Gap Analysis of Public Transport Needs

Measuring Spatial Distribution of Public Transport Needs and

Identifying Gaps in the Quality of Public Transport Provision

Graham Currie

A new approach to assessing the performance of public transport in • To assess the distribution and quality of public transport
meeting the needs of transport-disadvantaged people in the community service provided geographically, and
is described. It reviews previous and current research in this area and • To identify any needs gap between community needs and
describes how a new approach has been developed and applied with service provision.
Hobart, Australia, as a study area. The approach aims to identify geo-
graphical gaps in public transport provision where travel needs are high This paper is divided into four sections: a review of transport
but services are poor or nonexistent. It involves the use of readily avail- needs gap measurement, a new approach to needs gap measurement,
able socioeconomic statistics to quantify the distribution of needs in the key findings, and conclusions.
community with a single transport needs index. A public transport net-
work model measures the public transport accessibility to these groups
and a geographical information systems approach is used to display the TRANSPORT NEEDS GAP
distribution of the identified gaps between service and needs. The tech- MEASUREMENT: REVIEW
nique is highly relevant for smaller urban centers where the justifica-
tion of public transport subsidies is largely social-needs–based—that is, A literature review of quantitative approaches to measuring the geo-
where congestion and environmental benefits of transit are less critical. graphical distribution of people facing transport needs was under-
It is also relevant to recent work in transport accessibility audits and in taken by Currie and Wallis (2) and more recently by Nutley (3).
the assessment of community impacts of alternative transit development These reviews identified a range of approaches. Key components of
strategies. the methodologies applied include the following:

• A population measure, which quantifies needs relative to the

Catering for the needs of the transport disadvantaged remains an
number of people in a defined target population;
important objective and justification for the provision of public
• Socioeconomic measures, which consider the size and dis-
transport throughout the world. In smaller cities and in rural and
tribution of social groups considered to be in need of transport
regional settings, this objective is the principal rationale for the pro-
services;
vision of public transport subsidies. In these circumstances, other
• Measures of transport supply, which reflect the availability of
justifications for transit service provision, such as road congestion
transport so that needs can be assessed relative to supply; and
relief and the environmental impact reduction effects of public
• Measures of distance, cost, and accessibility to facilities (e.g.,
transport, are relatively minor.
work or shopping), which reflect the difficulties (or impedance) in
Despite the apparent importance of addressing community needs
gaining access to desired facilities and help to identify situations in
in public transport system design, comparatively little quantitative
which accessibility is poor.
work has been undertaken to ensure transit systems have been
designed in relation to these needs. In particular, the allocation of
In North America, the mobility gap approach has been applied to
services geographically often can be based more on historical and
Colorado and Montana (4, 5). This measures the difference in trip
political precedents than on a rational assessment of the distribution
rates between households with and without cars. Households with-
of potential users in the community.
out a car displaying low trip making are considered to be in need.
This paper presents a review of research in this area. It also
The approach identifies a quantum of transport service required to
explains the results of a research project undertaken in a small
fill this gap. A major benefit of this approach is its simplicity. How-
Australian city, Hobart, Tasmania (1), to further develop the
ever, it relies on the availability of the U.S. National Personal
methodologies. The aim of the approach taken is as follows:
Transportation Survey (6 ) to identify trip rates. Surveys of this kind
are not available in many countries, particularly for smaller settle-
• To measure the geographical distribution of transport needs in
ments. Furthermore, it does not consider needs in any great depth. For
the community,
example, transport needs for people in households with a car are not
considered. This is a substantive issue for families who live in single-
Institute of Transport Studies, Department of Civil Engineering, Building 60, car households. The technique is also limited in its consideration of
Monash University, Clayton, Victoria 3800, Australia.
the quality of existing transport services provided.
Transportation Research Record: Journal of the Transportation Research Board, One of the more refined approaches in relation to public transport
No. 1895, TRB, National Research Council, Washington, D.C., 2004, pp. 137–146. identified in the literature was that adopted by Searle (7), Martin &

