A Study on Index of Journey Time Reliability to Evaluate Road
Projects in Rural Areas
Topic Area:
D5
Paper Number:
Authors:
By Masaru FUJII, Hironobu HASEGAWA, Mikiharu ARIMURA , Tohru TAMURA
Title:
A Study on Index of Journey Time Reliability to Evaluate Road Projects in Rural
Areas
Abstract: Reliable trip time has been important point of view in rural roads
compared with shortening trip time in urban areas because trip that derived by local
economical activity sometimes has limitation of time-space path.
The discussion regarding the level of time-reliability as rural road performance has
not received as much research attention. However, now, local economy assigned
important rules in the national land planning by impact of the falling birthrate and the
aging population.
The aims of this paper are to show the reliability of trip time is an important index in
rural area activity by the stated preference survey, and to propose the time-reliability as
the road performance index.
Key Words: road administration evaluation, journey time reliability, Willingness to Pay
A Study on Index of Journey Time Reliability to Evaluate Road
Projects in Rural Areas
Masaru FUJII
Postgraduate student
Division of Civil and Environmental Eng.
Muroran Institute of Technology
Address: 27-1 Mizumoto, Muroran, Hokkaido,
050-8585, JAPAN
Fax: +81-143-46-5289
E-mail: s1891005@mmm.muroran-it.ac.jp
Hironobu HASEGAWA
Postgraduate student
Division of Civil and Environmental Eng.
Muroran Institute of Technology
Address: 27-1 Mizumoto, Muroran, Hokkaido,
050-8585, JAPAN
Fax: +81-143-46-5289
E-mail: s1121071@mmm.muroran-it.ac.jp
Mikiharu ARIMURA
Docon Co., Ltd.
Address: Atsubetsu-chuo 1-5-4-1, Atsubetsu-ku, Sapporo, Hokkaido,
004-8585, JAPAN
Fax: +81-11- 801-1520
E-mail: ma1517@docon.jp
Tohru TAMURA
Professor & Dr. of Engineering
Dept. of Civil Engineering and Architecture
Muroran Institute of Technology
Address: 27-1 Mizumoto, Muroran, Hokkaido,
050-8585, JAPAN
Fax: +81-143- 46-5288
E-mail: tamura@mmm.muroran-it.ac.jp
1. Introduction
Since July 2005, the Highway Agency of the United Kingdom (HA) has been using
“journey time reliability” as an outcome index for evaluating road repair/construction
projects. Road administration aims to guarantee the quality of services to counter
congestion on urban trunk roads by assigning numeric values to certain variables.
HA measures the time required to pass through each of the 103 routes it manages in
15-minute intervals between 6:00 a.m. and 8:00 p.m. on weekdays. The data obtained
are then used to determine the time required for a trip through a certain section of each
route, and the index of journey time reliability is derived from this.
Japan and the United States also list the solution of traffic congestion as an index for
evaluating road administration. In Japan, however, achievement of the solution of
traffic congestion is not compulsory, and the level of guaranteeing results with numeric
values as in the United Kingdom has not yet been reached.
In a number of countries including Japan, discussions regarding traffic in farming
villages and other rural areas have centered on traveling speeds and road safety, and
journey time efficiency has not been discussed. This is because congestion frequently
occurs in urban areas while hardly ever occurring on rural roads. From the standpoint
of the activity-based approach, which is premised on the idea that “traffic is a demand
derived from activities,” activities in rural areas have time-space constraints.
Therefore, discussions regarding a given road’s journey time efficiency are also
regarded important.
This study thus has two purposes. The first is to reveal how important journey time
reliability (road performance) is for the activities of local people, since it shows the
maximum time required to travel through a certain section of a road in a rural area,
which plays an important part in national land planning. The second is to examine the
index of journey time reliability as an index to evaluate road administration, with a
focus on intercity traffic in rural areas.
2. Road functions required when constraints on activities exist
The subject of this study is Nakagawa Town, where Hokkaido-style agriculture
differing from that in other prefectures in Japan is practiced. For people in an
agricultural community like this, wide-area traveling to major local cities is a common
way to supplement production, living, exchange and other functions that cannot be
fulfilled in their own area. These people therefore efficiently take time away from
their work for living-related activities under constraints of both time and space.
