DSpace at VNU: The effect of attitudes toward cars and public transportation on behavioral intention in commuting mode choice-A comparison across six Asian countries

Transportation Research Part A 69 (2014) 36–44

Contents lists available at ScienceDirect

Transportation Research Part A
journal homepage: www.elsevier.com/locate/tra

The effect of attitudes toward cars and public transportation
on behavioral intention in commuting mode choice—A
comparison across six Asian countries
Hong Tan Van a,⇑, Kasem Choocharukul b, Satoshi Fujii c
a

Department of Civil Engineering, Ho Chi Minh City University of Technology, 268 Ly Thuong Kiet District 10, Ho Chi Minh City, Vietnam
Department of Civil Engineering, Chulalongkorn University, Phayathai Road, Pathumwan, Bangkok 10330, Thailand
c
Department of Urban Management, Kyoto University, C1-2, Katsura Campus, Kyoto, Japan
b

a r t i c l e

i n f o

Article history:
Received 4 November 2008
Received in revised form 8 August 2014
Accepted 12 August 2014

Keywords:
Attitudes toward travel modes
Commuting mode choice

Mobility management
Psychological methods

a b s t r a c t
This study investigated the contribution of psychological factors in explaining the choice of
transportation mode in six Asian countries. Data were collected from 1118 respondents in
Japan, Thailand, China, Vietnam, Indonesia, and the Philippines. The dependent variable
was the intention to use one of three modes for work travel after getting a job: car, public
transit, or other modes. The explanatory variables were three attitude factors taken from a
previous study, including: 1/symbolic affective, reflecting affective motives of travel mode
use; 2/instrumental, referring to functional attributes of travel modes; and 3/social
orderliness which represents for environmental friendliness, safety, altruism, quietness
et cetera. Several logit model estimates were made using the samples from the six
countries separately and together. We obtained three main findings. First, attitude
variables about the car were all significant determinants for the entire sample from Asian
countries. Second, the social orderliness aspect of public transit was a common concern of
respondents from developing countries in selecting this mode for work trips. Third, in
countries in which the intent to use a car was not very high, attitude factors about the
car were found to be significant determinants of the behavioral intention to commute by
car but were less significant in countries in which the desire to use a car was high.
Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction
Newly industrialized countries in Asia are experiencing rapid economic growth and urbanization, resulting in soaring travel demand in many large cities (Institute for Global Environmental Strategies, 2004; Poverty and Development Division,
2006). The increases in income and standard of living in urban areas are followed by the increased trend in car use
(Hayashi et al., 2004). In fact, recent figures showed that the rates of increase in car ownership in some developing Asian
cities are much higher than those in developed Asian countries with the same per capita level of income (Morichi, 2005).
From another perspective, this trend toward the car for private travel has contributed to problems such as traffic congestion,
environmental pollution (Hayashi et al., 2004; Morichi, 2005), and global warming. Such negative effects threaten the quality
of life and mobility within societies (Mészáros, 2000; Steg and Gifford, 2005).

⇑ Corresponding author. Tel./fax: +84 8 38637003.
E-mail address: (H.T. Van).
/>0965-8564/Ó 2014 Elsevier Ltd. All rights reserved.


