2
43 COMMENT on "The Measurement of Accessibility" CHRISTOPHER FOSTER Centre for Urban Economics, London School of Economics and Political Science I have read the paper by Dalvi and Martin with considerable interest. It is an extremely useful explanation of the problem of measuring accessibility, put clearly, and with some discussion of various difficulties involved. What it does in terms of analysing the consequences of different measures of accessi- bility when applied to Inner London is also interesting in a wider context. However, whilst I accept broadly what Dalvi and Martin are trying to do, we should ask what are the uses of a notion of accessibility. Underlying their use of it here is an idea that either relative accessibility or integral accessibility is a concept useful for planning purposes and with both behavioural and normative justification. If one thinks about the relative measure in relation to an individual, then it infers that it is a measure of utility and that a reduction in travel cost Ci] may be associated with an increase in utility, not merely monotonically but as possibly a fairly com- plete measure of it. When aggregation of individuals into a community takes place then the argument is again that any measure which increases (or reduces) their joint C-. is one which approximates to a measure of changes in rl their joint utility. There is nothing very special about this, as the limitation of generalised cost as a measure of utility is well known. Integral accessibility, however, does pose a further problem. It assumes that some general measure of generalised cost, irrespective of direction and of any independent measures of the relative utility to be attached to different journeys, can measure utility in relation to some proposed changes in the system. While I see that this is convenient and possibly of practical necessity, it is a grand simplification. Dalvi and Martin disaggregate this into different measures of accessibility related, by and large, to different journey purposes and they find, not surprisingly, that the measures of accessibility

Comment on “The measurement of accessibility”

Embed Size (px)

Citation preview

43

COMMENT

o n

"The Measurement of Accessibility"

C H R I S T O P H E R F O S T E R

Centre for Urban Economics, London School of Economics and Political Science

I have read the paper by Dalvi and Martin with considerable interest. It is an extremely useful explanation of the problem of measuring accessibility, put clearly, and with some discussion of various difficulties involved. What it does in terms of analysing the consequences of different measures of accessi- bility when applied to Inner London is also interesting in a wider context. However, whilst I accept broadly what Dalvi and Martin are trying to do, we should ask what are the uses of a notion of accessibility.

Underlying their use of it here is an idea that either relative accessibility or integral accessibility is a concep t useful for planning purposes and with both behavioural and normative justification. If one thinks about the relative measure in relation to an individual, then it infers that it is a measure of utility and that a reduction in travel cost Ci] may be associated with an increase in utility, not merely monotonically b u t as possibly a fairly com- plete measure of it. When aggregation of individuals into a communi ty takes place then the argument is again that any measure which increases (or reduces) their joint C-. is one which approximates to a measure of changes in rl their joint utility. There is nothing very special about this, as the limitation of generalised cost as a measure of utility is well known.

Integral accessibility, however, does pose a further problem. It assumes that some general measure of generalised cost, irrespective of direction and of any independent measures of the relative utility to be attached to different journeys, can measure utility in relation to some proposed changes in the system. While I see that this is convenient and possibly of practical necessity, it is a grand simplification. Dalvi and Martin disaggregate this into different measures of accessibility related, by and large, to different journey purposes and they find, not surprisingly, that the measures of accessibility

44

then vary very greatly. If one presses this point, one wonders whether further disaggregation would not lead to equally divergent sub-classifications of the measures they use. Moreover, there is an aggregation problem in that reflecting them in decisions that need to be taken logically, implies a weighting of the utilities to be attached to different journey purposes. Thus, while measures of accessibility can, I suppose, be defended as a necessity in practical transport planning, I do believe their theoretical underpinnings are insecure and I would not have been surprised to see Dalvi and Martin make more of this fact. However, they could argue that this matter is discussed elsewhere (though personally I do not know where). However, it does to my mind somewhat question the relevance of what they say in economic terms (indeed it is interesting that none of their references are to economic literature). I find that my doubts about this carry over into my interpreta- tion of their results when they vary the zoning conventions. How, again, does one really measure the importance of the differences they find, except in relation to some decision rules and the effects upon decisions of changing the conventions?