Figure 1 presents the speed distribution in 12 hours during the daytime in the
non-snow (October) and snow (February) seasons at Bifuka on National Route 40,
which usually connects Nakagawa Town and Nayoro City in approximately 90 minutes.
The axis of length in the graph shows the generation probability at the speed. The
axis on side shows the regularized speed data.
The distribution was defined by the probability density function. From this it can
be seen that, in the non-snow season, the interval between the probabilities of low- and
high-speed traveling is short, and the traveling environment is relatively stable. The
traveling environment in the snow season, however, is greatly deviated from that of the
3
non-snow season in terms of journey time reliability since frequency of low-speed
traveling increases and the interval between speed probabilities is longer.
Probability
Probability
0.45
0.4
0.35
0.3
0.45
Non-snow season: 12-hour speed
distribution in October
n=372
VAR=5.8
MaxV/MinV=1.24
Snow season: 12-hour speed
distribution in February
n=325
VAR=59.4
MaxV/MinV=5.44
-2.00
-1.00
0.3
A large deviation in speed distribution occurred
0.25
0.25
-3.00
0.4
0.35
0.2
0.2
0.15
0.15
0.1
0.1
0.05
0.05
0
0.00
1.00
2.00
3.00
4.00 -7.00
-6.00
-5.00
-4.00
-3.00
-2.00
-1.00
Regularized speed data
0
0.00
1.00
2.00
3.00
Regularized speed data
Fig. 1 Speed distribution at Bifuka on National Route 40
This clearly reflects the importance of journey time reliability as a traffic service when
traveling for activities related to production, living, interests and recreation (Table 1).
Table 1 Road performance needs by activity type
Required road
performance
Purpose of transportation
Production Living-related Interests
Speed
○
△
△
Journey time reliability
◎
◎
○
Safety
△
○
◎
3. Relationship between travel limits of urban and rural areas and traffic services
Figure 2 presents a schematic diagram of the theory of population movement. As
young people leave rural areas, marginal productivity decreases, which in turn leads to
population movement from rural to urban areas to achieve higher marginal productivity.
As a result, the productivity of society as a whole decreases even further.
Point where production of society as a whole
(MP) reaches the maximum
Marginal
productivity
Marginal productivity curve of the
rural area MPa
Point where production of
society as a whole (MP')
reaches the maximum
(MP>MP')
Marginal productivity curve of
the urban area MPc
Marginal productivity
when the population of
the urban area is Nc1
Marginal productivity curve of the rural
area in the case in which young people
moved out
Marginal productivity
when the population of
the rural area is Na1
M arginal productivity decreases.
N3
Rural area a
Ｎ2
Ｎ1
Urban area c
Population
As a result of decreases in the number of
young people, more of the population of
the rural area moves to the urban area for
higher productivity.
Population moves from the
rural to urban area for higher
productivity.
Fig. 2 Marginal productivity and population movement
4
Compared with young people who can travel light, traveling costs are said to be
higher for the elderly, since they have more territorial connections and blood relations as
well as many intangible assets, such as comfort of living, in their places of residence.
It is thus assumed that the productivity of society as a whole will drop even further as
long as the current population movement and aging continue to progress in Japan’s rural
areas.
Then what type of social space structure is necessary to promote settlement of
population in rural areas and improve the productivity of society as a whole?
Figure 3 is a conceptual diagram of movements of market- and raw
material-oriented companies when attempting to maximize profits from the aspect of
total transportation cost, including supply and distribution costs. Orientation to the
right indicates profit maximization for market-oriented companies as the total
transportation cost is minimized, while orientation to the left indicates profit
maximization for raw material-oriented companies.
Transportation
cost
Point of equilibrium between the market- and raw material-oriented economies
Total transportation cost of a marketoriented economy
Total transportation cost of a raw materialoriented economy
Distribution cost
Supply cost
 
Point of raw material
 
t
In a market-oriented economy:
the supply cost at point t is αt,
the distribution cost at point t is (T-t)β,
the total transportation cost is αt + (T-t)β
Whenβ>α, the total traffic volume becomes the minimum at the point of
market.
χ
Point of market
Ｔ
In a raw material-oriented economy:
the supply cost at point t is αt,
the distribution cost at point t is (T-t)β,
the total transportation cost is αt + (T-t)β
Whenα>β, the total traffic volume becomes the minimum at the point of raw
material.