137
138 Transportation Research Record 1895

Voorhees Associates (8), and Moseley (9), which is often termed the NEW APPROACH TO NEEDS
Lewes approach. In this case, the focus of analysis was rural settle- GAP MEASUREMENT
ments in East Sussex in the United Kingdom. The scale of transport
needs was identified by examining census records for the number of Reasons for Change
persons in social groups considered to be needy—for example, peo-
ple living in households with low car ownership. The quality of sup- The Department of Infrastructure Energy and Resources in Tasmania
ply was measured by examining public transport schedules to classify was seeking methods to assess the performance of public bus services
access to particular trip purposes, such as shopping, as impossible, relative to the distribution of travel needs in the community. In 2002,
poor, medium, or good. A needs gap was identified when a settle- management consultant Booz Allen and Hamilton was commissioned
ment had high concentrations of persons in needy groups and poor to use the methods developed by Currie and Wallis (2) to investigate
or no access to public transport. needs gap issues in Hobart, the capital city of the state of Tasmania,
This research aims to combine the assessment of both social needs which has a population of 192,000.
and the quality of public transport. There is a range of literature on Some specific issues regarding the Hobart needs gap project
either side of this equation, notably a range of techniques in measuring required further development of the methodology:
public transport accessibility (10, 11).
There have been more recent moves to examine the match be- • A local area assessment within a major urban area was required.
tween transport needs and the service supply in the United Kingdom This contrasted with more recent applications of the methodology,
as part of work by the social exclusion unit in relation to transport which identified needs gaps for settlements as a whole.
(12). These approaches have called for an accessibility audit to be • The supply side focus of the analysis was on public bus service.
undertaken as part of local transport plans. They compare accessi- Hence, a more detailed methodology was required to measure the
bility to employment, health care, and education facilities by all quantity and quality of the service provided, including time-of-day
forms of transport with due consideration given to the distribution and day-of-week analyses.
of socioeconomic groups that face transport difficulties, such as • It was hoped to apply geographical information systems (GISs)
people without a car, young people, older people, and people with to display the results of the analysis graphically to improve under-
disabilities. standing of the results.
Accessibility audits mirror earlier approaches proposed by Currie
and Wallis (2). Needs gap assessment first measured needs by the
following: Revised Approach

Overview of Approach
• Readily available census and social services information was
used to identify socioeconomic indicators that would measure In general, the proposed approach combined the needs indexation
the scale of transport needs faced by residents in a local area. These approach suggested by Currie and Wallis (2) and a more detailed
indicators were sourced from an analysis of the Adelaide House- assessment of transport supply measurement based on the Lewes
hold Travel Survey (13) by comparing socioeconomic groups approach. The latter involved the development of a fairly standard
demonstrating low trip-making behavior. public transport network model that measures the quantity and quality
• An accessibility measure [in this case, travel distance to the of public transport provision to a high level of detail. The TransCAD
central business district (CBD)] was adopted to identify locational transport modeling system (20) was adopted to undertake the network
disadvantage. modeling. The same system was adopted to display the results of the
• A single needs score was generated that combines the socio- analysis with GIS. Analysis was undertaken for Hobart’s 387 census
economic and accessibility indicator to give each location a score collector districts (CCDs). This is the smallest unit of analysis where
between 0 and 100 (with 100 being the location with the highest of census data could be collated for the needs analysis. Figure 1 presents
the combined indicator values of all the areas analyzed). the key steps in the analysis.
• The quantity of the supply of transport was measured by devel-
oping an indicator that includes the following:
– A public transport supply measure: the density of vehicle Network Supply Modeling
kilometers provided in the daytime interpeak per square kilo-
meter, A public transport network was constructed from an analysis of the
– A community transport (or paratransit) supply measure: the bus routes, stops, and timetables in Hobart. The network model
number of community transport vehicles supplied by area, and was similar to those used for transit and multimodal network plan-
– A taxi scheme usage indicator: the number of persons in the ning in most cities. In summary, the modeling process involved the
community who are members of the taxi subsidy scheme. following:

A single supply score was generated by combining the compo- • The location of facilities, shops, and so forth for 14 trip pur-
nent indicators and generating an index valued between 0 and 100, poses was defined (see Table 1).
with the highest score representing the highest level of supply. • The transport model measured the quality of travel by public
A needs gap was identified where the needs scores were high but transport (routes, access and egress times, frequencies, travel times,
supply scores were low. The preceding approach was applied in and fares) for five time periods including the following:
Adelaide (13). Other applications have been undertaken throughout – a.m. peak (07:00 to 08:59),
Australia and New Zealand (14–19). – Interpeak (09:00 to 14:59),
Currie 139

