The Department for Transport’s Decarbonisation Plan projects the decline of domestic transport greenhouse gas emissions from the present 120 MtCO2e a year to approaching zero by 2050 (see figure above). There is considerable initial uncertainty about the pathway but the range of projected outcomes narrows over time as the proportion of zero emission vehicles increases. The modelling is largely based on the Department’s long-established National Transport Model. Although little detail is provided, there are headline numbers for cumulative emission reductions over the period 2020 to 2050.

For cycling and walking, the projected savings from investments and policy initiatives are put at 1-6 MtCO2e, a notably wide range. For cars and vans, the savings are estimated to be 620-850 MtCO2e, a proportionately narrower range, reflecting greater confidence in the impact of policy to phase out the internal combustion engine. What is striking is the relatively tiny decarbonisation contribution expected from increases in active travel, at best one percent of that from car decarbonisation. This is surprising, given the prominence in the Plan of the intention to promote active travel, including expenditure of £2 billion over five years. This looks like a case of virtue signalling by DfT, wanting to be on the side of the angels.

Counting on minimal decarbonisation benefit from more cycling is consistent with the evidence from Copenhagen and other European cities that you can get people off the buses onto bikes but that it is difficult to get them out of their cars. In any event, 80% of car carbon emissions arise from trips of more than five miles, implying limited scope for savings from mode switching from car to active travel.

More generally, the DfT’s Plan relies very largely on technological innovation to achieve net zero for domestic transport by 2050. Others see a need for significant reduction in travel demand, including the Climate Change Committee (CCC) in its Sixth Carbon Budget report of December 2020, which envisages 20% of transport emission reductions from reduced demand. The CCC recommended commitment to 78% reduction in overall emissions by 2035 compared with 1990 levels. This was accepted by the government and is to be implemented through sector plans, of which the Transport Decarbonisation Plan is the first. 

The question of the need for travel demand reduction is crucial, given this could be both unpopular and difficult to achieve. One can see why the DfT might shy away from measures to reduce demand, such as significant increases in the cost of travel. But aside from the political sensitivities, could such measures be justified on the basis of existing models that generate conflicting conclusions and whose validity is unproven?

The National Transport Model is an elaborate model of the surface transport sector. Other relevant models are essentially energy models of the whole economy, including the transport sector. All these models are complex and opaque, with many parameters whose magnitude requires professional judgement. Given the timescale to 2050, it is not possible to validate models by comparing forecast with outturn. Models are therefore prone to optimism bias, whether unconsciously because modellers want to please their clients, or consciously in aid of achieving some higher purpose.

Greater transparency of the National Transport Model would allow us to understand whether there has been undue optimism about the prospects for decarbonisation by technology. However, such transparency seems unlikely. The DfT has always resisted allowing others to use its model on the grounds that its components employ proprietary software developed by consultants. Not that this is unique. Most transport modelling utilises proprietary models owned by consultants. This contrasts with practice elsewhere.

The Treasury’s model of the UK economy has been available for external use for many years. The Department for Business, Energy and Industrial Strategy has developed its energy modelling in close collaboration with academia and plans to increase transparency. Climate modelling is an international, open, collaborative effort that feeds into the findings of the Intergovernmental Panel on Climate Change. The epidemiological modelling of the coronavirus pandemic, which has informed decisions on lockdowns and vaccine deployment, has been carried out not within government, but by university modellers, collaboratively and transparently.

Transport modelling needs to move on, to become transparent and collaborative rather than opaque and proprietary. More effort needs to be devoted to validating models by comparing forecast with outturn where that is possible, for instance over the initial years following the opening of a new element of infrastructure. For the period through to 2050, the best that can be done is for modellers to run their models on common assumptions, to understand why forecasts differ, and then to vary assumptions to test the sensitivity of forecasts to bias, both optimism and pessimism, whether concerning technological innovations or behavioural change.

We need an informed consensus from the modellers of transport decarbonisation to inform the development of policy.

The text above was published in Local Transport Today edition of 18 October 2021. Since this piece was drafted, two further relevant publications have become available.

CREDS, the Centre for Research into Energy Demand Solutions, a consortium of university groups, published a substantial report on the role of energy demand reduction in achieving net zero, including the energy associated with the transport system. The report concluded that the UK could halve its demand for energy by 2050, which would substantially ease the task of meeting that demand with zero carbon emissions. For transport, an ambitious set of assumptions were made, including that single occupancy car use becomes socially unacceptable and that the car fleet is reduced substantially. The model employed represents the whole energy system and is known as UK-TIMES. The model and assumptions are set out in some detail and the code has been published, which is very creditable, although it would take a professional modeller to fully appreciate the content.

The government has just published its Net Zero Strategy. This covers the whole economy including transport, although little is added to the previously published Transport Decarbonisation Plan, which placed minimal reliance on changes in travel behaviour. A technical annex sets out the modelling and assumptions used to justify the pathway to net zero by 2050, again employing the UK-TIMES model. For transport, the only behavioural change assumed is that the share of journeys in towns by active travel increases from 42% in 2019 to 55% in 2035. The trajectory of emission reduction for domestic transport on page 154 is similar to that shown in the Department for Transport decarbonisation plan, although the modelling framework used is different.

It is evident that the emission consequences of a wide range of travel behavioural change possibilities are being projected using different kinds of model. The recent publications reinforce the case for transparency and collaboration amongst the modellers.

The research literature – papers on aspects of travel and transport in peer-reviewed journals – has burgeoned in recent years. There are more papers in established journals and new journals created, often on an open access basis whereby the researchers pay the cost of publication, rather than journals relying on Libraries taking out subscriptions. The commercial basis of these new open access journals is not always clear, but certainly some are operated by for-profit publishers. There may therefore be an incentive to relax standards in the peer-review process to generate more income, lessening the overall quality of the research literature, which accords with my subjective impression. Some of the not-for-profit open access journals appear to lack editorial oversight by academic researchers.

