Evidence on performance
The evidence relating to the impact of changes in fare studies tends
to take three forms. The first is provided by simulation models that aim
to simplify complex interactive situations and allow the researcher to
test incrementally the affects of changes to key drivers such as fare
and service levels. Such models use base data to seed the models (demand
data, fares data, service data and cost data) and apply established elasticities
(for fare, service, GDP etc) to estimate new levels of demand. The advantage
of these models is that they can isolate the impact of different fare
levels which in real life would be difficult because of changes to a host
of other variables that take place at the same time, e.g. service levels,
income, employment etc.
The estimation of fare elasticities themselves provides very useful information
and can be seen as another form of evidence on performance. The data used
for estimating the elasticities will take one of two forms: 1) actual
data which reveals people’s choices when faced with a real life
situation, often referred to as revealed preference (RP) data; or. 2)
what people state they would do if faced with a choice between different
scenarios, often referred to as stated preference (SP) data..
A third approach is to use case study data, preferably incorporating
before and after studies. This is the most data intensive of the three
approaches outlined and so the most expensive. It is also difficult to
isolate the affect of reducing fares from others changes, such as service
level changes.
Simulation Studies
LEK – Achieving Best Value for Public Support in the Bus Industry(2002)
Context
This study was commissioned by CfIT and attempted to assess the best
use of public support (concessionary fares and fuel rebate) within the
UK bus industry. To help in this assessment a bus model was constructed.
Standard values of times, quality values, diversion factors and elasticities
were used on the demand side. On the supply side bus operating costs were
estimated using the CIPFA formula (CIPFA, 1974) and augmented with costs
associated with different quality packages.
The model was provided with base data by operators which included details
about service levels, journey times, passengers and fare levels. All the
data provided was heavily anonymised,. Data for a number of different
types of routes was provided but in this section we only report the results
from the large radial route model which was based upon a busy major radial
route of approximately 12 kilometres in length in a large city. The services
operate along a single route with the following frequencies in each direction:
• 10 buses an hour: Monday to Friday, peak and interpeak and Saturdays:
• 2 to 4 buses an hour: Monday to Friday evenings, Saturdays early
and evenings, and Sundays.
Thus the service runs every 6 minutes during the main operating periods.
The services are paralleled for part of the route near to the city centre.
In short a very well served bus route.
Impacts on Demand
A number of scenarios were run using the model and these were based upon
four key attributes of bus services. The attributes and their levels are
outlined below and give a possible 189 combinations, however in this section
we only report the scenarios that examined different fare changes.
• 7 fare levels (+20%, +10% as now, -5%, -10% -20% and –50%);
• 3 frequency levels (as now, +20%, +50%);
• 3 journey time levels (as now, -5%, -10%); and,
• 3 quality combinations (as now, medium and high quality packages)
The model outputs were a mix of financial and quantitative data and represent
the change from the base case, which is presented in Table 14.
Table 14 Base Case Scenario (weekly data in £s)
Profits |
Bus Revenue |
Bus Cost |
Bus Pax
|
Car Pax
|
Bus Pax
Kms |
Car Pax
Kms |
Bus Veh
Kms |
£16,934 |
£40,374 |
£23,440 |
82,166 |
889,625 |
412,515 |
3,869,986 |
20,445 |
The models runs were only the fare level was changed are reported in
Table 15. The table reports the change of each indicator as compared to
the base case. A number of abbreviations are used, these are:
• CS (Consumer Surplus)
• Car Pax (car drivers/passengers)
• Bus Pax (bus passengers)
• Bus Pax kms (bus passenger kms)
• Car Pax kms (car driver/passengers kms)
Note that in these tables the change in net benefits is the sum of the
changes in consumer surplus, profit and any investment costs to the Local
Authority.