1.
1. Service
Service Level
Level Quantification
Quantification 2.
2. Activity
Activity Location
Location Review
Review

3.
3. Public
Public Transport
Transport
Network
Network Model
Model

4.
4. Area
Area Transport
Transport Need
Need Measurement
Measurement

- Socioeconomic Need Indicator Collection

-- Relative
Relative and
and Total
Total Needs
Needs Modelling
Modelling
-- Need
Need Score
Score Building
Building and
and Assessment
Assessment

5.
5. Public
Public Transport
Transport Gap
Gap Analysis
Analysis 6.
6. Reporting
Reporting

FIGURE 1 Revised needs gap analysis approach.

– Evening (18:00 to end of service), Where it was possible to walk directly to the nearest destination
– Saturday p.m. (12:00 to 18:00), and (without using a bus), this was considered to be preferable up to a dis-
– Sunday p.m. (12:00 to 18:00). tance of 800 m. Some origin zones were very large and hence walk-
• The analysis generated a matrix of generalized travel cost ing to and from buses was considered inappropriate given the lack of
results for 14 trip purposes by five time periods and for 387 travel routes within these areas. Distances above 400 m were identified as
zones (some 27,000 trip cost outputs). the threshold for feasible walk access to and from buses.
• For each time period, the transport model measured walk Figure 2 presents an example of the distribution of facilities for the
access time to bus stops, wait time, fare and travel time on buses, pharmacies (or drug stores) trip purpose. It also illustrates the gener-
and walk egress time. Table 2 presents the key assumptions for alized travel costs output from the model. Dots indicate the location
generalized cost modeling included in this analysis. of pharmacies in the greater Hobart region. Shading shows the
• Where more than one option was available for travel, the lowest- quality of travel by bus in categories of generalized travel cost (2003
cost path was chosen. Australian dollars), ranging from low cost (light shading) to high cost
(heavy shading). The darkest shade identifies where travel by bus (or
direct walk) was not possible. Results are for Sunday afternoon.
The output from this step in the analysis is a series of total gener-
TABLE 1 Trip Purposes Adopted: Bus Travel
Quality Modeling alized costs by area to each of the 14 trip purposes. These are then
summarized into categories such as trip not possible, very high cost,
CBD—Hobart CBD high cost, and medium cost.
Pools—public swimming pools
Shops—major groups of shops Area Transport Needs Measurement
Universities—major tertiary education facilities
The methodology for measuring needs involves assembling trans-
Sports—key recreational sporting facilities port needs indicators for a series of areas and defining a single needs
Pharmacy—chemists score for each area based on the relative indicator values. Transport
needs indicators used in the analysis are identified in Table 3.
Regional—larger regional shopping centers
Accessibility is the only indicator not readily available from gov-
Employers—larger-scale employers’ main location ernment statistics. Accessibility measures the natural convenience
Schools—major primary and secondary schools or difficulty a person is faced with when traveling from home to
basic services. It is a measure of locational disadvantage. The acces-
Hospitals—major clinics and hospital sites sibility measure used was the distance traveled to the CBD along
Food Stores—convenience shopping/local stores public roads (there are no rail services in Hobart). This was sourced
from a road network model for Hobart with the TransCAD modeling
Cinema—movie houses
system.
Child Care—site for a child-care center or crèche The formula for calculating needs scores is as follows:
Doctors—individual surgeries or clinics
needs score a = ( SI1a × WI1) + ( SI 2 a × WI 2) + L + ( SI 7a × WI 7)
140 Transportation Research Record 1895

TABLE 2 Generalized Cost Assumptions: Bus Travel Quality Modeling

Element of Travel Assumptions

Walk access/egress time • Measured between residential zone centroid and stop or from a stop to
the facility destination.
• Walking is made along the streets of Hobart rather than as the crow
flies.
• A walking speed of 4.32 kph is used. A weighting of 2 was applied to
walking time to model passenger perceptions of walk quality.
Fare • Based on an analysis of revenue and boarding data from the Hobart
ticketing system. Included an average fare for travel between zones
including higher fares for Hobart coaches routes and also
consideration of concession fares for particular passenger groups.
Wait time • Based on half the effective headway of routes operating between on
and off stops. Headways based on an analysis of bus schedules.
• A weighting of 2.0 was applied to wait time to model passenger
perceptions of waiting.
Value of time • Time was valued at $Aust 8.69/hour (or 14.48 cents per minute) based
on values used elsewhere in the transit planning industry.
Transfer time • A transfer penalty of 20 minutes was added to the time of those
transferring between bus routes.

