Michael Baker - Dissertation - Equity in Transport Planning
About | Contents
Previous: Section V | This: Difficulties of the "Alternative Approach" | Next: Appendix - Problem of the Use of Average Figures
There are several areas of difficulty inherent in the Alternative Method put forward in Chapter 5. These are problems of its own, those which it has in common with cost-benefit analysis, those due to the transport input, and those which will occur if the worst of the deficiencies of the transport planning methodology are removed.
There are many possible groups, and types of costs and benefits, which will lead to many cost-benefit distribution surfaces being produced in each evaluation. Either all the surfaces could be presented, in which case the effect may be totally confusing, or a single combined surface could be produced, in which case assumptions on inter group utility comparisons must be made, or a small number of composite surfaces could be produced, which would entail a trade-off between confusion and assumptions.
The types of groups which can be identified include those by income, by car ownership, by tenure, etc. Costs and benefits can be separated into those accruing to users and non-users. They can be further divided. For example user benefits can be sub-divided by trip purpose, and non-user costs can be classified as being due to causes such as noise and visual intrusion.
If all the surfaces are presented to the decision maker, and there could be as many surfaces as the product of the number of groups and the number of cost and benefit types, he may not be able to comprehend the overall position. This is a similar problem to that of comprehending a Planning Balance Sheet (for example see Lichfield (1966)).
On the other hand, if the surfaces are combined, an assumption will have to be made regarding the weighting of the costs and benefits. The simplest method would be to use the same weighting for all costs and benefits, or in other words to use no weighting. This however still makes an assumption that all groups attach the same utilities to monetary values for all cost and benefit types.
Enumeration. There are still the questions; has everything which is relevant been considered? and how are intangibles such as 'visual intrusion', or 'severance' to be measured? However, if, at the same time, other evaluations are being carried out, and this modified cost-benefit analysis is no longer considered 'comprehensive', these two questions are not so important.
Evaluation. As with cost-benefit analysis there are several difficulties in evaluation. The market does not operate perfectly, so prices will not necessarily reflect social costs or values. Intangibles, because they are not traded have no market value, so they have to be valued by other methods which will probably require the valuer to make 'value judgments'. There are still the difficulties of time valuation.
There is also an additional problem with time valuation and other values. Should an average value be used or not? An average value could be justified on the grounds that it will ensure that benefits (for instance time saving) accruing to the higher income groups would not be over valued compared to other groups' benefits. This would be an 'equity' valuation. On the other hand it could be argued that the values of each group should be used. In the case of quantities such as time savings these values would be the same as the behavioural values used for modelling purposes. Both behavioural values and equity values will tend to bias the information presented to the decision maker. If behavioural values are used the most noticeable quantities, and so those most liable to be changed, will tend to be costs and benefits to the rich, whereas if equity values are used those quantities which attract attention may not be valued at all by those they affect. This again is an inter group utility comparison problem.
Time Preference. If some of the costs or benefits for any group change significantly through time all costs and benefits will have to be discounted. This raises the question, what discount rate should be used, the current financial rate of return, or the social time preference rate of return? The groups which receive costs or benefits which have to be discounted may well have different time preferences. If a social time preference is to be used the question of whether an average or a group value should be used will have to be asked.
Just as the accuracy of the output from a transport model depends upon the accuracy of the input, so the accuracy of any cost-benefit analysis will depend upon the accuracy of its input. If sensitivity tests are carried out it may well be found that the size of costs and benefits is most sensitive to the socio-economic forecasts, and that the incidence of costs and benefits is most sensitive to the transport system. For example a doubling of growth in G.N. P. will probably increase everyones' costs and benefits by about the same factor provided it is not accompanied by a fundamental change in the distribution of incomes. On the other hand the costs and benefits due to the construction of a radial transport facility to the north of a city will fall mainly on those living to the north, whereas if the facility was on the south of the city most of the costs and benefits would fall on people living to the south.
Policies. If more policies are explicitly pursued in transport planning more evaluations than those of cost-benefit analysis will have to be made. One possible policy concerns equity, and the alternative method put forward in Chapter 5 is to deal with it. However, if, in the same iterative process of plan preparation, other policies are also considered, some way will have to be found by which the results of the evaluations for each policy can be combined.
Transport Model. There are several improvements which could be made in the transport modelling process. Account could be taken of the effect that provision has on demand. Consideration could be taken of alternative land use patterns, and the effect transport has on land use changes, instead of the single "trend" projection which has been used in the past. A more realistic assignment procedure which can split each set of interzonal trips between several routes could be used rather than the "all or nothing" assignment procedure which places them all on one route. Each of these improvements will increase the complexity of calculation, particularly of the incidence of costs and benefits. The first two will create great difficulties in calculations of costs and benefits, because there will no longer be an easily calculable "do nothing" situation upon which to base them.
One of the main problems is the combination of quantities which are valued differently by different groups. Should a single value be used? or should each groups' value be used to value their own quantities? and if so how can the values be determined? Just as welfare economics offers no satisfactory solution to the problems of making interpersonal utility comparisons, it would appear that no satisfactory answers can be given to these questions.
About | Contents
Previous: Section V | This: Difficulties of the "Alternative Approach" | Next: Appendix - Problem of the Use of Average Figures
Copyright © Michael Baker 1974,2005. All Rights Reserved.