Monday 28 April 2014

Finland's path to road pricing raises some big questions

I wrote on 27 March that the a Finnish Government Working Group had recommended that the country move away from fixed taxation of motor vehicles towards a distance based approach.  I've now had a chance to read the full report (in English) of the Working Group and to digest its analysis and approach, and it demonstrates the one rule of thumb I've often seen in road pricing studies across the world - every country has considerably different contexts, but many common issues.

To read the report yourself, it is available here as a pdf.   

For those who don't want to dive into the detail, I'll summarise my key thoughts and views first, and then put down some more detailed reflections afterwards.  It's particularly pertinent given the upcoming ITS Europe Congress in June, being held in Helsinki.

Finland, most of the population is along the southern coast
On 3 February 2012 a working group was established to explore how Finland could move towards "fairer and smarter transport systems and study long term strategies to introduce road pricing systems, with a specific mandate to assess the feasibility of GPS road pricing".   The focus was on looking at global experience, what objectives road pricing could serve, what technical solutions could be viable, what impacts would arise and how and over what timescale it could be introduced.  The working group was also to look at privacy and whether any other services could be provided using the GPS platform.

This follows previous investigations into congestion pricing for Helsinki alone, which came to the conclusion that the best approach would be some form of distance based charging that varies by time and place.


The key points from the study are as follows:

- Existing fixed taxes for cars should be replaced with a distance based tax.  Those taxes include car tax (paid on first registration of vehicles in Finland) and annual vehicle tax and vehicle motive force tax.

- The tax considered for the purposes of the study would be €0.033/km (US$0.074/mile) on all cars with another €0.02/km (US$0.045/mile) for non-petrol cars to offset the lower diesel tax.   Some variants on this were tested based on a regional differentiation.

- The distance tax would not replace fuel tax.  Finland's fuel tax rates are €0.6729/l (US$3.52/gallon) for petrol and €0.4966/l (US$2.60/gallon) for diesel, compared to the EU legal minimum rates of €0.359/l for petrol (about US$1.89/gallon) and €0.33 (US$1.72/gallon) for diesel.  (note these are US Gallons for the sake of comparison, not imperial).

- Motoring taxes in Finland already recover more than five times the state spending on road maintenance, with none of the money hypothecated for roads, so the argument for distance tax was not based on revenue generation or protection,  but rather dynamic improvements in economic and environmental outcomes.  By law, such a charge can currently only be a tax in Finland.

- Shifting from fixed taxes to fuel taxes was considered, but ruled out because it would not offer the potential to target congestion and environmental impacts by location and time of day.

-  The distance tax would not apply to buses and coaches because no fixed taxes apply to them.

-  The distance tax would also not apply to HGVs because the fixed taxes that apply to them could not be reduced sufficiently to make the tax work, as the rates are not far above EU minimum rates.

- The net impact of a shift to distance tax by 2025 is estimated as being a 30 million reduction in annual car trips, with a 4% drop in CO2 emissions from cars and around the same proportionate drop in serious car accidents.

Change in passenger km estimated in 2025 from introducing distance charging in Finland
- Most car users would pay less, with commuters in cities and high usage rural users paying more.  The net effect is to reduce barriers to car ownership, but encourage car usage when and where alternatives are not available.

- The estimated capital costs for such a system, for 3.5 million cars, were €89m-133m (US$123m-US$183m) (which I believe to be far too low), but operating costs were estimated at between €116m-133m (US$161m-US$183m) per annum.  The report indicated more scrutiny is needed over these costs.

- However, these higher costs compared to the current tax system would be recovered by the reduced costs to the economy of fewer accidents and emissions.  The savings from reduced congestion were not calculated, but given evidence from previous studies ought to deliver substantially higher economic gains, even if charges did not vary by time of day or location.

- It was considered that the primary benefit from the change would be to allow for charging by specific road and time of day, so congestion charging could be introduced equitably, targeting congestion and locations where there are reasonable public transport alternatives.

- It was strongly recommended that any distance tax have strong privacy protection, so that all data on specific trips be kept with the on board unit and in the possession of the vehicle owner, with only charging data released.  Other data could only be accessed if the owner wished to query the tax calculation or if there is suspicion of systematic fraud.

- There could be industry development potential in allowing such a system, as it would encourage local business to develop complementary systems and potential applications to use alongside the distance tax.

So, Finland is considering a car only based distance tax to replace taxes on owning cars.  That's interesting and revolutionary, but I also think it is missing some key points, which I come to below.  These are:

- Fuel tax ought to be included down to EU minimum rates;
- By including fuel tax, all other vehicles should be taxed by distance and mass;
- The rates of distance tax should reflect, at a minimum, infrastructure costs;
- By shifting to distance tax Finland cannot evade a strong case for partial hypothecation of revenues, in which case it may be wise to consider part of the charge not being a tax, legally speaking;
- The capital and operating costs need far closer scrutiny;
-  The benefits of shifting to distance based charging may be overstated around accidents, due to technology around automation, but understated by excluding the deadweight costs of existing taxes.