One feature of many recent publications is the theoretical modelling of a new technology. This may be useful where there is a clear practical need, for instance the optimal deployment of charging points for electric vehicles. Yet there is also extensive modelling activity in relation to the deployment of autonomous vehicles (AVs), where experience of on-road behaviour is extremely limited thus far. Because model outputs depend on assumptions about AV performance parameters, the conclusions of such studies are very varied and provide little in the way of useful guidance to practitioners and policy makers.

Another feature of the literature is the excessive formal analysis of survey findings, for instance of the responses to surveys of the expected impact of a new technology, such as AVs, whether of drivers or city planners. State of the art analysis is reported in tabular form, with statistical significance specified numerically. Rarely are findings reported as charts, bar charts or scatter diagrams, with uncertainty shown visually, which would make clear the common limited significance of the findings.

A further feature of the recent literature is the systematic review, in which formal search methodologies are employed to identify all relevant papers on a topic. One problem is that because of the deteriorating quality of the literature, it becomes difficult to see the wood for the trees, as every paper needs to be cited. Systematic review originated in the medical literature where the aim of such meta-analysis is to identify every relevant study of a condition or treatment, with a ranking by quality such that only the highest quality papers contribute to the conclusions of the review. But for transport studies, such quality ranking is not practical, in part because findings may be specific to particular locations or circumstances.

Another problem with formal searching of the literature is that relevant papers may be missed because of the difficulty of specifying appropriate search terms. A recent paper by a distinguished transport researcher addressed a topic on which I had published some years ago without mentioning my contribution. When I raised the matter, I received an apology that his search had failed to identify my papers.

I have noticed increasing reference in the recent literature to transport researchers as ‘scholars’, a term hitherto largely reserved for those working in the humanities. Generally, those involved in transport research have seen themselves as based in disciplines such as engineering, economics, planning and the environmental sciences. The purpose of research within such disciplinary frameworks has been to advance understanding and thereby contribute to practical solutions to the problems of the transport sector. We have not, I think, seen ourselves as primarily involved in developing a branch of knowledge through scholarship that focuses on the extant literature. Indeed, the inward-looking processes of scholarship are cluttering up the literature with findings of little use and thereby may be displacing contributions of more practical value.

For instance, I have been attempting, without success, to get published in a peer-reviewed journal a paper on Digital Navigation, by which I mean the combination of satnav, digital mapping and route guidance algorithms that are in widespread use by road users. Highways Magazine, read by practitioners, has published a short account of my analysis, but a fully documented paper seems not to fit the current fashion for what’s hot, as seen by journal editors.

I have an article published in Highways Magazine, the text of which is below.

Recent advances in a number of digital technologies in combination are having a significant impact on travel behaviour on the road network by providing route guidance that takes account of traffic conditions. What may be termed ‘digital navigation’ involves the use of satellite navigation (satnav) to provide spatial positioning to high precision; digital mapping; the ability to detect vehicle speeds and hence the location of traffic congestion; and routing algorithms to optimise journeys. The combination of satnav location and digital mapping provides a navigation service that offers turn-by-turn route guidance.

While digital navigation is in widespread use by road users, remarkably little information is publicly available about performance, in particular how routes are optimised, the suitability of recommended routes, the accuracy of estimated journey times, and the impact on the functioning of the road network as a whole. Nevertheless, there is evidence to indicate an impact on the use of minor roads, of major roads, and on traffic congestion and the optimisation of the road network.

Recent revisions to British road traffic statistics appear to show that there has been a substantial growth of motor vehicle traffic on minor roads in recent years, an increase of 26% between 2010 and 2019, while traffic on major roads increased by only 12%.  One factor contributing to this growth is the increase in van traffic, including that arising from the growth of online shopping with home deliveries. However, in 2019 van traffic amounted to 15% of traffic on urban minor roads, and 19% on rural minor roads, cars being responsible for 82% and 78% of traffic respectively. So, the growth of van traffic on minor roads has been responsible for only part of the overall traffic growth on these roads.

The most likely main contribution to the large growth of traffic on minor roads is the widespread use of digital navigation, which makes possible the general use of minor roads that previously were largely confined to those road users with local knowledge, as well as extending such local knowledge. Diversion to minor roads is likely to occur when major roads are congested and represents an effective increase in the capacity of the road network, so generating additional traffic.

As well as encouraging use of minor roads, digital navigation may divert traffic from local roads to roads intended for longer distance traffic. One case that I have analysed where such diversion may have occurred is the widening of the M25, the London orbital motorway, between junctions 23 and 27 to the north of the city. There was substantial growth in traffic above the level that had been forecast, much of which arose from diversion of local trips, such as home to work, to take advantage of faster travel on the motorway, despite the greater distance and higher fuel costs incurred. The contribution of digital navigation in facilitating such diversion cannot be inferred from available data, but it is plausible. Regular users of digital navigation would have up-to-date information for each journey, while irregular or non-users would likely be aware that diversion to the motorway would offer the fastest journey.

The M25 case study suggests that local traffic may be expected to take advantage of the capacity increase of major routes in the vicinity of urban areas that generate much traffic, which are the locations where the Strategic Road Network is under greatest stress and where investments to increase capacity are thought to be most needed. However, this local traffic negates the benefits expected for long distance road users and so undermines the economic case for the investment. The growing use of digital navigation would tend to contribute further to weakening the case for such investment.

While the M25 case study is an illustration of the maxim that we can’t build our way out of road traffic congestion, nevertheless the development of digital navigation offers probably the best means available to mitigate the impact of congestion. Congestion arises in or near areas of high population density and high car ownership, where the capacity of the road network is insufficient to cope with all the trips that might be made. Drivers are deterred by the prospect of time delays and so make other decisions – to travel at a different time, by a different route, by a different mode, to a different destination (where there are options, as for shopping), or not to travel at all (by shopping online, for instance). Congestion is therefore substantially self-regulating, in that if traffic increases, delays worsen and more potential users are deterred on account of the time constraint.