Fig. 3 Movements of market- and raw material-oriented companies
from the viewpoint of transportation cost
Since there are of course companies that place importance on both marketing and
raw materials, X indicates the equilibrium point where profit maximization for such
companies is achieved.
What will happen if these are replaced by rural and urban areas, and if it is assumed
that the total transportation cost consists of expenses for production and living-related
activities?
Although the values of transportation costs for production and living-related
activities vary between rural and urban areas, profit maximization of society as a whole
5
is achieved at the  point of traveling time shown in Fig. 4, and this is thought to
correspond with the time of 1 to 1.5 hours required in Japan’s national land planning
theory. It is considered necessary to maintain this level of traffic service in the future
to ensure sustainable regional development in rural areas where there are considerable
time and space constraints on production and living-related activities. However, it is
difficult to ensure transport services within the limit of 1 to 1.5 hours in all regions with
limited financial resources. It is thus necessary to at the very least ensure reliability to
the extent that there are no significant variations in the current traveling time required.
Point of profit maximization of society as a whole
Transportation
cost
Total transportation cost of a person from the rural
area (cost for living-related activities > cost for
production activities)
*Rural residents can receive medical treatment and do
their shopping with many constraints.
Total transportation cost of a person from the urban area
(cost for production activities > cost for living-related activities)
* Urban residents can receive medical treatment and do their
shopping with few constraints.
Transportation cost for
production activities
Transportation cost for
living-related activities
 
Rural area
t
Time
In the rural area:
the cost for living-related activities at point t is αt,
the cost for production activities at point t is (T-t)β,
the total transportation cost is αt + (T-t)β
Whenα>β, the total traffic volume becomes the minimum at
the central point of the rural area.
 
χ
Urban area
In the urban area:
t he cost for living-related activities at point t is αt,
the cost for production activities at point t is (T-t)β,
the total transportation cost is αt + (T-t)β
Whenβ>α, the total traffic volume becomes the minimum at the central point of
the urban area.
Fig. 4 Travel limits in urban and rural areas
Therefore, in this study, values for speed and reliability of traveling as seen from the
rural areas were measured using the stated preference survey.
4. Evaluation of journey time reliability in a rural area
4.1 Summary of the questionnaire survey
In this study, a questionnaire survey involving door-to-door interviews was conducted
regarding items that might affect the subjects’ willingness to pay, including age, gender,
family makeup, frequency of visits to the central city and income, as well as their values
of speed and reliability. Table 2 provides a summary of the survey.
Table 2 Summary of the questionnaire survey
Investigation period
Person for investigation
Questionnaire method
Extraction condition
Number of extractions
Number of collections
November 6～11, 2006
Nakagawa-cho resident（Number of homes）
Visit interview investigation（Catching type）
Occupation（Agriculture/Civil servant/Commerce/Another）
72 families
54families（Collection rate　75％）
6
4. 2 Questioning method
The questionnaire survey conducted in this study consisted of three evaluation items
– speed, reliability and willingness to pay, and was designed for respondents to select
from a road of the current improvement level or a reference road with a different level
of improvement.
The road of the level with different travel time and delay generation probability was
set the number of patterning, and whether it was compared roads where the delay of 60
minutes or less during winter occurred six times in ten times by 90 minutes, and which
preferable were answered for the travel time that became a standard.
For the question on willingness to pay, the Yes/No question method (respondents
were asked to express their willingness to pay for the evaluated amount in the form of
YES or NO) was adopted to simulate the purchase condition of actual goods, in which
purchasers decide whether or not to make the presented purchase, and to avoid bias
caused by improper presentation of interviewers. The form of payment adopted for
measuring benefits was the “tax-use method,” in which taxes paid to local governments
every year are used for road construction. Fig. 5 shows one part of the said
questionnaire.
The time required for traveling in winter is 90 minutes before the measure is taken,
and a delay of up to 60 minutes occurs six times out of ten. Please take these facts
into consideration when answering the questions below.
The decision to take this measure will be made according to your will. It will be
implemented immediately if you agree with it, but the timing of implementation will
not be decided if you disagree with it.
Q2-1. If there was a plan to improve the average time required for traveling in winter to
60 minutes with a delay of up to 60 minutes occurring three times out of ten, would you
agree or disagree with the plan?