H.T. Van et al. / Transportation Research Part A 69 (2014) 36–44

37

Therefore, to move toward increased and widespread use of public transit in developing Asian countries, it is important
for planners to manage the use of private transportation modes. This can take the form of travel demand management (TDM)
or mobility management (MM), strategies aimed at changing the behavior of travelers.
The literature shows that TDM strategies using economic or regulatory tools with some tangible result are believed to be
particularly appropriate in developing countries (Litman, 2004; Victoria Transport Institute, 2005). However, psychological
strategies aimed primarily at changing attitudes have not yet been considered by policy makers in many Asian countries.
This is partly because very few studies have investigated the psychological factors relevant to travelers in these countries.
Meanwhile, MM measures such as the travel feedback program (Fujii and Taniguchi, 2005), TravelSmart (Department of
Transport, Western Australia, 2000), Travel Blend (Ampt and Rooney, 1999), and Individualized Marketing (Brög, 1998), have
proven to be effective and low-cost solutions for urban transport in several developed countries. Specifically, through
communication and persuasion to provide detailed travel information and incentives or to use marketing techniques
focusing on individual travel behavior, such measures have successfully induced changes in attitude, reduced car use, and
increased the use of public transit (Department for Transport, UK, 2004; Fujii and Taniguchi, 2006). In many Asian
developing countries, where a bias seems to exist toward private car versus public transit (Action Plans for Reducing
Vehicle Emissions, 2002), such psychological measures could be very important for managing transportation demand.
In considering such a strategy, it is widely known that behavioral intention is an important psychological factor that
determines actual behavior, and that attitude directly influences behavioral intention, as assumed in Ajzen’s theory of
planned behavior (Ajzen, 1991). Indeed, attitudes toward travel modes are among the determinants in deciding on the mode
of travel for commuting (Vredin Johansson et al., 2006; Kuppam et al., 1999; Recker and Golob, 1976; Schwanen and
Mokhtarian, 2005). It is therefore essential in developing countries to investigate and quantify the role of attitudes toward

travel modes on the behavioral intention of choosing a commuting mode before embarking on any psychological approaches
to MM.
Past studies have generally acknowledged that attitudes toward the car and public transit are composed of three main
components: the symbolic, the instrumental, and the affective (Gatersleben, 2004, 2007; Steg, 2003, 2005; Steg et al.,
2001). Symbolic aspects are related to the way people express social and personal identity. Affective aspects refer to emotional feelings of travelers, while instrumental aspects are mainly about benefits of using transport modes. Many studies
have used different methodologies to investigate attitudes toward the car and public transport. However, many of them,
for example, Anable and Gatersleben (2005), Hiscock et al. (2002) and Stradling et al. (1999), did not go beyond qualitative
analyses. Among quantitative studies, the research by Steg (2003, 2005) can be considered seminal in factorizing attitudes
toward travel modes. Using principal components analysis (PCA), Steg (2005) found three factors: symbolic affective, instrumental, and independence. This finding was in line with Dittmar’s (1992) model which suggested that material possessions
are strongly related to social and personal identity. In addition, after testing a theoretical model on motives regarding car use
by applying attitudinal factors, Steg (2005) concluded that symbolic and affective motives play important roles in explaining
the level of car use.
Most studies discussed above were carried out in European countries whose transportation systems differ in several ways
from those in Asian countries. Following the quantitative approach by Steg (2005) using PCA, a more recent study by Van and
Fujii (2011) in Japan, Thailand, China, Vietnam, Indonesia, and the Philippines, confirmed these three components, and added
a fourth, viz. social orderliness. This newly identified factor, social orderliness, was related to environmental friendliness,
safety, altruism, and quietness, among others. This factor reflected the diversity in culture and social life among Asian
countries in which traffic situations of some developing countries seem to be ‘‘chaotic’’ rather than orderly, especially in
urban areas, which make them different from those of developed countries. On the whole, those factors reflected how people
in Asian countries think about the car versus public transit.
This article is a continuation of the above study by Van and Fujii (2011) in which the symbolic affective, instrumental, and
social orderliness factors of attitudes toward the car and public transit were used as explanatory variables for estimating the
behavioral intention of commuting mode choices. For each country’s data, the analysis was aimed at weighting attitudinal
dimensions in terms of their contribution to explaining the future mode choice for commuting trips. The significant
determinants were expected to have an effect on future intervention policies, especially on psychological methods for MM.

2. Data
2.1. Participants
The target subjects of the study were first-year engineering students attending universities in six Asian countries, which
were selected to include both developed countries (i.e. Japan) and developing countries (i.e. China, Thailand, Indonesia, Vietnam and the Philippines). Those countries have certain similarities in culture, thus this motivates us to explore common and

different features in terms of travel behavior. All surveys were administered in late 2005 and early 2006.
Note that students are not necessarily representative of the whole population in each country. However, it was necessary
to find some common ground so that results could be compared across all the countries. If the sampling methods were
different in each country, differences in the results might be attributable to the sampling methods, rather than to country
differences. A survey targeting engineering students was our way of reducing sampling differences across countries.