Digital navigation that takes account of congestion in real time can offer less congested routes, so making better use of the existing road network and reducing road users’ exposure to congestion. One problem that may arise is that traffic may be diverted on to unsuitable roads, where local environments and neighbourhoods may be adversely affected, or even where large vehicles can become obstructed. Diversion onto unsuitable routes is a problem that could be mitigated through collaboration between digital navigation providers and road authorities.

Beyond the rerouting of traffic to less congested roads, there is a feature of digital navigation that mitigates the unwelcome experience of traffic congestion – the prediction of journey time, or estimated time of arrival (ETA). When road users are asked about their experience of congestion, both in surveys and in discussion, the evidence from their responses indicates that the uncertainty of journey time is a more important adverse consequence than lower speed. Accordingly, an important benefit of digital navigation is the forecast of ETA in the light of prevailing traffic conditions on the selected route, in this way substantially reducing journey time uncertainty.

While diversion onto less congested routes may be helpful for users of digital navigation, there is a question as to whether this is optimal for users of the road network as a whole. Digital navigation employs proprietary algorithms whose performance is difficult to assess externally. An algorithm might response to build up of congestion by diverting all traffic to a single alternative route until that became congested, repeating the process to spread traffic across available routes until congestion abated. Or the algorithm might spread traffic across all available routes at the outset. And the algorithm might anticipate the build-up of congestion based on historic experience. But in any event, the routing algorithm used by one provider would not take account of the activities of another provider. The providers of digital navigation services are very secretive and there is almost no published information on their design and performance.

The road system is generally well regulated to achieve safety and efficiency. Given the potential scale of impact of digital navigation devices on network operations, arguably a licensing regime would be appropriate for providers. This might require information to be exchanged with road authorities, guidance to be accepted to avoid adverse environmental and social impacts, and mutual collaboration to optimise the operational efficiency of the network as a whole, while at the same time optimising the experience of individual road users.

The Department for Transport has published its 218-page detailed Decarbonisation Plan. Boosting cycling and walking features first amongst the measures for reducing the carbon emissions of individual modes of travel, with £2 billion to be invested over five years to make active modes the first choice for many journeys. The aim is to have half of all journeys in towns and cities cycled or walked by 2030.

However, the expected carbon reductions are only 1-6 MtCO2e over the period 2020 to 2050 (page 53 of the Plan), which is tiny in relation to UK domestic transport carbon emissions of 122 MtCO2e in the single year 2019 (page 15).

In contrast, the emission reductions expected from electrification of cars and vans amounts to 620-850 MtCO2e over the period 2020-2050 (page 87).

The DfT’s very small expectations of carbon reduction from increased investment in cycling is consistent with the evidence that you can attract people off the buses onto bikes, but it is much harder to get them out of their cars, as seen in Copenhagen and elsewhere.

Altogether, it seems that the DfT’s apparent enthusiam for cycling is virtue signalling, with low expectations of real carbon reduction benefits.

This blog post is the text of an article in Local Transport Today.

Cycling is widely advocated as a desirable means of travel – healthier, cheaper, more environmentally friendly and barely slower than the car for short-to-medium length trips. The Government seeks a step-change increase in cycling with £2 billion new funding, as a cost-effective way of reducing transport carbon emissions.

Certainly, there is substantial scope to increase cycling by investment in better infrastructure, witness Copenhagen with dedicated cycle lanes on all major roads, where 28% of all trips are by bike, compared with 2.5% in London. So when, at the outset of the pandemic, the Mayor of London announced his ambition to increase cycling by tenfold, you could see that this should be possible with the requisite investment. However, when you’re in Copenhagen, you are aware of the considerable amount of general traffic (and viewing Scandi noir crime dramas set in that city, you see very few of the characters using a bike). In fact, with 32% of all trips by car, Copenhagen is only slightly less car-dependent than London with 35%.

Aside from cycling, the other big difference between these two capital cities is that public transport use in Copenhagen is only half that of London, 19% versus 36% of trips. This indicates that you can get people off the buses onto bikes, but that it is much harder to get them out of their cars, even in a small, flat city with excellent cycling facilities where almost everyone has experience of safe cycling. Yet we don’t want to diminish the use of buses, which are an efficient means of moving people in urban areas, the diesel engines of which can be replaced by electric or hydrogen propulsion. Fewer bus passengers mean less fare revenue and less frequent services.

Data for other European cities indicate that Amsterdam is similar to Copenhagen, with 32% of trips by bike and 17% by public transport. In marked contrast, both Zurich and Vienna have excellent public transport responsible for 40% of trips, with cycling accounting for only 7-8%. More generally, while the pattern of urban travel reflects both local geography and history, we don’t find cities in developed economies with high mode shares of both cycling and public transport.

In seeking to reduce transport carbon emissions, we should be careful not to underestimate the attractions of the motorcar, which is useful for longer journeys and for shorter trips with less sweat, for carrying people and goods, including child seats and the stuff left permanently in the boot. The car is well-suited for meeting our needs for access to people and places, for door-to-door travel where there is road space to drive without unacceptable congestion delays and the ability to park at both ends of the trip.

But there is more to car ownership than the ability to go from A to B. The growth in popularity of SUVs suggest that there are feel-good factors that motivate purchase of these costly vehicles (it would be interesting to see the findings of the market research carried out by the car manufacturers, regrettably proprietary). The fact that cars are generally parked for 95% of the time is a good economic argument for car sharing. But conversely, this also indicates the value we place on individual ownership, to have vehicle available when we want it, a vehicle that reflects our personal consumer preferences. Cars are not unique in this respect. My washing machine sits unused more than 95% of the time. I could share with others at the laundrette, but it’s more convenient to have my own.

Car sharing in its various forms is advocated as a means to reduce car use, road traffic congestion and carbon emissions. Sharing has been facilitated by online digital platforms, which have been transformative of many aspects of the economy. For travel, we have the disruptive impact of ride-hailing as exemplified by Uber, and of online booking of trips by rail and air. By contrast, the growth of car sharing has been relatively slow, indicating the development of niche markets, with substantial replacement of private ownership looking unlikely.