More detail

The proposal recommended for further work is to replace existing fixed taxes for cars (not heavy vehicles) with a distance based tax.  These fixed taxes include Car tax (a tax paid on first registration based on price and CO2 emissions) and annual vehicle tax and vehicle motive force tax (the latter for non-petrol vehicles).

Finland's fuel tax rates are €0.6729/l (US$3.52/gallon) for petrol and €0.4966/l (US$2.60/gallon) for diesel. Diesel tax is lower to offset perceived higher road freight costs due to the country's size and isolation from the rest of Europe.  The report specifically said that fuel tax couldn't be considered because the rates are not far above EU minimum levels, but they are actually around €0.32/l higher than the minimum for petrol and €0.16/l higher than the minimum for diesel.  So both could be cut by say €0.16/l if the differential is to be maintained, or cut to the minimum levels altogether, so that there could be a charge based on vehicle characteristics and road usage, rather than fuel.

This would suggest that there is scope to replace part of these taxes with a distance based charge, but that is not explored further.  The focus is on replacing fixed taxes only.

This seems unusual, given the obvious trend of more fuel efficient vehicles eroding revenues from fuel taxes. Notwithstanding that it would appear that Finland can raise fuel taxes far more easily, in political terms, than many other countries, this doesn't remove the sustainability problem of the fuel tax.  Finland obviously cannot lower fuel tax below the rate specified in EU Directive 2003/96/EC, but it can take its taxes to that level, which would create a range of benefits in themselves.  By lowering an input into transport costs, but offsetting it with a charge on road usage, it would more fairly distribute the incidence of that tax based on actual road usage, not a proxy for it.  Bear in mind that the minimal fuel tax rates are not low and can be argued as more than recovering imputed social costs of CO2 emissions.

Of course by including fuel tax there would be a need to charge all vehicles, which raises the next issue - what is the purpose of charging for road use at all?

The kilometre tax proposal is not driven by a need to protect revenue for roads, as existing revenues from all motoring taxes (excl. the general VAT) are over five times greater than what is spent on road maintenance. Including VAT, revenues from motoring taxes comprise 15% of all government tax revenue, so road transport more generally funds other expenditure.

In Finnish law, all such taxes are treated as general revenue. There is little interest in changing this, although it is admitted it becomes more difficult if a distance based tax varies by location and time of day, this was not explored in depth.

So a proposal based on charging cars is not about setting charges based on recovering infrastructure costs, because if it was it would mean charging based on road space and mass.  In short, the base charge for vehicles should be higher for heavier vehicles, and should be lower on major highways and higher on more lightly built roads, to reflect relative infrastructure costs. Beyond that, there may be variations based on environmental factors and congestion, but the report touches on none of this.

This is a major gap, yet the strongest case for pricing is to have efficiently allocation of resources around transport.  The starting point should be transparent charging of such costs.  If there is a political imperative to charge more than that, beyond the minimum fuel tax levels, then that should be considered on top of that.  Whether

It is important to note the proposal is to apply to cars only.  Trucks are taxed differently, with the "HGV motive force taxes" and diesel tax said to be close to minimum EU levels by law.  So they wouldn't be charged by distance, neither would buses and coaches which face no taxes other than fuel tax.   Given that there remains scope to reduce fuel tax to near the EU minimum levels it seems like this deserves a rethink.

There are 3.5 million cars in Finland.

The impact assessment was based on replacing taxes with like for like charges, with some scope to vary charges by region.  The overall effect was to make it cheaper to own a car, and for most car users to pay less, but for a minority to pay more.  The proposed rate was €0.033/km (US$0.074/mile) on all cars with another €0.02/km (US$0.045/mile) for non-petrol cars to offset the lower diesel tax.

The net impact was that 30 millions fewer car trips would result, with commensurate reductions in kilometres travelled by car (with all other modes seeing an increase in travel).

CO2 emissions would drop by 4% more than retaining existing taxes and accidents by the same proportion.

Under kilometre tax the more a car is driven the more it pays relative to fixed taxes

The effect is seen clearly in this figure from the report, showing that beyond a threshold of around 17000km per annum, a motorist would pay more with kilometre tax, but below that less.   It's hardly surprising, as fixed taxes effectively mean everyone is charged an average, penalising those using the roads the least.