SUN_PM
Cost to Phamacy
0 to 10
10 to 20
20 to 30
30 to 40
40 to 50
> 50
Other
0 4 8 12

Kilometers

FIGURE 2 Example trip purpose locations (pharmacies) and bus travel quality (total generalized cost) modeling results.
Currie 141

TABLE 3 Transport Needs Indicators and Weights Applied

Need Indicator Source Weight
Adults without cars Census 2001 and BAH analysis a 0.22
b
Accessibility BAH analysis 0.17
Persons aged over 60 years Census 2001 0.16
Persons on a disability pension Centrelink and BAH analysis c 0.14
Adults on a low income Census 2001 and BAH analysis d 0.11
Adults not in the labor force Census 1996 and BAH analysis e 0.10
Students Census 2001f 0.10
NOTES:
BAH = Booz Allen Hamilton.
a
Based on the number of cars per household and the number of persons aged over 16 (Census 2001).
b
Based on the distance to Hobart central business district (General Post Office) travelling on public roads
(Australian Bureau of Statistics 2001).
c
Based on the number of persons on a disability pension in a postcode grouping (Centrelink 2001). This was then
spread across CCDs based on number of persons aged over 60 (Census 2001).
d
Based on household income < $200 per week (Census 2001) and Hobart statistic of on average 2 persons aged
over 16 in each household (Census 2001).
e
Based on persons not in labor force in 1996 for the 1996 CCDs (Census 1996). This was spread across the
matching 2001 CCDs. When a one-to-one mapping did not exist, the data were spread in equal proportions—
bounded above by the number of people in the 2001 CCD (Census 2001). As the population of Hobart did not
grow in the period 1996–2001 (Australian Bureau of Statistics 2001), the total number of persons not in the labor
force in 2001 was assumed consistent with that estimated in 1996.
f
Based on persons enrolled in an educational institution, including primary and secondary school, university, and
technical and advanced further education.

where KEY FINDINGS

SI1 = standardized indicator 1 = adults without cars, Quality of Bus Access
SI2 = standardized indicator 2 = accessibility,
SI3 = standardized indicator 3 = persons aged over Figure 3 presents the average generalized cost of travel measured for
60 years, Hobart by time period and trip purpose. This indicates the following:
SI4 = standardized indicator 4 = persons on a disability
pension, • The cost and time of travel by bus vary considerably by trip
SI5 = standardized indicator 5 = adults with a low income, purpose as well as by time period.
SI6 = standardized indicator 6 = adults not in the labor • Travel for trip purposes with a greater number of facilities and
force, more localized facilities is easier. Food stores, for example, are con-
SI7 = standardized indicator 7 = students, sistently the easiest to get to. There is also little variation in the travel
WI1 to WI7 = weight for indicators I1 to I7 (see Table 3) [weights time and cost to get to food stores by time period. This suggests that
were sourced from an analysis of low-trip-making walk access is a key feature of access to these more localized facilities
behavior from the Adelaide Household Travel (instead of bus access).
Survey (13)], and • Trip purposes with consistently easier (lower cost) access are
a = area a (387 areas were used in the analysis). (in order):
A single needs score is derived from the indicators by first stan- – Food stores,
dardizing each value. This involves resetting the scores to a value – Schools,
between 0 and 100 based on the relationship of the score to the high- – Doctors,
est value in its series. Each standardized value is then weighted and – Pharmacies (drug stores), and
added together and a finalized needs index is generated. This is then – Child day care.
standardized to obtain needs scores between 0 and 100 for all areas • Trip purposes with the most difficult (longer time and higher
in the analysis. cost) access are (in order of difficulty):
– Hospitals,
– CBD, and
Needs Gap Analysis – University.
• With few exceptions, this pattern of access is replicated by time
The needs gap analysis is undertaken by comparing the needs scores period but with an overall decline in access quality for all trip pur-
with the network supply costs for each time period and trip purpose. poses on weekends.
To assist in understanding the considerable quantities of data this
analysis produced, all values were classified into categories: very Analysis also indicated that the bus network could not provide
low, low, average, high, and very high. Hence, an area with very service to all areas. This was because some zones were very large
high needs could be identified and its quality of supply measured. and had dispersed low-density development. It was impractical to
Supply measures included a categorization of trip not possible service these areas with a bus. Hence, up to 19% of zones in the
where the public transport service did not enable travel for most a.m. peak had no service. This increased to 35% on Sundays.
residents within a zone. Interestingly, the share of areas without bus access was more
142 Transportation Research Record 1895