38

H.T. Van et al. / Transportation Research Part A 69 (2014) 36–44

In Japan, an open recruitment campaign on the campus of the Tokyo Institute of Technology attracted 403 students by
offering an equivalent of US $1.50 to complete the survey. One response was eliminated due to missing data. In Thailand,
China, Indonesia, the Philippines, and Vietnam, surveys were administered with the help of local teachers in selected classes
and year levels at Chulalongkong University (Bangkok), Nankai University (Tianjin), Bandung Institute of Technology
(Bangdung), University of the Philippines–Diliman (Metro Manila), Ho Chi Minh (HCM) City University of Technology,
HCM City University of Architecture, and Hong Bang University (HCM City). Only those students who agreed to participate
received the questionnaire. In Thailand, 100 respondents submitted their survey forms, while 107 did so in China. In total,
122, 178, and 209 usable questionnaires were generated in Indonesia, the Philippines, and Vietnam, respectively. Descriptive
statistics of those samples are shown in Table 1.
The table shows that except for the samples from the Philippines and China, which were approximately equal in terms of
gender, the rest had more males than females. All samples had similar mean age values. The percentage of respondents with
a driver’s license and a car was highest in Thailand and lowest in China and Vietnam.
2.2. Questionnaire
The questionnaire for the study was first written in English for use in the Philippines, and then translated into the official
languages of the other countries. Each translation was checked by an editor who understood both English and the local
language, and who resolved any inconsistencies with the translator.
Information on attitudes toward travel modes was collected using the semantic differential technique with a five-point
bipolar adjective scale. For example, for the pair ‘‘poor-rich’’, ‘‘poor’’ was written on the utter left and ‘‘rich’’ was placed on
the utter right side of the five-point scale. Accordingly, respondents can choose one out of five checkboxes to imply whether

the image of the car/public transport is nearer to ‘‘poor’’ or ‘‘rich’’. The ideas of car and public transport modes were
stimulated in the respondent’s minds by the large printed words ‘‘Automobile’’ and ‘‘Public Transport’’ in the instructions.
The questionnaire then showed randomly arranged beliefs in the form of pairs of opposite adjectives, which covered several
aspects of travel modes such as the symbolic, instrumental affective, and social moral attributes. In total, 32 beliefs were
used in common for both modes of car and public transit; however, one belief was eliminated due to mistranslation (see
31 beliefs the Appendix A). For example, ‘‘austere–luxurious’’ and ‘‘cheap–expensive’’ were used to measure the symbolic status dimension, ‘‘inconvenient–convenient’’ and ‘‘useless–useful’’ for the instrumental aspect, and ‘‘destructive–constructive’’
and ‘‘negative–positive’’ for social orderliness.
After completing the two sets of 32 beliefs for car and public transit, the respondents were asked to choose which kind of
travel mode they intended to use to commute to work in the future after getting a job. The proposed alternatives in the
choice set were car, public transit, walking, motorbike, bicycle, or other choice specified by the respondent. However, since
the study’s objective was focused mainly on choices between the car and public transit, the others were combined into an
‘‘other modes’’ category. The responses of mode choice by country are summarized in Table 2.
As shown in Table 2, the car was the predominant intended travel mode in all countries except Japan. This was especially
true in the samples from the Philippines, in which nearly 90% of the respondents intended to drive to work after getting a job.
Note that students’ intended behavior for commuting is expected to have a long-term effect on transportation conditions
in the region. This is because commuting behavior is usually habitual. While the habit is usually formed immediately after
getting a job, such a habit is expected to be influenced by the behavioral intention developed before actually starting the job,
i.e., during the study period (Fujii and Gärling, 2003).
2.3. Attitude factors
To obtain the attitude factors from the beliefs rated by the respondents, this study took advantage of the work by Van and
Fujii (2011). In that study, the pooled data set of 31 beliefs for car and public transport evaluated by respondents in all six
countries was subjected to a principal components analysis (PCA) using varimax rotation. The result of the PCA identified
three main factors, accounting for 52.6% of the total variance underlying the attitudes toward the car and public transit.
The first factor accounted for 36.9% of the total variance. Belief loads on this factor mostly described the symbolic and