Where road capacity limits car use in city centres, both public transport and active travel are attractive alternative modes. Agglomeration economics have led to increased population density in successful cities, which shifts travel away from the car. The growth of higher education in urban centres has contributed to reduced car use by young adults.  However, these trends may weaken post-Brexit and post-Covid. And while car use can be impeded in low traffic neighbourhoods in favour of cycling and walking, the aggregate impact may not be great.

We need to be careful to avoid optimism bias when projecting the impact of measures to reduce transport carbon emissions. The models that are used for this purpose are complex and opaque, with many input assumptions and parameters to be specified. Optimism bias arises when modellers make choices, consciously or unconsciously, that tend towards achieving a strategic purpose. Yet optimism bias leads to outcomes that fall short of those that are forecast. 

It is now part of the culture of transport planning to place emphasis on the opportunities for promoting cycling. But caution is needed. When addressing the impact of changing mode share, attention should be paid to the modes from which the shift to cycling is expected. For instance, the well-established Propensity to Cycle Tool, which assesses the potential to increase the amount of cycling, assumes that commuters are equally likely to shift to cycling from any prior mode. However, the evidence from Copenhagen and elsewhere indicates that a shift to cycling from public transport is much more likely than from car use, which would substantially reduce the carbon reduction benefits assumed from boosting cycling.

If optimism bias informs assumptions about mode shift from cars to bikes, or about the scope for car sharing, then disappointment is likely to ensue.

The House of Commons Transport Committee is holding an inquiry into zero emission vehicles and road pricing. I submitted evidence set out below.

Main points

  • There is a case for road pricing both to replace fuel duty revenues lost as ZEVs replace conventional vehicles and to help manage road traffic congestion.
  • The charge for road use might comprise two elements: one generating a revenue stream for the Exchequer and another for the local authority, which would allow substantial devolution of responsibilities for transport provision.
  • There would be attractions in the incremental introduction of national road pricing, building on the successful congestion charging arrangements in London.

Rationale for road pricing

The move to ZEVs will result in the loss of revenue from road fuel duty, as well as from VED were ZEVs to remain zero rated. Revenue from the former amounts to some £28 billion a year and from the latter some £6 billion. This prompts consideration of some form charging for road use to make up the loss.

The case for zero VED for electric vehicles (EVs) is to incentivise their uptake, a reason that will become irrelevant as the capital cost of EVs falls and as sales of new conventional vehicles are phased out. So, in due course VED could be applied to road vehicles generally, and if set at a rather higher rate could cover the annual cost of capital and current expenditure on national and local roads of £8 billion, obviating the need for road pricing to ensure that drivers pay for the roads they use.

One argument some make for road pricing is that without fuel taxation or a similar charge related to distance travelled, the running costs of EVs would be substantially lower than that of conventional vehicles, which would result in more miles travelled and thus more congestion, carbon emissions and other externalities. One scenario of the Department for Transport’s (DfT) 2018 Road Traffic Forecasts illustrates this expectation[1]. However, in recent years the average distance travelled by car has remained stable, being limited by the time available for travel, the speed of travel and the proportion of households owning cars, none of which have increased in this century. It is therefore not to be expected that the replacement of the internal combustion engine by the electric motor would have much impact on vehicle use.

Another argument for road pricing is to alleviate road traffic congestion, the intention of the London congestion charge. Experience has shown that reduction of congestion is quite limited at the level of charging typically employed, particularly in a prosperous city like London where many are able to afford the charge[2]. Charging for road use benefits those who can readily afford to pay by displacing those who are less able, generating increased inequalities in use of the road network that historically has been a relatively egalitarian domain. Nevertheless, congestion relief is in principle a possible aim of the road pricing regime, although the magnitude of the charge would need to reflect both the level of congestion and affordability in the locality if congestion is to be effectively ameliorated. Related to this is charging polluting vehicles, as in London’s Ultra Low Emission Zone (ULEZ), the rationale for which will diminish over time as EVs are increasingly used, yet which may remain relevant in respect on non-tailpipe particulate releases from vehicles.

Revenue from the London congestion charge and the ULEZ are retained by the city authority, as are charges for low emission zones planned elsewhere, ring-fenced for expenditure on transport services, and likewise revenues from parking charges. A national scheme for road user charging might comprise revenues both for the Exchequer as well as for local authorities, the latter setting levels of charges to reflect local conditions, including congestion and other environmental impacts of traffic, as well the need for revenues for road maintenance. There would be attractions in allowing local authorities to set their share of the road user charge to cover the full cost of local transport provision, obviating the need for grants from the DfT (other than perhaps for ‘rebalancing’ purposes). The Exchequer element of the charge could depend on the type of road, for instance higher for motorways that are funded nationally, and could vary by region to aid ‘rebalancing’ policies.

Introduction of road pricing

There would be attractions in introducing road pricing for EVs alone, on the rationale that they should pay their fair share of the costs of the road network that conventional vehicles are paying via fuel duty. However, the lower operating costs of EVs are a necessary incentive to purchase while capital costs remain higher. As capital costs fall, as is expected, scope would develop to charge users of EVs by introducing a road pricing regime from which conventional vehicles were exempt.

Introduction of general road pricing on top of fuel duty would be invidious for lower income motorists who are likely to continue to use conventional vehicles for longer than the better off. Accordingly, one possibility would be to introduce a general road pricing system but crediting conventional vehicles with the fuel duty they pay. This is the basis of a voluntary pilot scheme in Oregon[3]. Because this scheme is voluntary, uptake is incremental, in contrast to an obligatory scheme that would have to be imposed all at once.

It is worth considering options for incremental roll-out of national road pricing, given the potential difficulties of overnight national implementation of charging and enforcement technologies. In London, the existing congestion charging system functions sufficiently well and is publicly acceptable, but its scope is limited by the fixed charge for entering the charging zone.