1. I would agree.  Please answer Q2-2.
2. I would disagree.  Please proceed to the next page.
3.I don’t know.  Please state the reason below
Q2-2. How much would you be willing to pay for the plan?
Please circle “Yes” or “No” for questions 1) to 10).
The plan is to be conducted under the “tax use” method. However, you are supposed to
pay ¥200,000 in tax from which ¥50,000 is to be appropriated to social welfare, among
other budget items. If you pay any of the following burden charges, levels of services for
social welfare and hygiene will be reduced. Please take these facts into full consideration
when answering the questions below.
Payment
A nswers
1)
¥0
?
1. Yes
2. N o
2)
¥1,000
?
1. Yes
2. N o
3)
¥3,000
?
1. Yes
2. N o
4)
¥5,000
?
1. Yes
2. N o
5)
¥10,000
?
1. Yes
2. N o
6)
¥15,000
?
1. Yes
2. N o
7)
¥20,000
?
1. Yes
2. N o
8)
¥30,000
?
1. Yes
2. N o
9)
¥40,000
?
1. Yes
2. N o
10)
¥50,000
?
1. Yes
2. N o
Fig. 5 Example of questions
7
4.3 Respondent attributes
The heads of 54 households (75% of the potential subjects) responded to the
questionnaire. Due to the progression of aging, approximately 30% of the respondents
were 60 years of age or older, which is nearly identical to the age structure of Nakagawa
Town. Regarding occupation, approximately 75% were engaged in dairy and upland
farming and 25% were self-employed, public servants and other workers. Looking at
the family structure, approximately 60% of respondents were living with their children
or children’s families who will eventually become their successors (Fig. 6).
【Occupation】
【Age】
70years old or
more
20%
【Cohabitation family structure】
The unemployed
Civil servant
20～29years old
7%
4%
Commerce and
9%
30～39years old
7%
industry and
independent
enterprise
40～49years old
9%
19%
Consort＋Parents＋
Child 17%
Parents＋Child
4%
Parents（Parents
of contain it. ）6%
60～69years old
11%
Stock raising
42%
50～59years old
39%
Fig. 6
Consort
28%
Consort＋Child
Agriculture
38%
（Plowing a field）
33%
Consort＋Parents
7%
Respondent attributes
Respondents who visited Nayoro City at least once a month on average in winter
(November to March) accounted for approximately 94% of all purposes and 65% of the
purpose of receiving medical treatment. Nearly half of those who visited Nayoro for
medical treatment also shopped there.
Percentage distribution (%)
Percentage distribution
50.0
Total
Medical treatment
45.7
40.7
40.0
40
31.5
34.8
30.0
22.2
20.0
10.0
50
30
20
17.4
10
5.6
2.2
0.0
No visit
Once/month
Fig. 7
2 to 3 times/month
0
4 times or
more/month
Number of visits to Nayoro in winter
4.4 Evaluation model
In this study, the effect function was the function shown in Eq. (1) when a respondent
presented a road improvement level higher than the current level, and parameters were
estimated using the maximum likelihood method.
U j    x j1    x j 2    x j 3
(1)
Where,
Uj: effect function of the improvement level of road j,
8
xj1: willingness to pay for the improvement level of road j (yen/year)
xj2: time shortened by the improvement level of road j (min.)
xj3: improvement in journey time reliability by the improvement level of road j (min.: in winter)
 ,  ,  : parameters of xj1,2,3
Improvement in journey time reliability was obtained using the function shown in Eq.
(2), assuming that the probability other than that of the maximum lag time (60 minutes)
follows a uniform distribution within the range of 0 to 60 minutes.
xj 3  Tmax 60  Pmax 60  Tm  1  Pmax 60 
(2)
Where,
xj3: improvement in journey time (min.), Tmax 60 : maximum lag time (60 minutes),
Pmax 60 : probability of the maximum lag time,
Tm : average lag time within the range of 0 to 60 minutes.
The presumption value of θ that used the log likelihood function shown in the
expression (3) in the calculation of the maximum likelihood estimate and maximized
this was requested. Using the likelihood function of the logarithm shown in Eq. (3) ,
andθ that maximizes L was calculated.