The capital costs claimed for the system are, in my view, heroically low, at between €89m-133m (US$123m-US$183m) which appears to imply that on board units will somehow be supplied commercially under the operating costs, which are themselves more realistic, at between €116m-133m (US$161m-US$183m) per annum.    However, in both cases I think more scrutiny is needed over costs.

The big news is that whilst a distance tax would cost more to administer and collect than ownership taxes, the net economic impact would be to almost fully offset that through the cost savings in lower accidents (presumably due to less traffic, so less exposure to risk) and lower emissions.   However, I would question the accident point, given technology in the next two decades may drastically cut that risk anyway with greater automation, at least to allow for incident evasion.   Yet, nothing is said of the deadweight economic cost of fixed taxes, which was highlighted in Australia, by a study undertaken for the New South Wales Treasury, indicating that the entire operating cost of shifting to distance based charging is offset by corrections to distortions in the economy due to fixed taxes on vehicle ownership.

The table from the report below indicates the Finnish working group's view on the impacts of a kilometre charge overall.   This, of course, does not consider how charging by time of day or location could cut congestion costs on top of that (and concurrently emissions).

The operating costs of a distance tax in Finland would be largely offset by reduced externalities.

In conclusion, the report indicates that further work be done on designing a tax system based on a kilometre based charge for cars, and that this should focus on ensuring that the technology to achieve this can be introduced, that the costs estimated are accurate and that trials can be developed to allow for an incremental introduction of the technology.  It recommended expressly that all data collected on vehicle trips be kept in the on board unit, and be transmitted only to calculate taxes, and that the data be owned by the vehicle owner - and only be available if the tax is challenged or if fraud is suspected.  This is considered to be a way of protecting privacy.  Below are some other interesting graphs from the report.

Distribution of impacts by geography and intensity of car use.

My view

I strongly supportive Finland's interest in moving away from fixed taxes towards usage based taxes.  Although there is an economic case for having some charge to reflect the fixed costs of highway network maintenance (which are not usage based), the externalities generated by usage are such that shifting to usage based taxes is likely to create more efficient outcomes.  

However, there is one big gap in the analysis.  The purpose of this reform would be to better link taxation to transport policy objectives, yet the most direct way to do that would be for motoring taxation to reflect costs rather than being political determined.

That means undertaking analysis as to the highway infrastructure costs in Finland, and allocating costs among vehicle classes so that they are charged by distance across all vehicles.  Beyond that, factors for pollution and regional and time of day factors for congestion, can be added.  However, to get more efficient allocation of resources across society, and also to better reflect efficient modal choices, road users should pay for the roads.   I understand Finland's legal framework meaning motoring charges are taxes, but should that get in the way of charging for usage based on costs?

After all, if the intention of a tax is not to change behaviour, then it would automatically rule out charging by location and time of day.  Yet, if a tax is to be levied efficiently, should it not be done to avoid distorting economic signals - and the most obvious one would be to charge according to objectively defined costs that are being recovered.

What that means is that trucks need to be included, and buses, and to do that means having the new charge also reflect a reduction in fuel taxes to just above the EU minimum levels.  There are likely to be dynamic benefits in reducing fuel taxes and recovering the difference from distance and weight based charging for heavier vehicles, but the main effect will be to reduce costs for smaller trucks, increase them for larger trucks, but also incentivise (if done correctly) truck configurations that reduce wear and tear on the roads.  It can also increase the relative competitiveness of rail and sea based transport (although for only a limited range of goods).  It will seem rather unfair for cars, but not trucks to face paying for road use, especially when it comes to charging for pollution and congestion.   That is a public relations issue when it comes to telling all car users that they need to pay by distance, but trucks don't, particularly when it comes to congestion charging when large trucks use scarce road space on congested roads in city streets, exposing residents to pollution - and only pay fuel tax while they drive.

So Finland should continue to investigate distance tax for cars, but must also include other vehicles and a means of refunding part of the fuel tax in the process.  Beyond that the question may be asked as to who should set those charges, with an argument that could be made that a distance charge should be a fee - not a tax - set by companies running the roads themselves, with taxes on motorists above the revenue needed for roads, levied on top of this.  That's an answer for taking a more consumer and commercial approach to road management.

What's interesting is how it is the replacement of "unfair" fixed taxes in Europe that are driving interest in charging cars on a distance basis, unlike the US where it is the difficulties politically of raising fuel tax and the growing fuel efficiency of vehicle fleets undermining that tax.

Finland is not immune from the latter, and so should consider how it can protect itself from that, within the boundaries of EU law.  It may also, from an industrial policy perspective, consider itself a pioneer if it can start to move away from fuel tax, as other European countries will face this in due course too.

Finally, Finland ought to look at what has gone on in New Zealand with electronic road user charging on all roads.  It wouldn't be the first country to charge (some) cars by distance on all roads.

No comments:

Post a Comment