Optimal Taxes and Charges for Transport and Energy

David’s Directorship of the DAE beginning in 1988 coincides with the flourishing of his applied work, albeit strongly rooted in the micro-foundations of his earlier research. This work was very much focused on transport and energy. Although his energy work is the better known of the two, I particularly appreciate his research on transport and start with this.

Following his work on optimal taxes in a developing country context, David became very interested in the pricing of goods in developed countries that were often mispriced by the standards of optimal tax theory. One such target of his writing was the pricing of road use, where David became a leading public advocate of the use of road pricing in Britain in the 1990s. Indeed, it was this work which attracted the most media attention in his career. I still remember—in the early days of my own collaboration with David—waking up to hear him being interviewed about road pricing at a busy junction on BBC Radio 4’s flagship morning news programme.

David’s public views on road pricing arose directly from his writing. The central problem to be addressed was how to appropriately recover the costs of the road system from its users. This was addressed in his 1988 paper ‘Road User Charges in Britain’ (Newbery 1988a). This paper discussed the theoretical basis for road charging, calculated the likely amounts raised from optimal charging—for congestion, road damage, and accidents—and compared this to the actual charges paid by road users. Among the themes that emerged were the fact that passenger cars should be heavily taxed because of their contribution to congestion and accidents, not because of their contribution to road damage (which is negligible), while optimal charging would only collect 40% of the total damage costs imposed by heavy goods vehicles (HGVs). Overall, road-related taxes only recovered 70% of total cost, indicating significant room for improvement in tax policy. David’s conclusions suggested that the failure to properly account for congestion and accidents in charging meant that decisions on future road investment were unlikely to be sensible. Indeed, a central implication of his work emerges here: new roads, which reduce congestion and accidents, while raising additional fuel taxes, were likely to be highly socially beneficial and self-financing for the Treasury.

David continued his interest in road damage costs in his 1988 Econometrica paper: ‘Road Damage Externalities and Road User Charges’ (Newbery 1988b). This paper won the Frisch Medal of The Econometric Society in 1990, awarded every two years. It theoretically explored optimal road damage charging. What the paper shows is that under certain circumstances ‘the externality caused by vehicles damaging roads, which raises the operating cost of subsequent vehicles, exactly cancels out when averaging over roads of different ages’ (ibid.: 313). This has the implication that if vehicles are charged per mile in proportion to the direct damage they cause, that would be optimal. This paper further shows that optimal charging of maintenance costs is likely to under-recover such costs while optimal charging for congestion will over-recover marginal capacity costs. Thus, considering both costs together might yield optimal charges which both provide optimal price signals and recover total road network costs. This holds out the possibility that rebalancing road user charges on an optimal tax basis might allow the road system to be self-financing.

This last idea was further explored in Newbery (1989). Here, David shows theoretically and empirically how an optimal road user charge for maintenance and congestion will cover capital and maintenance costs for the road network in the UK. This is interesting because the road user charge should be levied on an equivalent standard axle (ESA) factor basis (which measures the damaging power of each vehicle axle), while congestion should be measured on a passenger car unit (PCU) basis (i.e. relative to the congestion of a representative car). On an ESA basis, an HGV is in orders of magnitude more damaging than a car but in terms of congestion only represents 2-3 PCUs. David’s empirical results suggest that while current aggregate road charges may be in line with costs, the misalignment between actual charges and optimal charges means that road investment decisions are unlikely to be optimal. He thus advocates road pricing—well ahead of the Electronic Road Pricing in Singapore (which began in 1998) and the London Congestion Charge (from 2003).

David’s most comprehensive paper on road pricing is his Oxford Review of Economic Policy paper from 1990. This paper is familiar to generations of Cambridge undergraduates because it includes all of the key diagrams in David’s lectures on road pricing to final year students. In this paper, a number of economically correct (but often unpopular with politicians and environmentalists) arguments are strongly made. These include the facts that efficient road pricing would probably justify further road expansion (by correctly showing the positive NPV and self-financing nature of such investments) and reduce the quality-adjusted costs of public transport (a positive externality!). This point about public transport would be especially true if the extra revenues generated by substitution away from using correctly priced urban roads were spent on improving quality. This last point was amply justified by the subsequent positive experience of exactly this happening in the case of the London Congestion Charge.

Another key idea advanced by David (in Newbery 1994) was the idea that a public road authority should be created, drawing on the experience of private regulated network monopolies. The idea was that a commercial entity should be created to own and operate the road network in Great Britain. This entity (which would not necessarily have to be privatised) would have a balance sheet and could consider investment decisions on a commercial basis. The main point of doing this would be to allow proper financial decisions to be made about road investment, following the arguments in David’s earlier papers about road charging. It would also allow a more sensible debate about the costs and benefits of road investment and allow the roads authority to borrow to invest where there were socially profitable (and often financially profitable) road projects. This is an idea whose time has not quite come, but it remains a powerful suggestion, which would free road investment from the political business cycle and represent a considerable supply- side benefit to UK plc.

A favourite paper of mine on road transport is David’s 1995 piece discussing the Royal Commission on Transport and the Environment (Newbery 1995a). This Commission reported in 1994, recommending a substantial rise in fuel duty as part of the effort to avoid the forecast doubling of road traffic in the UK out to 2025. David’s economic hackles are rather wonderfully raised by this presumption that restraining road transport would be good for society. A particular focus of the paper is the ignoring, by the Commission, of the congestion externality and its overestimation of the emissions impacts of road transport. David sharply points out that doubling fuel duty, as the Commission recommended, would massively distort the relative taxes on emissions between sectors (by raising it by ?600/tonne of carbon) and could not be justified on optimal tax grounds. He also makes the point, with which I still delight my own supervision students, that the problem in the UK—given the huge cost of road congestion to the economy—is too few roads, not too many!