There are a number of incremental developments of the London scheme that might be feasible:

  • Encourage participation via a smartphone app by offering a discount from the standard daily charge;
  • Take advantage of location awareness of smartphones to identify when the user is both in a vehicle registered for the charge and in the charging zone, backed up by camera enforcement as at present;
  • Encourage entry and exit from the charging zone outside times of peak congestion by offering a discount from the standard daily charge;
  • Increase the standard charge but offer discounts to encourage use when and where traffic is less congested;
  • Extend the charging to other areas of London where congestion is a problem;
  • Make the charging and enforcement systems available to other cities that wished to manage local traffic, incentivised by the revenues that could be used to provide alternatives modes of travel to the car.

Once a number of cities were using road pricing, there would exist the basis for national adoption in the form of an established charging system, which would need to be supported by the national roll-out of camera enforcement (unless a better enforcement system could be devised). This would be accompanied by the reduction and eventual abolition of road fuel duty, perhaps with the public assurance of no net increase in revenues from road users. There would need to be a daily penalty charge for those evading payment for road use, which, if not paid when requested, might be added to the annual VED charge, failure to pay which could result in clamping.

Whichever way to bring it about, a decision to adopt national road pricing would need to be strategic, commanding wide acquiescence, analogous to the decision to phase out internal combustion engine vehicles.

Conclusion

Adoption of a scheme of national road pricing would allow loss of revenue to the Exchequer from road fuel duty to be offset. A scheme that generated revenues for both central government and local authorities would allow substantial devolution to the latter of responsibilities for funding the provision of their transport services in the light of local needs. A national road pricing scheme might be developed incrementally from the congestion charging arrangements in London.

21 January 2021


[1] Department for Transport, Road Traffic Forecasts 2018, Scenario 7.

[2] Metz, D. Tackling urban traffic congestion: The experience of London, Stockholm and Singapore. Case Studies on Transport Policy, 6(4), 494-498, 2018.

[3] https://www.myorego.org/

Below are the main points and implications of my analysis of the outcome of widening of the M25 motorway between Junctions 23 and 27, published as ‘Economic benefits of road widening’, Transportation Research Part A, 147, 312-319, 2021. Abstract available at https://www.sciencedirect.com/science/article/abs/pii/S0965856421000872 Manuscript available from david.metz@ucl.ac.uk

  • The M25 motorway was widened between Junctions 23 and 27 as part of the Smart Motorway investment programme implemented by Highways England. Detailed traffic monitoring reports were published before the scheme was opened and for three years afterwards.
  • There was some increase in traffic speeds at Year One after opening, compared with Before opening, but this gain was lost subsequently account of increased volumes of traffic. At Year Three, average daily traffic was up by 16% compared with Before, and up 23% at weekends. This contrasts with an increase of 7% for regional motorway traffic growth.  
  • The conclusion of the Year Three monitoring report states: ‘These results show that increases in capacity have been achieved, moving more goods, people and services, while maintaining journey times at pre-scheme levels and slightly improving reliability.’ However, this could not have been the basis of the investment case, which in general suppose that travel time savings are the main benefit of transport infrastructure investment. Accordingly, reports of the traffic and economic modelling were obtained; these utilised the long-established SATURN and TUBA models.
  • The traffic model projected increased traffic volumes and speeds for the scheme opening year, comparing the ‘do something’ investment case with the ‘do minimum’ case without the investment. However, the increase in traffic volume was less than the observed outturn and the increase in speed forecast failed to materialise beyond the first year after opening.
  • The modelled economic benefits derived very largely from time savings for business users. There were also time savings for local users, commuters and others, but these were almost entirely offset by increased vehicle operating costs. This was the consequence of local users rerouting trips between unchanged origins and destinations to take advantage of short journey times made possible by diverting to the motorway, travelling somewhat greater distances.
  • The benefits forecast for business users were the main input to the economic appraisal that generated a benefit-cost ratio of 2.9, which was the basis for the investment decision. However, the time savings benefits did not materialise beyond the first year after opening, on account of the additional traffic above forecast.
  • The nature of this additional traffic cannot be deduced from the traffic monitoring. It is likely that much, possibly most, comprises local trips rerouting, of no net economic benefit; indeed, these trips would be of negative benefit on account of the additional externalities (carbon etc) arising from the increased distance travelled. The outturn BCR must be much less than the forecast 2.9, possibly even negative.
  • This M25 case is likely to be typical in that the Strategic Road Network comes under greatest stress in or near major urban centres where local traffic competes for carriageway with long distance users. Highways England has 10 smart motorway schemes in its current investment programme, with an average BCR estimated as 2.4. This likely reflects considerable optimism bias in the modelling.
  • The modelling to support decision making distinguishes between different classes of road user, yet the traffic monitoring does not allow such a distinction. The monitoring is therefore of limited use in refining the models and countering optimism bias. What is needed is monitoring of representative samples of road users over time to see how their travel behaviour changes as the result of the road investment. Such longitudinal studies, as they are known, are common in the areas of health and social sciences, but almost unknown for travel and transport.

The House of Commons Transport Committee is holding an inquiry into major transport infrastructure projects. Written evidence has been published. My evidence is set out below.

This submission is concerned with the effectiveness of the Government’s decision-making and appraisal processes for transport infrastructure projects, a subject on which I have researched and published in peer-reviewed journals.

Main points

  • The standard approach to the economic appraisal of transport investments, based largely on the value of time savings, does not reflect reality. An independent review of the methodology would be desirable. 
  • The value of transport investment lies in better access to people and places, making possible more opportunities and choices. Improved access changes the built environment, the consequences of which can be valued.
  • Digital navigation (satnav) is affecting traffic flows, and thus the outcomes of road investment, in ways that are little understood. A study is needed.