Table 1
Descriptive statistics of samples in Japan, Thailand, China, Vietnam, Indonesia, and the Philippines.
Sample

N

Japan
Thailand
China
Vietnam
Indonesia
Philippines

402
100
107
209
122
178

Gender

Age

Driver’s license

Car ownership

% male

% female

Mean

SD

% having

% not having

% having

% not having

91
87
61
84
88
53

9
13
39
16
12
47

20.35
21.09
20.67
21.99
19.67
19.16

2.08
1.66
2.08
2.37
2.06
1.70

50
75
6
10
56
26

50
25
94
90
44
74

10
53
2
1
10
9.7

90
46

98
99
90
89

Note: In cases of less than a 100% sum, the remaining percentage is attributable to missing data.


39

H.T. Van et al. / Transportation Research Part A 69 (2014) 36–44
Table 2
Percentage indicating a stated intention of choice for each mode by country.

Car
Public transport
Other modes

Japan (N = 402)

Thailand (N = 100)

China (N = 107)

Vietnam (N = 209)

Indonesia (N = 122)

Philippines (N = 178)

22.4
47.0
30.6

71.0
18.0
6.0

47.7
18.7
33.6

49.3
17.2
33.5

59.8
11.5
28.7

89.9
3.9
5.1

Note: In cases of less than a 100% sum, the remaining percentage is attributable to missing data.

affective characteristics of travel modes. For example, beliefs such as richness, luxury, superiority, and ‘‘coolness’’
represented prestige values, while comfort, excitement, and relaxation were related to affective feelings. This factor was thus
referred as symbolic affective. The second factor, instrumental, made up 10.9% of the total variance. Covering the functional
attributes of the travel modes, this factor consisted of beliefs such as usefulness, convenience, simplicity, and speed. The last

factor, whose beliefs reflected the social orderliness aspect of travel modes included the terms environmentally friendly, safe,
altruistic, and quiet.
The rotated factor loadings of 31 beliefs on the three scales are shown for reference in the Appendix A. Note that the
beliefs were translated so that respondents in each country could similarly understand their meanings and could rate car
and public transport according to the same scales. In PCA, factor loadings of each belief on each scale are computed from
the constituents of the eigenvector of the correlation matrix. This eigenvector determines the direction of that scale. Then,
the new attitudinal variables are computed by linearly combining all observed variables whose weights are determined by
their factor loadings. Using this transformation, the three new variables, generated by PCA in that study, were therefore
evaluated in an absolute sense.
In addition, it should be noted that to make it possible for comparing psychological factors between car and public
transport, the pooled dataset was used in PCA to generate psychological variables for both car and public transport while
assuming that car and public transport have similar structure of attitudinal factors. To check the validity of this assumption,
we compared the result of PCA on pooled dataset with the results of two PCAs, implemented separately for car and public
transport. A similar result was found, that is attitudes toward car and attitudes toward public transport separately can also
be classified into three similar conceptual scales, namely symbolic affective, instrumental and social orderliness. Specifically, a
majority of the beliefs similarly had high factor loadings on respective conceptual scales (between car and public transport
and between two ways of analysis). This implies that constructing attitudinal factors of car and public transport on a same
scale structure would not yield a large difference from constructing attitudinal factors of car and public transport on different
scale structures. Thus the generated variables in Van and Fujii (2011) can be used in the following analysis for comparing
attitudinal factors across countries.
The descriptive statistics of these variables in Table 3 show that a general contrast existed between the car and public
transport in terms of symbolic affective values; that is, the symbolic affective factor for the car was larger than that for public
transport. In addition, if comparing across countries, we can speculate that respondents in a higher income society like Japan,
where it may be economically easier to own car, will perceive having a car as having a lower symbolic status compared to
those in lower-income countries. Regarding the instrumental aspect, the result indicated that scores for the car were higher
than those for public transport in some countries, but lower in others; for example, among the six countries, the score on the
instrumental aspect for the car was highest in Japan and lowest in Indonesia. For the social orderliness of travel modes, the
relationship of scores between the car and public transport were different across countries. Details of the comparisons across
different countries and travel modes can be found in Van and Fujii (2011).