Ns
L     it ln P (i | z t ,  )
(3)
t 1 iCt
4.5 Calculation results
Parameters were estimated for “activity purpose” and “occupation” attributes by
assigning variables determined by the stated preference survey to Eq. (1), and the values
of time shortened and improvement in journey time reliability were compared based on
these estimated parameters and differences in sensitivity against the “willingness to
pay” parameters, which indicate the weight of each explanatory variables. Table
3shows these results. The t values of estimated parameters were large and satisfied
sign conditions. Improvement in journey time reliability exceeded time shortened, and
the difference was estimated to be roughly 1.5 times.
Table 3 Estimation results
1)Willingness to pay
2)Time shortened
3).Improvement in journey time reliability
1)/2)
4)＝1)/2)÷12×4
5)=1)/3)
5)/4)
Hitting rate
（n=54）
Estimated parameter
1.433
-0.017
-0.033
-83.9
-28.0
-43.2
1.5
78.20%
t-value
13.11
-4.15
-3.98
－
－
－
－
－
Note: 1)/2) is the annual value, so it is converted to the winter (December to March) value in
4).
9
5. Journey time reliability in rural areas
5.1 Journey time reliability as an evaluation index
Since July 2005, the Highway Agency of the United Kingdom (HA) has been using
“journey time reliability” as an outcome index for evaluating road repair/construction
projects. Road administration aims to guarantee the quality of services to counter
congestion on urban trunk roads by assigning numeric values to certain variables (Fig.
8).
numbre of
Observation
Reference value of Travel time T1
（Travel time when it is not congestion）
※ It makes it based on figure of reference
document 2.
Distribution of travel
time of point-to-point
Of trip of ten times
Nine times : in this time.
Time that can be moved T0.9
Subordinate position 10%
of travel time
Mean value of trip T0.1
Travel time
18 minutes
Fig. 8
33 minutes
38 minutes
Conceptual diagram of journey time reliability
Specifically, HA measures journey times on all the 103 routes under its jurisdiction
every 15 minutes between 6:00 a.m. and 8:00 p.m. on weekdays. Using these data,
they determine journey times for certain sections of every route and derive the index of
journey time reliability according to the following set procedure: 1) Designate the
average journey time of congestion-free data groups collected in the morning and
evening as 1 . 2) Extract 10% of data groups with the longest journey times out of all
the data collected and designate the average journey time as  2 . 3) Define the index
of journey time reliability as    2  1 .
HA has pledged to improve the standard values measured between August 2004 and
July 2005 by the time of measurement between April 2007 and March 2008. Specific
improvement measures include shortening of traveling time by careful handling of
traffic accidents through cooperation with police and accurate handling of road
construction.
Japan and the United States also list the solution of traffic congestion as an index for
evaluating road administration. In Japan, however, achievement of the solution of
traffic congestion is not compulsory, and the level of guaranteeing the results with
numeric values as in the United Kingdom has not been reached. Even in the United
10
Kingdom, establishment of an index of journey time reliability on rural roads has yet to
be considered.
5.2 Journey time reliability index in rural areas
Meanwhile, in a number of countries including Japan, discussions regarding traffic in
farming villages and other rural areas have centered on traveling speeds and road safety,
and journey time efficiency has not been discussed. This is because congestion
frequently occurs in urban areas while hardly ever occurring on rural roads. From the
standpoint of the activity-based approach, which is premised on the idea that “traffic is a
demand derived from activities,” activities in rural areas have time-space constraints.
Therefore, discussions regarding a given road’s journey time efficiency are also
regarded important.
Reliability of road traffic in terms of time required is said to consist of two elements:
system reliability related to defects of the system itself, such as network disruption, and
journey time reliability related to failure to fulfill functions due to the occurrence of a
certain event even when the normal status of the system is maintained. Of these,
journey time reliability is thought to be closely related to the traffic demand that is
constantly emerging within a city. In the case of intercity connection in a rural area,
however, journey time reliability is highly related to the emergence of disturbances,
such as heavy rain, heavy snow and snowstorms, as well as extraordinary traffic
demands, and it is presumed that different service levels of reliability are required
within and between cities.
(1) Service index of journey time reliability
Now, assuming that a road user traveled between two cities, A and B, the desired time
required t 0 can be expressed as follows:
t0  tm  
(4)
tm : actual average time required,  : time of safety allowance.