Time is not saved

The Department for Transport (DfT) has a well-developed methodology for the appraisal of transport infrastructure investments, set out in considerable detail in its Transport Analysis Guidance (TAG, formerly WebTAG, reflecting the pioneering effort to make the material available via the internet). The Guidance is consistent with the Treasury’s Green Book, and indeed appears to be regarded as an exemplar of good practice in economic appraisal in the public sector. However, the Green Book does not deal with the specifics of transport investment, so that the recent changes to it, while welcome, do not bear on what follows.

The main economic benefit supposed to arise from investment in transport infrastructure is the saving of travel time. Accordingly, the DfT has commissioned considerable analysis designed to attribute values of time to different classes of user of the transport system, according to mode and purpose of travel. Travel time savings typically amount to 80% of overall economic benefits that are set against capital and other costs in the cost-benefit analysis employed to help reach investment decisions.

Average travel time has been measured regularly in the National Travel Survey (NTS) and has barely changed over almost fifty years, at close to an hour a day for travel by all modes except international aviation. So, there is a paradox in that huge past investments in transport infrastructure have been justified by the expected value of travel time saved, which yet is not apparent in any change in travel time averaged across the population.

The explanation is that travel time savings are quite short run. In the long run, people take the benefit of faster travel to travel further, to gain access to more people and places, to have more opportunities and choices. Thus, while average travel time remained constant, the average distance travelled increased from 4500 miles a year in the early 1970s to around 7000 miles in the late 1990s. The economic benefits to users of transport infrastructure investment therefore are not time saved for more work or leisure, but rather relate to enhanced access to, and choice of, jobs, homes, schools, shops and other services.[1]

Such access is subject to diminishing returns. For instance, 80% of urban residents in Britain have access to three or more large supermarkets within 15 min drive, and 60% have access to four or more, a level of choice likely for most to remove the need to travel further for yet more choice[2]. On the other hand, access increases with the square of the speed of travel, since what is accessible is defined by the area of a circle whose radius is proportional to the speed of travel. The combination of access increasing with the square of travel speed yet subject to diminishing returns means that travel demand to achieve access saturates, that is, it ceases to grow.

Such demand saturation, also known as market maturity, is of course a standard feature of consumer markets generally and it is to be expected for daily travel. The finding of the NTS that the average distance travelled has not increased over the past twenty years is consistent with this expectation. So rather than planning for ever more transport infrastructure investment based on time savings, we should limit investment to meeting specific access deficiencies in what is generally a mature transport system. This points, for instance, towards investment in urban rail that can increase economic density and hence agglomeration benefits, as the National Infrastructure Commission has argued[3], rather than on inter-urban roads where the benefits are overstated, as the following case study illustrates.

In 2015 the London orbital M25 was widened from three to four lanes in each direction between junctions 23 and 27, part of the so-called ‘Smart Motorway’ programme of Highways England, with detailed monitoring of traffic volumes and speeds before and for the first three years after opening. The growth of traffic was substantially greater than on major roads in the region and greater than forecast in the transport modelling used to justify the investment. The model predicted a significant increase in traffic speed, which generated travel time savings contributing to a benefit-cost ratio (BCR) of 2.9, which represented high value for money. However, beyond the first year, no increase in traffic speed was seen on account of the additional traffic. The evidence suggests that this is due mainly to local users rerouting between unchanged origins and destinations to take advantage of shorter journey times via the motorway, while incurring greater fuel costs. The outturn BCR must be quite small.[4]

This M25 case is likely to be representative of much planned investment in new capacity on the Strategic Road Network, which comes under greatest stress in or near urban centres where local traffic competes for carriageway with the long-distance users for whose benefit the investment is primarily intended. Rerouting to take advantage of new capacity is facilitated by the widespread use digital navigation (satnav) devices that offer routes with the shortest time.

At present, optimism bias in modelling means that traffic growth is underestimated and time savings overestimated. Monitoring of traffic flows before and after opening is too crude a measure to understand the consequences of the investment for the different classes of road user. Accordingly, we need to monitor the changes in travel behaviour of a representative sample of users, employing the travel diary technique as used for the National Travel Survey. This would allow transport models to be better calibrated, such that changes in access could be identified and valued, and externalities (carbon emissions, pollutants etc) that are related to vehicle-miles travelled better estimated. In particular modelling needs to take account of the impact of digital navigation on traffic on the road network.

Built environment is changed

The focus on travel time savings in the standard approach to appraisal means that the impact of transport investment on the built environment is not properly taken into account. Consider a proposal to construct a bypass around a village, motived by concerns about the local environmental impact of traffic. The economic case would be based largely on the value of travel time savings from a faster route. However, a bypass may make land more accessible for development, for instance for housing, subject to decisions of the planners and the prospective return to developers. A bypass scheme with housing is clearly different from one without, as regards both traffic and economic benefit. The standard approach to appraisal disregards the benefit of new housing when estimating the BCR on the grounds that this would double count the user benefits, which might be shifted to others such as land owners but which would not change in overall magnitude. In reality, the impact of the scheme with housing is very different from that without and they need to be appraised separately.

More generally, the real-world outcomes of transport investments depend on decisions by planners and developer. It has been attractive for the DfT to operate in a silo, disregarding changes in land use, initially on grounds of simplicity that were perhaps justified in the heyday of motorway construction. But now that we have a mature network of transport infrastructure in place, with only fairly marginal increases feasible on account of the high cost of civil engineering work, we need to focus on the benefits of new schemes beyond the traditional user benefits. Decision-making needs to be tripartite, involving planners, developers and transport authorities.

The broad objectives of investment in transport infrastructure are threefold: to stimulate economic growth; accommodate population growth; and mitigate environmental harm. Accordingly, we need an approach to appraisal that helps reach investment decisions relevant to these objectives, or whatever more specific versions may be decided by those holding devolved budgets. Changes to the built environment need to be recognised explicitly since they are important to achieving objectives, whether to make sites accessible for new housing or for business expansion. Moreover, changes to the built environment are spatially located, whereas time savings are not, yet location of benefits is very relevant to investment decisions. For instance, the economic case for HS2 was based largely on benefits to users of the new route and was silent on spatial distribution, whereas the strategic objectives were concerned to rebalance the economy in favour of regions beyond London[5].