120

Saturday P.M.
110
Average Weighted Travel Time Equivalent (min)

100

Sunday P.M.
90

Interpeak Evening
80

70

60

50

40
A.M. Peak

30

20
CBD Child_Care Doctors Cinemas Employers Food_Store Hospitals Pharmacy Regional Schools Sports Shops Pools University

Trip Purpose

FIGURE 3 Bus travel total generalized cost shown by equivalent travel time by time period and trip purpose.

variable by time period than it was by trip purpose. This demon- Figure 5 indicates that Sandy Bay M and Risdon–Risdon Vale A,
strates that the expansion and contraction of the bus network by both inner metropolitan areas, owe a large part of their total needs
time period is more significant than its connectivity to local and score to the low car ownership indicator, and Claremont and Kingston
regional destinations. owe a reasonably high proportion of their scores to high numbers of
people aged 60 or over.
In general, the other very high total needs score areas have high
Distribution of Transport Needs scores due to high values in all the remaining indicators. This is a
significant conclusion, particularly for the fringe areas consistently
Figure 4 presents the distribution of Hobart transport needs. This mentioned so far.
includes a blowup of the areas in inner Hobart. Needs are indicated It is also significant that fringe areas score highly not only because
in Figure 4 by the use of shading; darker shades are areas with higher of being less accessible, as may be expected, but also because they
needs and lighter shades indicate lower needs. have high concentrations of people with low car ownership, high
In general, the distribution of needs is patchy, suggesting a scat- levels of disability, and so forth. It can be concluded that, in fringe
tered distribution of high and low scores with no particular trend localities, people most vulnerable to transport disadvantage live in
toward inner versus outer areas being either high or low scores. areas where public transport is more likely to be limited relative to
There are some fringe areas with clear concentrations of very high inner city areas.
or high scores including the New Norfolk area and developed parts
of Bridgewater and Gagebrook.
Fringe areas with concentrations of high needs scores include Public Transport Needs Gap Analysis
Kingston; Sorell; and parts of Snug, Primrose Sands, and the South
Arm–Opossum Bay peninsula. Figure 6 shows the important needs gap identified in the analysis.
In general, undeveloped areas have very low needs scores. This The heavily shaded areas are those with identified high needs but
is to be expected given low total population levels. Figure 5 shows relatively poor quantity and quality of public transport (i.e., where
the size of component indicator scores for the highest-rated needs costs of using buses are very high). This analysis presents results
areas. The component indicators represent the true contribution to for the weekday a.m. peak. It also provides a summary of access to
the total needs. These component values have already been multi- all trip purposes. Figure 6 presents seven sets of needs gap shading
plied by the associated weights and standardized between 0 and 100 categories. These represent the cases of needs gap identified in
so that the CCD total needs score is simply the sum of these com- Table 4. Key conclusions from this analysis are as follows:
ponents. Suburbs have been split into several zones with the same
suburb name plus an alphabetic indicator at the end to provide each • There are no areas with the worst-case combination of needs
with a unique zone name. and service (i.e., very high needs and no service).
 Bridgewater/Gagebrook

Sorell

New Norfolk

Primrose
Hobart CBD Sands

0 4 8 12
Kingston Kilometers
South
Arm

Total Needs
Very Low Need
Low Need
Snug Mid Need
High Need
Very High Need

(a)

Claremont

Risdon/
Risdon Vale
Berriedale

Lenah
New Valley
Town

Hobart CBD

Sandy Bay
0 1 2 3
Kilometers

(b)

FIGURE 4 Spatial distribution of total transport needs score categories.

