Recommendations for the design of optimal transport strategies

The following recommendations can be made for policy makers on the design of optimal transport strategies:
  1. Strategies should be based on combinations of measures, and should draw fully on the synergy between successful measures.
  2. The key elements of a successful strategy should be public transport measures and car user charges. In most cases, the public transport measures should include increased service levels and/or reductions in fares. However, the degree of such changes will clearly depend on the service and fare levels in the base case. Car user charges can be applied through road pricing or parking charges.
  3. There should generally be a distinction between peak and off-peak implementation of public transport and car user charge measures.
  4. Low cost road capacity improvements should generally be included in a successful strategy. However, it should be emphasised that such improvements should come from measures that genuinely improve traffic efficiency, given a fixed level of infrastructure. Such measures would typically include: traffic signal coordination and optimisation; telematics measures; and other traffic management measures. Low cost road capacity improvements should not be introduced if they have a negative effect on plans for city centre pedestrianisation, traffic calming in residential neighbourhoods, or enhancements to pedestrian mobility or safety.
  5. Large-scale public transport infrastructure projects (on top of those already planned) would typically not be part of an optimal strategy. However, medium-scale and small-scale infrastructure projects, such as guided bus or improvements to the public transport vehicle fleet, may be beneficial.
  6. In many situations, car user charges will generate sufficient revenue to finance the other elements of the optimal strategy, assuming that the initial capital costs can be raised and financed from revenue. It is important that the revenues are used to finance the optimal strategy, rather than to support other transport schemes which may detract from the strategy.
  7. In some circumstances, optimal policies (in terms of net social benefit) may include both car user charges and increased fares for public transport users (without a corresponding increase in service levels). The implications of this require careful consideration, since they suggest that transport policy can be used to subsidise other areas of public policy.

Recommendations for the involvement of the private sector

The following recommendations for policy-makers can be extracted from the discussion above:
  1. In many cities it will be possible, following the methodology outlined below, to identify strategies which are optimal, and whose costs over a 30 year evaluation period are met in full by payments from users. Care will needed to ensure that the pattern of charges on users is politically acceptable and legally feasible.
  2. Even in the circumstances in (1), city authorities may not be able to raise the finance required for initial investment in the strategy. In such situations, the private sector may be able to finance the strategy, and be repaid either directly by users or by the city authorities using payments by users. Examples include investment in new infrastructure and the introduction of road pricing schemes. In either case, the user charges should be consistent with the strategy: fares or charges imposed at higher than optimal levels to satisfy the private sector can significantly reduce the performance of the strategy. Furthermore, where the private sector requires a higher rate of return than the public sector discount rate, this may result in a more constrained, and less beneficial, strategy. The implications of this for involving the private sector need to be carefully assessed.
  3. Where the financial costs of the strategy exceed the revenues, it may still be acceptable for city authorities to finance them. The optimal strategies in FATIMA have been generated on the basis that the opportunity cost of using finance for them is fully justified. Where city authorities cannot raise the initial finance, it may be appropriate to involve the private sector, as in (2). However, the private sector will need to be reimbursed in part from future tax revenue, or from future revenue generated by reducing public expenditure on other sectors.
  4. Where public finance is limited, the optimisation procedures used in FATIMA can identify the appropriate modification to the strategy to achieve the optimal performance within the financial constraint. Such strategies will usually have smaller social benefits than those without such constraints.
  5. In the situation in (4), value capture may offer an opportunity for raising additional finance to help support the transport system. Such finance, which would not involve (later) repayments by transport users or the city authority, should be distinguished from the private finance arrangements in (2), which do involve such repayments. However, within the range of conditions tested, value capture appears not to offer the potential for significantly improving the overall strategy.
  6. Whether or not the private sector is involved in financing a strategy, there may be interest in private sector operation of the public transport service. Such involvement may possibly increase managerial efficiency which would enhance the performance of an overall strategy or alternatively it may lead to a reduction in net social welfare: evidence on the scale of these benefits or losses is unclear.
  7. Private sector operation has been implemented through deregulation, in which operators are free to determine service levels and fares, and through franchising, where the city authority specifies them. If a city authority decides that private operation is beneficial, it should use, where legally possible, a franchising model in which it specifies the objectives and the optimal service levels and fares.
  8. If a deregulation model is required (in order to comply with national law), private operators should not be given complete freedom to determine the operating conditions which meet their profitability target, even if the level of profitability is itself constrained as a result. There are typically a number of combinations (e.g. of fares and frequency) which achieve a given level of profitability, and not all will be equally effective in terms of public policy objectives.