An innovative approach to decision-making has been developed by the National Infrastructure Commission for its recent assessment of rail needs for the Midlands and the North[6]. This focuses on the way in which improved rail services can increase the effective density of city centres, which has long been recognised as boosting the productivity of businesses from agglomeration benefits through more efficient labour markets, better supply chains and enhanced knowledge sharing. The Commission has extended this analysis to capture the consumption impacts of agglomeration through access to increased amenities, which replaces conventional time saving benefits. Separately, the Commission has developed a property value uplift tool that allows the estimation of the impact of transport investment on property prices[7].

Digital technologies

The widespread use of digital navigation was mentioned above. The other important new technology is the digital platform, used by ride-hailing businesses to match demand to supply, as exemplified by Uber. Much is known about the impact of ride-hailing on traffic in US cities because the authorities are able to require provision of data as a condition of the companies’ operating licence. This has prompted the companies to volunteer data provision to help cities address urban transportation needs.

In contrast to ride-hailing, the providers of digital navigation are secretive. Little is known about the algorithms that calculate routes in the light of prevailing congestion, and how the guidance to users affects traffic flows generally. It is noteworthy that the DfT has recently revised its road traffic statistics to generate an increase of 26% of motor vehicle traffic on minor roads over the past ten years[8]. It is likely that this has been due in large part to use of digital navigation that makes minor roads usable to those without local knowledge.

The lack of appreciation of the impact of digital navigation is remarkable, given its likely influence on the functioning of the road network. The DfT’s Road Investment Strategy 2: 2020-2025 makes no mention of the use of satnav (although there is an illustration of a device on p38). Highways England created a ‘high-tech corridor’ on the A2/M2 in Kent to trial digital communications between roadside infrastructure and vehicles, to improve journey time reliability, yet this appears to pay no regard to the general use of digital navigation. The prospects for investment in such a publicly funded guidance system look poor, given the benefits provided without charge by the private sector providers of digital navigation.

A better approach to taking advantage of digital navigation to improve the operation of the road network would be through regulation. There is in fact legislation in place, but never used, to licence providers of dynamic route guidance. Licence conditions could include the provision of traffic information to road authorities and the avoidance of use of unsuitable roads.[9] Guidance to users is provided without direct charge, hence accommodating such licence conditions should not affect the business models of the providers, which depend either on selling direction-finding to retailers’ websites or mapping services to vehicle manufacturers.

Conclusion

The standard DfT approach to appraisal is no longer suited to decisions on investments in a mature transport system. The standard methodology has become vastly elaborate and misses the point that the benefit of investment is better access, which is seen as changes to the built environment. Decision-making needs to involve planners and developers as well as transport authorities. The National Infrastructure Commission has developed alternative approaches that better reflect the reality. An independent review of appraisal methodology would now be desirable.

Digital navigation seems to be having a significant impact on traffic flows. While the consequences are as yet little understood, it appears likely that the modelled benefits of road investment would be overstated if local traffic rerouting is disregarded. It would be desirable for the DfT to commission a study of the impact of digital navigation on the road network.

January 2021


[1] Metz, D. (2021) Time constraints and travel behaviour. Transportation Planning and Technology,44 (published online).

[2] Competition Commission, The supply of groceries in the UK market investigation, 2008. Fig 3.9

[3] National Infrastructure Commission, National Infrastructure Assessment, 2018.

[4] Metz, D. (2021) Economic benefits of road widening: discrepancy between outturn and forecast. Transportation Research Part A, (forthcoming).

[5] Department for Transport. Full Business Case: High Speed 2 Phase One. 2020.

[6] National Infrastructure Commission. Rail Needs Assessment for the Midlands and the North: Final Report. 2020.

[7] https://nic.org.uk/studies-reports/national-infrastructure-assessment/uplift-tool/

[8] Department for Transport. Benchmarking Minor Road Traffic Flows for Great Britain, 2018 and 2019: Methodology Report. 2020.

[9] Road Traffic (Driver Licensing and Information Systems) Act 1989.

Recent revisions to the road traffic statistics appear to show that there has been a substantial growth of motor vehicle traffic on GB minor roads in recent years, from 108 to 136 billion vehicle miles between 2010 and 2019, an increase of 26%. Traffic on major roads rose from 197 to 221 bvm over the same period, an increase of 12%.  (DfT Road Traffic Statistics TRA0102).

Road traffic statistics are based on a combination of automatic and manual traffic counts. Major roads are well covered in that traffic in all links is counted on typical days, although not every link in every year. Given the vast number of minor roads, however, it is only possible to count traffic at a representative sample of locations every year, and the observed growth is applied to minor road traffic overall. Estimates from a fixed sample may drift over time such that the sample becomes less representative of the changing minor road network. To account for any errors incurred in the fixed sample, the sample is revised through a benchmarking exercise every decade, involving a much larger sample of locations.

The most recent minor roads benchmarking exercise was published in 2020, based on 10,000 representative locations. Overall, the benchmark adjustment for 2010-2019 was 1.19, which is the factor to be applied to 2019 data from the original sample to bring this to the observed traffic level. Data for minor roads traffic for intermediate years are adjusted pro rata, to avoid a step change in the reported traffic data. There is significant regional variation in the adjustment factor, from 1.35 for Yorkshire to 1.09 for East of England, with London at 1.32. For B roads the factor is 1.25, for C roads 1.17; while for urban roads, 1.22, and for rural roads, 1.15. Applying the regional weightings yields an increase in traffic on minor roads of 26%, as noted above, whereas the increase based on the original sample would have been 6%.

The previous benchmarking exercise published in 2009 found a smaller overall adjustment factor of 0.95, with a regional range of 0.81 to 1.08.