Table 3
Mean (M) and standard deviation (SD) of symbolic affective, instrumental, and social orderliness values for the car and public transport.
Country sample

Car
Symbolic
affective
Instrumental
Social
orderliness

Japan (N = 400) M
(SD)

Thailand (N = 99)
M (SD)

China (N = 107)
M (SD)

Vietnam (N = 209)
M (SD)

Indonesia (N = 122)
M (SD)

Philippines (N = 173)
M (SD)

0.48 (0.52)

0.79 (0.53)

0.57 (0.66)

1.15 (0.73)

1.00 (0.67)

0.93 (0.61)

0.52 (0.83)
À1.09 (0.77)

À0.02 (0.85)
0.39 (0.67)

0.04 (0.79)
À0.31 (0.80)

À0.33 (1.06)
0.26 (0.78)

À0.50 (1.04)
0.01 (0.75)

0.31 (0.60)
0.22 (0.71)

À0.80 (0.58)

À1.01 (0.55)

À0.52 (0.63)

À0.90 (0.70)

À1.03 (0.55)

À1.19 (0.82)
0.01 (0.71)

0.09 (1.02)
0.32 (0.89)

À0.25 (1.11)
0.67 (1.01)

0.28 (0.98)
À0.25 (0.74)

À0.02 (0.92)
À0.47 (0.67)

Public transport
Symbolic
À0.71 (0.51)
affective
Instrumental
À0.02 (0.94)

Social
0.68 (0.78)
orderliness


40

H.T. Van et al. / Transportation Research Part A 69 (2014) 36–44

3. Multinomial logit analysis
To examine the extent to which attitude factors contribute to explaining participants’ behavioral intention regarding
future commuting mode choice, we used multinomial logit (MNL) models of the ‘‘mode of travel to work that respondents
intended to choose after getting the job.’’ The three attitude factors toward the car and public transit, that is, symbolic affective,
instrumental, and social orderliness, derived from PCA, were used as explanatory variables. For the choices of car and public
transit, the utility functions were parameterized as a function of the three attitude factors corresponding to each mode
and undetermined attributes of the alternative. However, the utility function for the choice of other modes only contained
undetermined components since the survey covered only the attitudes toward the car and public transit. Thus, the other
modes group provided a reference to compare with the other two modes. In addition, due to the variation in gender, age,
driver licensing, and car ownership among countries surveyed, these variables were incorporated as covariates in the utility
function. This was mainly to eliminate those effects in estimating the roles of the three attitude dimensions on the choice of
transportation mode. However, not all covariates could be included in some sample groups due to an inadequate size for
estimating the respective parameters. Two types of models were estimated, including one using the data from all countries
and the other six models using data from individual countries. The estimation results are given in Table 4.
 2 for the seven models ranged from 0.034 to 0.153, which were
As shown in Table 4, the adjusted goodness of fit indices q
2
quite low. However, the large v indicates that the models could explain the data set well. First, the estimation using the data
from all countries yielded the result that the attitude factors were positively related to the behavioral intention regarding
commuting mode choice, and all coefficients for the car were larger than those for public transit. As also shown in Table 4,
the constant for the car was significant and much larger than that for public transit. Moreover, gender and car ownership had