Table 15 Results of Fare Change Scenarios (weekly
data in £s)
No. |
Fare |
Profits |
CS |
Bus Pax
|
Car Pax
|
Bus Pax
Kms |
Car Pax
Kms |
12 |
+20% |
5049 |
-7828 |
-4,753 |
2,225 |
-27,347 |
13,262 |
21 |
+10% |
2630 |
-3977 |
-2,413 |
1,130 |
-13,906 |
6,742 |
30 |
-5% |
-1394 |
2033 |
1,232 |
-577 |
7,111 |
-3,446 |
39 |
-10% |
-2846 |
4103 |
2,484 |
-1,164 |
14,354 |
-6,957 |
48 |
-20% |
-5924 |
8345 |
5,046 |
-2,366 |
29,208 |
-14,155 |
57 |
-50% |
-16664 |
21956 |
13,206 |
-6,202 |
76,792 |
-37,200 |
The results illustrate that an increase in bus fare levels will, all
other things equal, reduce bus passengers and increase car passengers
& car travel. Although not shown in the table this will increase the
level of environmental externalities. A reduction in bus fare levels will
have the opposite impact with an increase in bus passengers and a reduction
in car travel. Financially, the operator will benefit from an increase
in the fare level and lose from a decrease. These results reflect the
fairly low fare elasticity of demand in the short run. It should be noted
that in the long run the elasticity of demand may well be closer to 1,
which would reduce the profitability connected to price rises and the
losses associated with price reductions.
Impacts on Supply
No impacts on supply were calculated.
Objective
|
Comment
|
|
Fare reductions are
likely to lead to reductions in car use will have contributed to
an efficiency improvement.
Fare increases are likely to lead to the opposite
impacts. |
|
Fare reductions are
likely to lead to a reduction in car use which will contribute to
a liveability improvement.
Fare increases are likely to lead to the opposite
impacts. |
|
Fare reductions are
likely to lead to a reduction in car use and so a reduction in environmental
impacts.
Fare increases are likely to lead to the opposite
impacts. |
|
There was no discernable
impact on equity and social inclusion from either a fares increase
or reduction. |
|
There was no discernable
impact on safety but it is likely that a reduction in fares will
reduce car use and reduce accident incidence and cost.
A fares increase is likely to lead to the opposite
impacts. |
|
Efficiency improvements
that are likely to occur from a fares reduction may help support
economic growth.
A fares increase is likely to lead to the opposite
effects. |
|
Reducing
fares is likely to lead to reduce bus revenues.
Increasing fares is
likely to lead to increases in bus revenues. |
Elasticity Studies
Dargay & Hanly – Bus Fare Elasticities (2002)
Context
This study estimated bus fare elasticities on annual data taken from
bus operators in Great Britain for years 1987 to 1996 on fares, bus demand
and a number of other variables that influence bus use, e.g. demographics,
GDP and motoring costs. The data was obtained from the STATS100A database
provided by the DETR and includes data returns from all GB bus operators
who are licensed for 19+ vehicles. Permission had to be obtained from
the bus companies first and it was sought from English operators with
fleets of 50 or more vehicles. Eventually, data was obtained from operators
who made up 87% of bus vehicle kilometres and 93% of passenger journeys
in England.
Impacts on Demand
A variety of models were estimated and the key results are presented
in Tables 17 and 18. In Table 17 the long run fare elasticities are greater
than in the short run, illustrating that passengers have more options
open to them for reacting to changes in fares (e.g. they can change jobs,
move house or purchase a car) as opposed to the short run. It is also
interesting to note that the fare elasticity increases as the fare does.
This is to be expected as, in monetary terms, a 10% increase in a high
fare will be greater than a 10% change for a low fare, making passengers
more sensitive to fare changes.
Table 17 Estimated Short-run and Long-run Elasticities Based
on Pooled Data for English Counties
|
Fare |
|
Short run |
Long Run |
Constant
Elasticity |
|
|
Constrained
Unconstrained* |
-0.33
-0.43 |
-0.68
-0.74 |
Variable
Elasticity |
|
|
Constrained
Min. Fare = 17p
Ave Fare = 56p
Max Fare = £1 |
-0.13
-0.41
-0.74 |
-0.26
-0.86
-1.53 |
Unconstrained*
Min. Fare = 17p
Max Fare = £1
Average GB |
-0.13
-0.44
-0.79
-0.33 |
-0.23
-0.75
-1.35
-0.62 |
*average of individual elasticties for all counties (...) elasticities
not significantly different from zero.
