84178


   THE CONCEPT OF AFFORDABILITY OF URBAN PUBLIC TRANSPORT
            SERVICES FOR LOW-INCOME PASSENGERS

                 Slobodan Mitric and Robin Carruthers, February, 2005

The concept of affordability is widely used when discussing urban public transport fares
in the presence of low incomes and poverty, but it is not universally accepted as being
meaningful.

Fare is the dominant feature of public transport services for low- income travelers.
Likewise, it is the key policy variable for both the operating companies and for the
government, local and even national. As passenger incomes increase, concern over fares
starts to be displaced by the concern for quality. Traveler reactions to fare increases or
decreases are of major interest to operators of transport services, since they affect
patronage therefore revenue. This aspect is captured by a standard economic notion of
price elasticity, with an established definition, measurement methods, and interpretation.

When a fare increase leads to a loss of passengers at the low end of the income scale, the
operator is concerned because of the loss of revenue, but not more than that.
Understanding the concept and having some knowledge of the values of price elasticity
in this context is necessary for the operator to make informed decisions about setting
tariffs (sufficiency includes a requirement to know also about costs).

Affordability refers to the same context, but the concern is different and from a different
point of view. It is a concern of the society for its least privileged members. When a fare
increase could lead to a loss of poor passengers, and the alternatives are few and inferior,
access to jobs and services may be reduced, and social interaction may be affected. This
chain of events may lead to social and economic exclusion of those at the bottom of the
income ladder. This chain of events needs to be taken into account when tariff changes
are being considered. So concerns about affordability are clearly nested in the larger
subject of poverty impacts of transport policy.

The debate about the meaning and usefulness of the concept of fare affordability can be
presented by stating the polar positions. Given that we are concerned about the social
impacts of the policies that we recommend, the polar minimalist position is not to
consider only the impacts on the revenues of the transport operator through the operations
of price elasticity. It is rather to treat affordability as an important measure of transport
system performance.

An affordability index could be defined as the fare expenditure made by a household as a
percentage of its income. The index would be computed for various income groups and
the results scrutinized with an open mind as to whether, using the evidence of the
proportion on income spent on fares, they are reasonable, high or onerous. The
knowledge of affordability indices for the same city at some other point in time, or for
other cities in the same country or elsewhere, may provide a useful basis for comparison.
Such comparisons can only be made with much caution, since the ceteris paribus
assumption cannot be made.

At the other end of the range, the approach could be to specify affordability norms to
guide policy action. A standard and often cited statement of this approach comes from
Alan Armstrong-Wright and Sebastien Thiriez, "Bus Services - Reducing Costs, Raising
Standards", World Bank Technical Paper No. 68, The World Bank, Washington, DC, July
1987:
        “Travel expenditure probably is perceived, particularly by low-
        income groups, as the most important criterion in their choice
        of mode and may lead many to choose to walk. The extent to
        which a bus service is affordable is dependent on the income
        level of the users. In developing countries, a reasonable level of
        household expenditure on bus travel should not exceed 10
        percent of household income…… In industrial countries ,
        households without cars may spend in the region of 3-5 percent
        of their incomes on commuting. In developing countries, at the
        other extreme, studies have found that certain very low- income
        groups may spend in excess of 30 percent of their income on
        travel (Nairobi, Sao Paulo)., while levels of expenditures in the
        region of 15 percent are not uncommon (e.g., Kingston,
        Jamaica; Calcutta, etc.).”

Whether or not the authors meant for the 10% reasonable ceiling to become a norm, it has
been treated in this manner by many professionals and by some local governments. In
many cases it is not necessarily this particular index value (10%) or any index value, but
rather the idea that there is some limit to what is an affordable fare. In these cases the
norm is not expressed in terms of an index value, but as some absolute level of fare
produced by the political process or chosen arbitrarily by someone in the decision making
process. In a few cases the norm is applied as the level beyond which subsidies are
applied, as with the vale-transporte program in Brazil. In this case when fare
expenditures exceed a given percentage of the disposable income of an employee
(typically six per cent) , the subsidy is applied.

This normative approach is not limited to urban public transport. The same concept
underlies the consumer basket commonly used to track price levels and household
subsidies used in Russia and China, and the entire basic needs approach to poverty
calculations.


While the minimalist approach to use of an affordability index has limited use in
informing policy advice, the many difficulties of the normative approach in urban
transport are given in Cities on the Move: a World Bank Urban Transport Strategy
Review, The World Bank, Washington, DC, 2002. In summary:
There are wide variations in proportion of income used for public transport services by
low- income households, reflecting different tradeoffs in the choice of residential location
and employment.
Fare may be the most important aspect of public transport services for the low- income
travelers, but this is not to the exclusion of other service parameters. Low- income people
are known to pay This also contributes to variations in modal choices by low- income
travelers and in related expenditures
These variations imply that the concept of “affordable fare” cast as a ceiling and used for
fare regulation has no basis and should be discarded. Instead, the fare setting process
should include demand forecasting coupled with subsidy forecasting.

