism than the current road user charging. They're a very good idea in Auckland and Wellington (and perhaps to a limited extent Christchurch). But they don't necessarily have to be a part of an overall move to road user charging. Of course, the marginal cost of adding full road user charging would be trivial given the infrastructure for congestion charging.

On the diminishing returns argument, isn't it generally the case that more fuel efficient cars are lighter and consequently generally impose less maintenance burden on roads? Again, it's not as good as directly weighing vehicles and charging accordingly, but I just don't know that it's really that far off: wouldn't the gains would be second order? Of course, if we move to a situation where the petrol taxes collected are less than the overall maintenance costs, we could simply increase the petrol tax by a couple of cents to get things back in line.

I didn't know that ACC levies on motorcyclists didn't cover costs. Another reason to privatize ACC: actuarily-fair rates.

I think your last argument is the most convincing. For me, the incremental benefit of road user charging over petrol taxes (for petrol, not diesel) are just so small for road maintenance that they're not worth bothering about. But if we're talking about moving to a system that looks an awful lot more like private roads, that's a different beast entirely - one that very plausibly would generate first order gains.

Anonymous said...

The first issue I can see with congestion charging is that Auckland doesn't have an adequate public transport system - that was the main reason why people did not support the proposed congestion charging of Auckland back in 2006. In Wellington, such a system could be implemented as they have an adequate public transport system; however, I don't think you would have the traffic to justify such a move.

While Road User Charges are a good idea, in comes the other issue - it encourages fuel inefficiency. The status quo encourages people to get more fuel efficient vehicles to reduce their fuel cost. Of course, you know the externalities of burning more fuel. The presence of Road User Charges is one of the reasons why I suspect that SUVs have become very popular in recent years; most of them are diesel powered and they pay the same amount of tax as a person with a diesel car.

Libertyscott said...

Crampton - Yes congestion charging would be more sophisticated than RUC, but the issue is how you do it. With the exception of Wellington, any "point" type charging scheme in NZ would distort costs because it would charge a high cost for a certain point in the network, rather than take into account the congestion costs imposed during the total trip. This is why distance/time/location based charging makes sense for Auckland, and could be an evolution from RUC for cars (e.g. it could be an option for motorists to not pay fuel tax, but pay by time, distance and location). Yes there are details I can answer, but distance avoids the "boundary" effect of London/Stockholm type schemes.

On the fuel efficient cars being lighter, frankly at that point the ESAL measure (impact on the pavement) below 3.5 tonnes is arguing about next to nothing. The petrol tax issue is also alternative fuels and equity. Hybrid cars pay half the petrol tax, but don't halve road costs. How do you address that? Fuel cell cars pay no petrol tax, neither do electric cars(both would have to pay RUC now anyway).

Agree on privatising ACC, this should be a risk assessment like the insurance sector, varying by vehicle owner and location.

Yes, road pricing is unfortunately seen as the end solution, when the bigger structural problems with road management, governance and financing need addressing. No country has done this. The Nats were going to address it by putting roads into SOEs and requiring them to run at a profit, borrowing for capital and charging road users (having the right to sign contracts with road users to opt out of fuel tax). It isn't perfect, but is in the right direction.

John-ston: Public transport is not a necessary pre-requisite for congestion charging - as congestion charging enhances the competitiveness of public transport, and as long as market entry is relatively easy, let the market provide. People can change time of day as well.

RUC doesn't encourage fuel inefficiency, truck operators are cogniscent of fuel costs, but they make a fair tradeoff - fuel vs time vs RUC. RUC charges for road use, fuel is an input as are tyres, oil, brake pads, labour.

Yes there are environmental externalities related to fuel, but consumption really only relates to one - CO2. Clean vs dirty engines are not always low vs high fuel consumption, catalytic converters increase fuel consumption for example. If you want to deal with environmental matters then fuel tax is a very poor way of doing so, because it charges the same whether you drive in the middle of the desert road or pass a playground - both have vastly different impacts.

It's a cheap blunt tool that used to be the best one available, but internationally (US, UK, Australia) it is widely recognised as having had its day.

Anonymous said...

"Public transport is not a necessary pre-requisite for congestion charging - as congestion charging enhances the competitiveness of public transport, and as long as market entry is relatively easy, let the market provide. People can change time of day as well."

It is a necessary pre-requisite to make it palatable to the population. When London introduced the congestion charge, most people already used the Tube, the buses and the trains, and so it didn't effect much of the population, and that made it acceptable - same with Stockholm and Singapore.

If you don't have the public transport infrastructure in place, then people will merely view it as another tax. Plus, leaving it to the market alone tends to provide an inferior product; buses tend to be viewed in a lower light than trains for instance. The market would provide buses, but probably not trains due to the prohibitive costs of setting up such an operation, and so you don't really have an alternative option that is politically palatable.

Further to that, you also have a time lag. Taking Auckland for instance, it would take between six months and a year to get extra buses, and would take three to four years to get new rail multiple units. You would need to gear your congestion charge around that as well; I seem to recall that London got some buses prior to the introduction of their charge.

"RUC doesn't encourage fuel inefficiency, truck operators are cogniscent of fuel costs, but they make a fair tradeoff - fuel vs time vs RUC. RUC charges for road use, fuel is an input as are tyres, oil, brake pads, labour."

Um, you forget about the other people that pay RUCs. In my household, we pay RUCs and it isn't because we own a truck. SUVs that use diesel have to pay RUCs and so to do car owners that use diesel. Unlike trucking firms, households are less likely to think of fuel costs when purchasing and operating their car.