Source: Dargay and Hanly (2002)
Table 18 again illustrates that in the long run fare elasticities will
increase over time. It also demonstrates how fare elasticities can vary
between location. The Shire counties (rural counties – such as Oxfordshire)
have higher fare elasticities than in the Metropolitan areas (large urban
areas – such as Greater Manchester). This is to be expected as car
use in the metropolitan areas is less advantageous given congestion, parking
costs etc Table 18 also contains a number of other elasticities that provide
a useful demonstration of additional impacts on bus demand. Service elasticities
are positive implying that an increase in the service levels will increase
demand. The motoring costs cross elasticity tells us that an increase
in motoring costs will also increase the demand for bus.
Table 18 Estimated Short-run (SR) and Long-run (LR) Elasticities
Based on Pool Data for English Counties
|
Fare |
|
SR |
LR |
Metropolitan
areas
Shire counties |
-0.26
-0.49 |
-0.54
-0.66 |
Note: elasticities in parenthesis are not significantly different from
zero.
Source: Dargay and Hanly (2002)
Impacts on Supply
- no impacts on supply were estimated.
Contribution to Objectives
Objective
|
Comment
|
|
No evidence presented
on this. |
|
No evidence presented
on this. |
|
No evidence presented
on this. |
|
No evidence presented
on this. |
|
No evidence presented
on this. |
|
No evidence presented
on this. |
|
No
evidence presented on this. |
Case Study Evidence
Sheffield Case Study
In this section evidence is presented from two different studies that
concentrated on the bus fare freeze policy that was implemented in South
Yorkshire between 1974 and 1984. The policy was supported by the County
Council and the South Yorkshire Passenger Transport Executive (SYPTE)
and resulted in a real fares fall of 69%. The main justifications for
the low fares policy was (according to Hay, 1986) to,
• Slow, halt or even reverse the decline in public transport;
• Contribute to planning and environmental objectives by reducing
road traffic and supporting retail and service activities in city centres
(and selected suburban centres and small towns);
• To contribute to social objectives by increasing the mobility
of transport-disadvantaged groups, and by making a nonstigmatising income
transfer to low-income households.
Hay - 1986
Context
This study analysed changes in travel behaviour in Sheffield-Rotherham
(1971-1981) and Manchester-Salford (1976-1982) with special reference
to the effect of bus fare levels in real terms, which fell by around 70%
in South Yorkshire but remained constant in Greater Manchester. The study
made use of weekday travel records that had been collected as part of
land-use transportation studies in both South Yorkshire (in 1973) and
Greater Manchester (in 1977). A repeat of these surveys was carried out
in 1981/82 in both areas.
Impacts on Demand
Analysis of the data enabled comparisons of bus trip rates per day to
be made, which are outlined in Table 19.
Table 19 Global Comparisons of Bus Trip Rates per Day on a Standard
Population Structure
|
Sheffield-Rotherham |
Manchester-Salford |
|
1972 |
1981 |
1976 |
1982 |
All Trips |
0.681 |
0.710 |
0.598 |
0.494 |
Households:
without cars
with cars |
0.873
0.421 |
0.957
0.509 |
0.738
0.372 |
0.663
0.302 |
Work
Education
Shop
Social |
0.333
0.067
0.113
0.090 |
0.239
0.083
0.139
0.106 |
0.261
0.069
0.080
0.071 |
0.217
0.088
0.091
0.046 |
Source: Hay (1986)
In terms of overall trips it can be seen that the number of trips made
by people in Sheffield-Rotherham has increased by just over 4% and fallen
in Manchester-Salford by around 17%. Interestingly, the only category
of trips to fall during the 1972-81 time period in Sheffield-Rotherham
are those for work (by nearly a third). This suggest either a decline
in employment within the region or that despite decreasing real bus fares,
people were choosing to travel to work by another mode (mainly car). In
fact if one looks at the percentage of motorised trips made by bus during
that time the pattern is one of bus catering for fewer trips (in relative
terms) in all categories (Table 20).
Table 20: Motorised Trips Made by Bus and Estimated Global Figures
by Purpose
|
Sheffield-Rotherham |
Manchester-Salford |
|
1972 |
1981 |
1976 |
1982 |
%
of motorised trips by bus |
All
Work
Shop
Social |
50
56
45
44 |
43
47
37
33 |
54
58
46
44 |
41
45
28
26 |
Source: Hay (1986)
The main conclusions of the study were that the low-fare policy had resulted
in higher levels of bus use in Sheffield-Rotherham than might otherwise
have been expected and that such levels cannot all be explained by short
run elasticities (e.g. low fares over a long period of time had encouraged
a bus travel culture). However, there was no evidence to suggest that
the low fares policy had made any contribution to reducing traffic congestion
or assisting in city centre activities.