144 Transportation Research Record 1895

100
90
Scaled Total Need 80 Students
70 Unemployed
60
Low income
50
Disabled
40
30 Aged 60+

20 Accessibility
10 No car
0

rD
I

rE

rB
G
tL

B
K
M

lk

le
on
lk

on
on

r fo

te

te

te
y

Va
r fo
Ba

wa

wa

wa
st

ht
m

No

No

ng

ig
re

on
y

ge

ge

ge
r
nd

w
a

Ki

sd
w
id

id

id
Cl

Ne
Sa

Ne
Br

Br

Br
Ri
n–
do
is
Census Collection Districts

R
FIGURE 5 Component indicator share of total needs score (very high category areas).

FIGURE 6 Areas with significant needs gap ratings, a.m. peak (average of all trip purposes).
Currie 145

TABLE 4 Transport Needs Gap Categories

Rating Transport Need Rating Public Transport Quantity/Quality

Worst Case Very High No Service
Very High Very High Cost
High Need No Service
Mixed High/Very High Very High Cost
Medium Need No Service
High Need High Cost
Less Worst Case Mixed Very High/Medium Mixed Very High/Medium Cost

• The needs gap areas are predominantly on the urban fringe CONCLUSIONS
including the most severe needs gap score category.
• Several large rural zones feature in the medium needs–no ser- This paper presents a review of approaches that measure the geo-
vice group. This is because they are too large to be effectively ser- graphical distribution of transport needs and compares this with the
viced by a bus. This group is interesting in terms of needs gap distribution of public transport service quality. The analysis has
assessment because it is unlikely that conventional bus routes will developed the concepts of travel needs measurement identified by
ever be able to effectively service these areas. Nevertheless, the Currie and Wallis (2) by undertaking a more in-depth measurement
analysis identifies medium travel needs, which require some form of public transport service levels with a public transport network
of public transport. model in conjunction with a GIS to display results.
The results provide interesting insight into the distribution of
Some urban areas were identified in the analysis: travel needs in the Hobart community. In general, large numbers of
people known to have travel issues are located in places with rela-
• Risdon and Risdon Vale A, in the highest needs gap group of the tively poor public transport options—not a good combination. These
urban data (high needs–no service): This is a satellite community, areas lie mainly on the urban fringe.
with many people in the high transport needs group including Risdon The analysis provides a reliable and defendable basis for identi-
Gaol. The presence of the jail is interesting. It represents a facility with fying priorities to adjust public transport services or to locate social
large numbers of low-income residents with zero car ownership. This facilities to better meet travel needs in the community. It is designed
is the type of facility the needs measurement technique will highlight. to be easy to apply with usually readily available census and transport
However, classification of inmates as representing examples of high modeling tools.
transport needs may be questionable.
• Claremont L, in the mixed very high or high needs and very
high or high cost group: This area has significant development, ACKNOWLEDGMENTS
including retirement homes around a peninsula with poor bus service
Parts of this paper are sourced from a paper by Currie et al. (21). The
levels.
author thanks David Enright, Craig Hoey, and Darryn Paterson for
• Dynnyrne–Tolmans Hill B, also in the mixed very high or high
their permission to use this material in this report and for their input
needs and very high or high cost group. This zone straddles the
to the research. The author also thanks Tony Richardson and Rita
southern outlet highway in hilly terrain. There are pockets of resi-
Seethaler for their support and encouragement with this paper and
dential development along short cul-de-sac side roads, which would
Geoff Rose and John Clements for assistance in reviewing the
be very hard to service by bus. Walk distances to services in these
document.
areas are too far for reasonable access to bus stops.
• Geilston Bay A, in the medium needs–no service group: A
large zone on the shore of the River Derwent with essentially a rural REFERENCES
residential distribution. It includes relatively remote settlements like
Store Point without bus access. 1. Booz Allen Hamilton. Hobart Needs Gap Analysis. Final Report. Trans-
port Policy Division, Department of Infrastructure Energy and
The analysis revealed that there was little difference in the distri- Resources, Hobart, Tasmania, Australia, 2003.
2. Currie, G., and I. Wallis. Determining Priorities for Passenger Transport
bution of needs gap occurrences by time period. Although public
Funding: The Needs Assessment Approach. Proc., 17th Australasian
transport service levels fall on weekends and evenings, this does not Transport Research Forum, Canberra, Australia, 1992.
affect the locations that have a needs gap. Instead, it affects the 3. Nutley, S. Indicators of Transport and Accessibility Problems in Rural
severity of the needs gap. Hence, a severe needs gap area may have Australia. Journal of Transport Geography, Vol. 11, 2003, pp. 55–71.
better service in the peak, but it is still poor relative to other areas. 4. Colorado Department of Transportation. Transit Needs and Bene-
fits Study. LSC Transportation Consultants, Inc., Tahoe City, Calif.,
The distribution of areas with a needs gap did vary by trip pur- 1999.
pose. The main factor was the distribution of trip purpose land use 5. Montana Department of Transportation. Montana Rural Passenger
locations. For more localized facilities, such as access to a local Needs Study. LSC Transportation Consultants Inc., Tahoe City, Calif.,
store, it was less common to identify a needs gap. For trip purposes March 2001.
6. National Personal Transportation Survey. U.S. Department of Trans-
with single or a few sites (e.g., the CBD, universities, or hospitals), portation, Washington, D.C., 1995.
the travel task was more onerous and hence a needs gap was more 7. Searle, G. Value for Money from Rural Public Transport Subsidies:
likely to occur. A Summary of the Lewes Approach. In Transport Subsidy Policy
146 Transportation Research Record 1895