Like his work on transport, David’s initial interest in energy was linked to optimal pricing. David’s early work in this area examined the dynamic consistency problem of government policy (similar to Kydland and Prescott 1977), applied to energy taxation. Fundamental to David’s approach was the analysis of the pricing problem faced by oil importing governments. David’s papers in this area took as a given that the producers had market power, but sought to model the impact of including the fact that consumers (large oil importing countries) also had market power. He did this in the context of the theory of exhaustible resources, where the producer price today had to be arbitraged against the price tomorrow, and hence a dynamic schedule of producer prices and taxes had to be calculated.

Newbery (1981) focuses on oil producers, the oil market being characterised as having a Stackelberg leader (cartel) facing a competitive fringe. This paper shows the extent to which the presence of the competitive fringe undermines the power of the cartel and provides incentives to renege on agreements among producers and with consumers.

The problem examined (with Eric Maskin) in ‘Disadvantageous Oil Tariffs and Dynamic Consistency’ (1990) was that the oil importing country with market power optimally wanted to define a time series of import taxes (known as optimal open loop taxes) which maximised its social welfare. It would likely be optimal to commit to higher taxes tomorrow in order to drive down prices today. However, this would be subject to a time inconsistency problem, in that when tomorrow came, it would be optimal to renege and reduce taxes—from their pre-announced level—in order to reduce consumer prices and increase oil consumption. This paper also suggests that the presence of low-cost storage might act as a strategic commitment to solve the time inconsistency problem.

The theoretical modelling of the market power of oil importers is further explored with Larry Karp in their 1991 paper, ‘OPEC and the U.S. Oil Import Tariff. This paper notes the similar concentration of oil consumers and oil producers and models OPEC’s position in the oil market as being that of a symmetric duopolist with a competitive fringe, the authors noting ‘it is somewhat surprising that no-one has proposed this solution concept before’ (ibid.: 305). The importing countries in this case should impose optimal import tariffs. The overall impact of the interaction between the three players in the market is that the initial price falls. The optimal modelled US oil import tariff is initially around half the final US consumer price.

This paper was followed by another with Larry Karp in 1992: ‘Dynamically Consistent Oil Import Tariffs’. This won the Harry Johnson Prize from the Canadian Economics Association in June 1993, for the best article published in the Canadian Journal of Economics in 1992. The paper continued David’s investigation of the dynamic inconsistency problem of oil taxation. It explores the differences between buyers and sellers of oil with market power. The results show that buyers with market power are more likely to be dynamically inconsistent while sellers are less likely to be so. The intertemporal price arbitrage of an exhaustible resource turns out to be a dominant effect, which means that complicated intertemporal oil import tax variations, for large importers, have relatively small effects on social welfare. Once again, the USA could benefit from the imposition of significant oil import tariffs. This paper is a good example of David showing how sophisticated theory does not necessarily support complicated (i.e. time-varying) government pricing, but does support obvious pricing of externalities (in this case the consumption externality of oil imports into a large country).

In his 1992 paper, ‘Should Carbon Taxes Be Additional to Other Transport Fuel Taxes?’, David combines his interests in energy and transport. The title question arises because transport fuel is already heavily taxed—relative to other sources of carbon emissions—in the UK and many other countries. The answer, according to David, is that if anything, the transport fuel tax should go up by at least the carbon tax content equivalent amount. This requires the maximisation of utility less the cost of gasoline, carbon emissions, road use, and congestion, subject to the presence of one pricing instrument—fuel tax. This theoretical paper contains a rather brilliant tax argument to explain why:

A carbon tax will lead to a reduction in the fuel used per km driven because it will encourage greater fuel efficiency. This in turn will reduce the tax base on which the congestion charge per km is to be levied and raise the required road user charge per litre (ibid.: 54).

David’s most comprehensive paper on energy taxation is his 2005 ‘Why Tax Energy? Towards a More Rational Policy’ (Newbery 2005a). This paper brings together the insights from his work on both transport and energy. A central theme of the paper is that the differentials in energy taxes between countries and within countries on different fuels cannot be justified on optimal tax grounds and should be harmonised further. These differentials substantially distort trade, especially within the EU Single Market. The paper contains a good summary of David’s work on optimal oil import taxes. This shows that the oil taxes in the EU were roughly optimal under certain assumptions, but that by implication, natural gas and coal (given its relatively high carbon content) taxes were far too low, being close to zero in many countries.

‘Why Tax Energy?’ is an important question because as an intermediate good, it is not clear why it should be taxed in a Diamond and Mirrlees (1971a, b) world. However, as David clearly argues, energy taxes can serve as optimal import tariffs, environmental externality prices, and road user charges. But he expresses doubts about the ‘double dividend’ argument for energy taxation, which suggests that energy taxes increase overall welfare by allowing the reduction of other distortionary taxation. This is because tax rates relative to other goods are likely to be close to being welfare optimal already, especially when taking distributional arguments into account, while higher energy taxes worsen income distribution. As ever, David wants to take us beyond the simple theory to theory which has taken the empirical realities into account.

< Prev   CONTENTS   Source   Next >