Impacts on Supply
- no impacts on supply were estimated.
Contribution to Objectives
Objective
|
Comment
|
|
No suggestion that
the low fares policy had reduced congestion and so improved efficiency.
|
|
No evidence presented
on this. |
|
No evidence presented
on this. |
|
No evidence presented
on this, however, a fares freeze policy should be expected to improve
equity and social inclusion. |
|
No evidence presented
on this. |
|
No evidence to suggest
that the low fares policy had made any contribution to reducing
traffic congestion or assisting in city centre activities. |
|
No
evidence was presented on this, however a fares freeze policy is
likely to have led to an increase in the amount of financial support
required from local government. |
Goodwin – 1983
Context
This study made use of the same data set utilised by Hay (1986) augmented
by additional postal questionnaire data and also face to face interviews.
The study placed much more emphasis upon assessing the social and travel
changes brought about by the low fares policy, and the key results are
outlined below.
Impacts on Demand
a) Effect On Other Methods of Transport
Car Ownership. This had grown in South Yorkshire, but at a lower rate
than in the adjoining county of West Yorkshire. A small number of households
found the high cost of motoring and low cost of bus a combination that
meant they would not be purchasing a car. However, the few number of people
who had actually forsaken their cars, had done so because of a change
in family circumstances, not because of the low fares policy.
Car Passenger Trips. The low fares policy had not affected the number
of car passenger trips. Lifts were mainly offered and accepted for reasons
of convenience and time saving.
Walking. Fares were at such a level that they were not the main consideration
when making a choice between walking and making a bus trip. More weight
was given to speed, security, weather, convenience of timing and knowledge
about the bus service.
b) Effect on Particular Groups
Employed. The purchase of a car appeared to be most influenced by the
journey to work. There has been a shift from bus use to car use for the
journey to work, as car ownership has continued to increase.
Shoppers. The use of bus for shopping had seen a large increase than
for any other journey purpose, with 23% of the weekday bus journeys in
1981 compared to 17% in 1972. The frequency of shopping trips by bus was
highest among non-car owners, the elderly and the unemployed. These groups
often see shopping as a recreational or social activity. The one type
of shopping where car still predominated was bulk shopping, e.g. weekly
groceries.
Unemployed People. Buses were not seen as the key means of looking for
work. They were however seen as important for facilitating other activities
such as shopping, visiting town, the library, recreational facilities
and friends. As such the low fares policy was seen to be helpful in assisting
the unemployed to maintain a ‘normal life’.
Retired and Elderly People. There was still a significant number of people
in this group who owned a car or had access to one. In some car owning
households more use was made of bus for certain journeys and there was
an appreciation that lower incomes and a reduction in savings might mean
that bus became more favoured over time.
Children. There had been an increase in bus travel by children despite
a decrease in the numbers of children born. The largest increase in trips
has been experienced during the morning and evening peaks during school
terms and also on weekends throughout the year.
In its conclusions, the study notes that the evidence of the impacts
associated with the low fares policy was consistent with long term as
opposed to short term fare elasticities. At the same time the policy appeared
to have had a much greater impact on the young than the middle-aged and
old sections of the population. This could be explained by the fact that
children and young people are much more influenced by conditions of the
time when forming habits and attitudes, compared to older people who experienced
different conditions as they grew up.
Impacts on Supply
- no impacts on supply were estimated.
Contribution to Objectives
Objective
|
Comment
|
|
No suggestion that
the low fares policy had reduced congestion during the peak periods
and so improved efficiency. |
|
No evidence presented
on this. |
|
No evidence presented
on this. |
|
Evidence that the
unemployed, the elderly and children were making considerably more
trips than in comparable areas. This suggests that equity and social
inclusion were improved. |
|
No evidence presented
on this. |
|
No evidence to suggest
that the low fares policy had made any contribution to reducing
traffic congestion during the peak. There was evidence to suggest
that it had helped to increase city centre activities particularly
for shopping purposes. |
|
No
evidence was presented on this, however a fares freeze policy is
likely to have led to an increase in the amount of financial support
required from local government. |
|