Assuming that the actual time required is ti , the difference from the desired time
required t can be expressed by the equation below.
t  t 0  ti
(5)
Here, the difference in effective time required T can be expressed by the equation
below, by taking the time of risk tolerance  , which would not hinder the fulfillment of
the target activity, or the origin demand, into account.
T  t  
(6)
This means that, if T  0 , the service level within the allowable limit of the user is
achieved by the time required.
(2) Level of service required by rural area residents
As shown in Fig. 9, the time of safety allowance or risk tolerance may vary between
the snow and non-snow seasons in rural areas of cold, snowy regions, compared with
urban areas, due to greater variation in traveling speed caused by disturbances. It is,
11
however, assumed that no differences exist in terms of the allowable limit of journey
time reliability throughout the year since the target activity itself does not change.
Probability
0.450
Rural area
0.400
0.350
0.300
The probability of
decreases in speed is
higher than in the urban
area.
-4.00
-3.00
-2.00
Fig. 9
0.250
0.200
-1.00
0.450
12-hour speed distribution in
February at Nakagawa on
National Route 40
n=336
ＶＡＲ=18.7（6.2）
MaxＶ/MinV=1.6（1.76）
* Values in parentheses are
those for October.
Urban area
0.400
0.350
0.300
0.250
0.200
0.150
0.150
0.100
0.100
0.050
0.050
0.000
0.00
1.00
Probability
2.00
3.00
4.00-4.00
Regularized speed data
-3.00
-2.00
-1.00
0.000
0.00
12-hour speed distribution in
February at Takasu on National
Route 40
n=336
ＶＡＲ=9.2（2.1）
MaxＶ/MinV=1.38（1.2）
* Values in parentheses are those
for October.
1.00
2.00
3.00
4.00
Regularized speed data
Speed distribution in rural and urban sections of National Route 40
From this, maintaining the time required within the time of the allowable limit can
be listed as a target for ensuring stable traffic service in rural areas in the future (Fig.
10).
Probability
発
生
確
率
Distribution of time
required
in the non無積雪期
snow
season
所要時間分布
Distribution of
積雪期
time required in
所要時間分布
w
s
tms tmw t0 s
w
the snow season
s
t0 w
許容限界
Allowable limit
所要時間
Time
required
Fig. 10 Concept of the service index of journey time reliability in a rural area
On the assumption that the difference in effective time required in the non-snow
season is Ts and that in the snow season is Tw , supplementation of human activities,
including smooth exchanges among regions and people, can be ensured by maintenance
of a level of traffic services that satisfies the equation below. This will then contribute
to the formation of a stable society in which people can continue to live in rural areas.
12
Ts  t 0 s  tis   s  0

Tw  t 0 w  tiw   w  0
(7)
Where,
t 0s  tms   s
(8)
t 0 w  tmw   w
(9)
t0 s : desired time required in non-snow season, tis : actual time equired in non-snow season,
 s : time of risk tolerance in non-snow season, tms : actual average time required in non-snow
season,  s : time of safety allowance in non-snow season, t0 w : desired time required in snow
season, ti w : actual time required in snow season,  w : time of risk tolerance in snow season,
tmw : actual average time required in snow season,  w : time of safety allowance in snow
season
6. Conclusion
The findings of this study are as follows:
1) The reasons behind high needs for journey time reliability in rural areas with
numerous time and space constraints were discussed using the theory of population
movement.
2) The value of traveling time as seen from the viewpoint of rural areas and the value
of journey time reliability were measured using the stated preference survey, and it
was found that the value of reliability was approximately 1.5 times as high as that of
speed.
3) The service index of journey time reliability was presented as an index to evaluate
road administration with a focus on intercity traffic in rural areas.
Acknowledgment
This study was conducted with the cooperation of the Muroran Development and
Construction Department, Hokkaido Regional Development Bureau, Ministry of Land,
Infrastructure and Transport. The authors would like to express their sincere
appreciation to this Department.
References
1. Hasegawa, Arimura, Fujii, and Tamura: A Study on the Journey Time Reliability
Requested of Roadside Residents in a Rural Area, FY2005 Research paper
collection of Japan Society of Civil Engineers, Hokkaido Branch, No.62, 2006.
2. Takahiro
MIYAO:
Urban
Economics,
NIPPON
HYORONSHA
CO.,LTD.PUBLISHERS ,1995
13