The substantially greater adjustment required following the recent benchmarking, compared with the earlier exercise, suggests that there has been a real change in use of minor roads, beyond errors arising from drift in the sample. Importantly, had the increase in minor road use been spread evenly across the national road network, the traffic estimation based on the sample would have been close to that from the benchmark exercise. Hence the major difference between sample and benchmark indicates considerable heterogeneity of minor road traffic growth. Moreover, the fact that the sample failed to detect the traffic growth suggests either that the process for establishing the sample was deficient in some way, or that significant changes occurred in use of minor roads over a decade.

DfT statisticians have created a revised minor roads representative sample (4,400 locations) from the latest benchmark data, which will be used for the coming decade. It would be desirable to have comparative analysis of the previous and the new samples, to gain insight into what has been happening on the minor road network. Regrettably, the statisticians only report findings, and do not attempt to explain them, which leaves uncertainty as to the nature and cause of the reported changes to traffic volumes. The representative nature of the new sample must be questionable if the reasons for the failure of the previous sample to reflect reality are not understood and addressed.

Transport for London has recognised this uncertainty. The recent Travel in London Report 13 discusses the implications of the minor roads traffic correction (p92). The revisions mean that, for 2018, the DfT estimate of vehicle kilometres was 20% higher than previously reported last year (and included in Travel in London Report 12). The previous estimate suggested a fall of 1.8% in vehicle kilometres in London between 2009 and 2018, whereas the revised series now suggests an increase of 17.9% over the same time period, this change wholly arising from revisions to the minor road estimates. TfL notes that it is currently working through how the DfT have made this assessment, and also what this could mean for London data sets. For the moment, TfL is relying on its own traffic monitoring data, although it does not report traffic on minor roads separately.

The National Travel Survey could provide a cross-check on the traffic data. Average distance travelled by car/van driver decreased from 3388 miles per year in 2010 to 3198 miles in 2019, a decline of 5.6% (NTS0303). The GB population grew from 60.95m in 2010 to 64.90m in 2019, an increase of 6.5%. The net increase in car use of about one percent is inconsistent with the new road traffic statistics which show an increase in traffic for all roads of 17% over the same period. The NTS employs a fresh sample of respondents each year, so sample drift should not be a concern. However, it is not clear that the travel diary technique would pick up rerouting to minor roads, given that respondents are asked to provide their own estimates of distance travelled for each trip.

Possible causes of increase in traffic on minor roads

One factor contributing to the growth of traffic on minor roads is the increase in van traffic, including that arising from the growth of online shopping with home deliveries. The number of vans (light commercial vehicles) registered in Britain increased by 28% between 2010 and 2019. Total van traffic increased by 34% over this period, with an increase of 49% on urban minor roads compared with 10% on urban ‘A’ roads, although ‘delivery/collection of goods’ was less important in respect of journey purpose than ‘carrying equipment, tools or materials’. However, in 2019 van traffic amounted to 15% of traffic on urban minor roads, and 19% on rural minor roads, cars being responsible for 82% and 78% of traffic respectively. So, the growth of van traffic on minor roads is responsible for only part of the overall traffic growth on these roads.

Another possible explanation of the apparent large growth of traffic on minor roads is the widespread use of digital navigation (satnav) that offers routes that take account of traffic conditions and estimated journey times. Such devices make possible the general use of minor roads that previously were largely confined to those with local knowledge. This is likely to occur when major roads are congested and represents an effective increase in the capacity of the road network, so generating additional traffic – the converse of the ‘disappearance’ of traffic when carriageway is reduced. Increased use of minor roads is problematic when policy is concerned to decarbonise the transport system and to promote active travel, which these roads facilitate.

The possible role of digital navigation might be investigated by an analysis of the correlation of the upward adjustment factor for each minor road sample location with traffic volumes on nearby major roads – to test the hypothesis that there would be more use of minor roads in areas where major roads were most congested. If so, this factor should be taken into account when setting up the new minor roads sample for the coming decade.

The use of digital navigation has been growing and may continue to grow in the future. A better understanding of the phenomenon would be important for forecasting road traffic growth by means of the National Transport Model and models at regional level and below.

A further possible cause of the changed distribution of traffic on minor roads arises from intentional interventions aimed at reducing such traffic. It has long been the practice to discourage ‘rat running’ on urban minor roads by means of suitable physical control measures, as are used in low-traffic neighbourhoods (LTN). Such measures would reduce traffic in certain locations while possibly increasing it in others through diversion. Some locations in the minor roads sample may be so affected. If LTNs and similar measures increase over time, the sample may become increasingly unrepresentative, a factor that should be taken into account in setting up the new sample. However, the net effect of intentional interventions would be to reduce traffic overall, so this cannot account for the reported growth of traffic on minor roads.

The growth of minor road use by through traffic apparently facilitated by digital navigation would strengthen the case for implementing LTN measures. Alternatively, or additionally, the providers of digital navigation might be encouraged to omit routes that direct through traffic along minor roads.

More generally, the impact of digital navigation on the functioning of the whole road network seems likely to be significant and therefore worthy of investigation.

The above considerations prompt a number of questions:

  1. How reliable are the statistics for motor vehicle use of minor roads, given the apparent sensitivity to the sampling of locations?
  2. How reliable are the NTS findings for car use?
  3. What information is available on the likely causes of the increase of traffic on minor roads?
  4. What is known of the impact of digital navigation on the road network?
  5. What are the implications of digital navigation for transport and traffic modelling?

Summary

The reported increase in motor vehicle traffic on minor roads over the past ten years is substantial and locationally heterogenous, for reasons that are unclear. This lack of understanding raises methodological questions about the sampling of minor roads. The reported increase in traffic is not consistent with the findings of the National Travel Survey, as well as being of concern to Transport for London. While interventions to reduce traffic on urban minor roads may increase the heterogeneity of the sample, they would not increase the volume of traffic. Hence this increase is most likely due to the growing use of digital navigation devices that allow minor roads to be used by those without local knowledge. This has implication for transport modelling as well as for policies to decarbonise the transport system and encourage active travel.

This blog post is the text of an article published in Local Transport Today 19 March 2021