Cities on the Move: a World Bank Urban Transport Strategy Review, The World Bank,
Washington, DC, 2002

A reasonable approach is to continue using affordability indices as valuable indicators of
transport system performance, useful in identifying problems that need to be looked at in
more detail. Lest this be thought too modest, it should be remembered that measures of
affordability may have some explanatory power but no predictive power, in contrast to
elasticity. Indeed, when the impending decision by a local government concerns the level
of fares to be inserted in a concession bid document, or even when there is a fare increase
being considered, it is well advised to take a cautious and much more analytic approach.

The affordability index was defined above in terms of actual spending for public
transport services and actual incomes (or total expenditures) by households. 1 Whether to
use household expenditure or income as the denominator in the Index depends on what
the Index is being used for (see later for a summary of the reasons for using income
rather than expenditure)

The empirical approach based on expenditure requires a sample drawn from a target
population, obtained through travel and/or expenditure surveys and generalized over the
population. The results are distributions that reflect actual travel experiences, including a
full range of coping strategies at the low end of the income scale. One drawback is that
surveys are costly and take time to do. The explanatory power of this index alone is low,
due to the multivariate nature and complexity of the underlying behaviors. Problems
therefore may arise in interpreting the results. For example, a low value of the Index in
the lowest income deciles may mean that transport services are provided at such
commendable efficiency that prices are satisfactory from low- income point of view. Or it
could mean that fares are so onerous that low- income travelers walk or take some other
mode. The approach to follow is to analyze affordability data together with other
measures of household economics and the transport system performance, including
further interviews, to get a better sense of what is happening.




1
 The question whether to use income or expenditures in the denominator of the affordability index is not
dealt with in detail here.
Problems of interpretation limit sharply the use of empirical affordability indices for out-
of-city and out-of-country comparisons, short of some form of category analysis to
eliminate the impact of essential local features such as the urban structure, transport
alternatives available, the regulatory arrangements, cost recovery policy, and
price/subsidy aspects for other essential goods and services.

An alternative to using data from travel and expenditure surveys to compile affordability
indices is to “synthesized” the data on fares and number of public transport trips based on
assumptions about household type and travel scenarios. A common scenario is to
postulate a 4- member household, with one or both adults employed and one child making
a school- related trip beyond walking distance. Fare expenditures are estimated for trips at
average and extensive lengths, say 5 and 10 km. Income data are drawn from surveys, if
available, otherwise some milestone wage is used, such as minimal wage or average per
capita income.

The advantage of the “synthetic” approach is that the needed data can be collected
relatively easily and at low cost. Also the assumptions are explicit and few in number,
and the information value of the resulting Index can be quite high. This approach is well
suited for diagnostic studies, and even as a complement to empirical studies. Finally, the
synthetic affordability index is more portable than the empirical one, since the same
household and travel scenarios and the same income milestones (e.g. the minimum wage)
can be used elsewhere. Index values compiled from different cities are more easy to
compare than those based on actual expenditures.

The above discussion should not be seen as favoring the empirical or synthetic approach
to computing affordability indices. The synthetic indices should be calculated whethe r or
not the surveys will be or have been done.

Affordability Indices are computed with an idea that ultimately this may be linked to
improvements for low- income and poor people. Indeed, in the urban transport planning
and regulatory context, especially in large cities with long travel distances for home to
work travel, the presence of onerous and/or unaffordable public transport services calls
for some form of corrective action.

In the most narrow frame, the corrective action focuses on PT fares, be this their
magnitude or structure (relation to distance and/or travel time, treatment of transfers).

A wider approach could involve adding at least the following dimensions:

   (1) organization and regulation on the supply side, affecting total operating costs of
       providing UPT services and the quality of service; since economic and financial
       requirements of transport companies have the largest bearing on fares (needed
       fare revenues) affordability can be increased considerably through non- fare
       action.
   (2) non-price features of UPT services, such as availability, travel speed, punctuality,
       reliability, comfort (including crowding, seat availability, climate control), safety
       and security. Various segments of the travel market (e.g. grouped by income and
       trip purpose) value the cash price and non-cash service parameters differently.
   (3) The subsidy policy, subsidy management and the government capacity to pay.
   (4) A full modal choice context in a given city, i.e. a sense of transport alternatives
       available in addition to public transport services. If UPT services are not
       affordable, how are these other alternatives and can one do something to improve
       them?
   (5) A full household income and consumption picture. Transport is only one of goods
       and services deemed essential by households. The available budget has to be
       divided among housing, food, energy, etc., with substitutions and trade-offs
       determined by household- and city-specific characteristics.


The work undertaken in the last year by the World Bank on affordability has been limited
to constructing an Affordability Index and estimating its values for twenty seven cities.
We are now calculating the values for more cities, and looking at the range of options for
making urban transport more affordable to those on low incomes, and looking at
experience in applying them to see what conditions are needed to make them work
efficiently. Two important lessons learned from many efforts to make urban transport
affordable is that subsidies generally have a very high “leakage” so that most of the
advantages do not go to those who really need them, and that unless a subsidy has a
reliable and sustainable funding mechanism, the negative impacts on the transport
operator can more than outweigh any benefits to poor passengers.

“Fares unaffordable to passengers should not be replaced by subsidies unaffordable to
the fare setting institution”.