Journals (S. Glaister, ed.), U.K. Department of Transport, Lewes, 16. Travers Morgan. Local Community Transport Pilot Project Stage 1:
United Kingdom, 1987. Demand Estimation. Director General of Transport and Home and Com-
8. Martin & Voorhees Associates. Lewes Area Public Transport Study— munity Care Program, South Australia, Adelaide, 1990.
Final Report. ECLRT Division of the Department of Transport, in con- 17. Travers Morgan. South West Corridor Community Transport Study.
junction with East Sussex County Council Highways and Transportation Department of Transport, Western Australia, Perth, 1990.
Department, London, 1981. 18. Booz Allen Hamilton. Mornington Peninsula Public Transport Strat-
9. Moseley, M. J. Accessibility: The Rural Challenge. Methuen & Co., egy. Frankston City Council, Mornington Peninsula Shire Council, and
London, 1979. the Department of Infrastructure, Canberra, Australia, June 1997.
10. Kerrigan, N., and D. Bull. Measuring Accessibility: A Public Transport 19. Booz Allen Hamilton. Potential Initiatives for Improving Public Trans-
Accessibility Index. Environmental Issues Selected Proceedings of Sem- port in Regional Queensland. Queensland Transport, Brisbane, Australia,
inar B, Public and Transport Research and Computation Summer Oct. 1998.
Annual Meeting, London, Sept. 1992. 20. TransCAD. Caliper, Newton, Mass., 2003. http://www.caliper.com/
11. Jordan, C., and S. Nutley. Rural Accessibility and Public Transport in tcovu.htm. Accessed June 26, 2003.
Northern Ireland. Irish Geography, Vol. 26, No. 2, 1993, pp. 120–132. 21. Currie, G., D. Enright, C. Hoey, and D. Paterson. Quantitative
12. Making the Connections: Final Report on Transport and Social Exclu- Approaches to Needs Based Assessment of Public Transport Services—
sion. Social Exclusion Unit, Office of the Deputy Prime Minister, Lon- The Hobart Needs Gap Study. Australasian Transport Research Forum,
don, United Kingdom, Feb. 2003. Wellington, New Zealand, Oct. 2003.
13. Strategies to Overcome Transport Disadvantage. Department of Prime
Minister and Cabinet, Social Justice Research Program into Locational The views expressed in this paper are those of the author and do not necessarily
Disadvantage, Canberra, Australia, 1992. represent those of Monash University, the Department of Infrastructure Energy
14. Travers Morgan. Mobility Study. Report for the Urban Transit Author- and Resources, or Booz Allen Hamilton. Any errors are the responsibility of the
ity, New South Wales. Urban Transit Authority, Sydney, Australia, author.
1989.
15. Travers Morgan. Palmerston North Public Transport Strategy. Palmer- Publication of this paper sponsored by Transportation Planning Needs and
ston North City Council, Palmerston, New Zealand, 1989–1990. Requirements of Small and Medium-Sized Communities Committee.