Methodological recommendations

The key steps for strategy/policy formulation, in the order they should occur, are given below. Of these steps, the most problematic in terms of practical transport policy-making are steps 1 and 2, and they should be given special attention by policy-makers:
  1. identify the policy objective(s) clearly;
  2. where a set of policy objectives is identified, indicate what the appropriate trade-off is between them (assuming, usually correctly, that they are to some extent in conflict);
  3. identify the set of policy measures which are to be considered, and which can be expected to have a strategic impact (in particular, list those which meet the latter requirement);
  4. specify the range(s) within which the measures in (3) can be applied, and the factors which limit that range (financial, political, legislative etc.);
  5. specify any other overall constraints (e.g. financial) on the specification of optimal strategies;
  6. employ a transport model which enables the full range of measures in point 3 to be assessed against all the objectives (from point 1), taking into account of all the user responses (mode, time of day, destination, frequency, route) of strategic relevance, and all the supply interactions (congestion, overcrowding, queuing) of strategic relevance;
  7. follow an optimisation procedure to identify the optimum, taking into account constraints where appropriate;
  8. check that this optimum is feasible and acceptable and modify if necessary;
  9. decide whether it is appropriate to use private finance or private sector operation, or both. If so, decide how best to employ them within the context of a socio-economically optimum strategy.

Clearly the steps listed above implicitly assume that all policy impacts can be measured quantitatively and that transport models are available for predicting the levels of such impacts. Further research, as recommended below, is required for dealing with situations where such assumptions are unacceptable.

Recommendations for further research

The general approach taken in FATIMA was well accepted, particularly by the city authorities involved in the consultation procedure. However, the approach could be further refined in a number of directions, as given below:

  1. The objective functions have no measure of equity in them. Issues of equity arise in two different aspects, amongst others. Firstly, there is the issue of intrazonal equity, concerning the relative differences in benefits received by different socio-economic groups living in the same neighbourhood. Secondly, there is the issue of interzonal equity, which concerns the differences in benefit received by inhabitants of different neighbourhoods. It is recommended that both issues of equity be considered in future research into the construction of objective functions. Furthermore, it is clearly necessary that the transport models used in the optimisation process can output the values of equity indicators; research is likely to be needed for the further development of transport models so that they can accomplish this.
  2. The choice of measures to be used for forming optimal packages did not include land use measures. This was due to the lack of availability of appropriate models for representing combined land-use / transportation policies at the start of the project. However, as such models are currently being developed (including ones for several of our case study cities), the possibility should arise for including land use measures in future optimisation work.
  3. Research should be carried out into making a comprehensive assessment of the consequences of private sector operation of public transport, both under regulated and deregulated regimes. Such an assessment should not be limited to financial costs of operation, but should also take into account potential external effects of private operation such as: increased levels of unemployment as a result of "downsizing"; changes in employees’ wages; changes in safety levels and changes in other environmental benefits or costs.
  4. The policy measures considered by FATIMA were, for each city, relative to a "do-minimum" strategy specified by the respective city authority. Typically, the measures involved with such strategies are those to which an authority has made a full commitment. It follows that the recommendations given make an implicit critique of a city’s committed policies. However, it could be argued that such an approach might misrepresent a city’s intentions since it might be planning "likely" policies to which it has not made a full commitment. Such policies were not considered explicitly in the FATIMA optimisation work. It is thus recommended that, in future optimisation work, a method be devised which takes account of potential city policies to which there has not yet been full commitment.
  5. In general, the FATIMA methodology needs to be extended to take into account policy impacts that cannot be measured quantitatively. This will require a different approach, in which the optimisation process incorporates both fuzzy data and qualitative judgements.