significant coefficients in the behavioral intention to choose the car. Specifically, female respondents and respondents
owning a car were more likely to intend to use the car for their future work trips. Though such a result might not be the
same for female in European countries, it could be reasonable for Asian countries where public transport services are
generally overcrowded (like in Tokyo, Japan) and chaotic (like in other developing countries). Women might have certain
feelings of insecurity when using public transit. Such fears of being harassed might overwhelm thinking about difficulties
in maneuvering a car, thus cause women’s stronger intentions to use car than men’s.
Investigating the differences between countries, we found that none of the coefficients for attitude variables in the sample
from the Philippines was significant, although this was not true for the other countries. Interestingly, samples collected in
developing countries including Thailand, Vietnam, and Indonesia, showed that the social orderliness aspect of the attitude
toward public transit had a significant coefficient. For respondents in a developed country of Japan, symbolic affective
and instrumental aspects of the attitude toward the car and the instrumental aspect of attitude toward public transit were

Table 4
Parameter estimates regarding the intention of mode choices for the entire sample of six countries and for the samples within each country.
Explanatory
variable

All (N = 1096)

Japanese
(N = 400)

Thai (N = 92)

Chinese
(N = 100)

Vietnamese
(N = 209)

Indonesian
(N = 122)

Filipinos
(N = 172)

B

t

B

t

B

t

B

t

B

t

B

t

B

t

Constant_car
Symbolic
affective_car
Instrumental_car
Social
orderliness_car
Gender
License
Car ownership
Constant_PT
Symbolic
affective_PT
Instrumental_PT
Social
orderliness_PT
Gender
License
Car ownership

0.64
0.62

3.16**
5.01***

À1.76

0.93

À2.48*
3.55***

1.63
0.08

1.29
0.13

0.40
0.68

0.89
1.77#

À0.80
0.46

À1.63
1.66#

0.66
0.33

1.50
1.03

2.73

0.71

3.8***4
1.38

0.50
0.78

6.22***
8.89***

0.81
0.19

4.43***
1.08

0.37
0.53

0.90
0.99

0.84
0.76

2.59*
2.26*

0.69

0.51

3.95***
2.17*

0.35
0.22

1.61
0.80

0.65
À0.13

1.23
À0.30

À0.50
0.08
0.81
0.14
0.32

À2.48*
0.42
2.56*
0.60
2.31*

0.51

0.19
À0.03
1.21
0.35

0.79
0.63
À0.07
2.69**
1.57

0.21
1.95
À1.45
À0.39
À1.43*

0.17
1.81#
À1.28
À0.27
À2.04

À0.42
–
–
0.45
1.11

À0.80

–
–
0.57
1.64

0.90
–
–
À0.71
0.40

1.88#
–
–
À1.42
1.25

–
À0.15
–
À1.48
À0.58

–
À0.35
–
À2.06*
À1.15

À0.79

À0.51
À0.12
1.32
1.38

À1.05
À0.58
À0.09
1.25
1.19

0.24
0.60

3.09**
6.57***

0.34
0.16

2.87**
1.13

À0.64
1.93

À1.46
2.71**

0.25

0.79

0.68
1.93#

0.56
0.48

2.87**
2.29*

0.33
1.29

0.91
2.30*

0.30
0.21

0.53
0.25

À0.34
0.18
0.62

À1.49
0.90
1.76

À0.72
À0.08
0.32

À1.73#
À0.31
0.75

À0.28
0.96
À2.00

À0.21
0.77
À1.60

À1.26
–
–

À1.75#
–
–

À0.14
–
–

À0.26

–
–

À
À0.15
–

À
À0.22
–

À1.05
À0.15
0.36

À0.89
À0.11
0.18

v2 (d.f.)

q

314.4 (12)
0.131

64.9 (12)
0.061

23.3 (12)

0.111

40.1 (8)
0.153

Note: Variable of Gender is a dummy variable (0 = female; 1 = male).
License and Car ownership are also dummy variables (0 = not having; 1 = having).
#
p < 0.1.
*
p < 0.05.
**
p < 0.01.
***
p < 0.001.

56.1 (8)
0.111

17.0 (8)
0.037

9.0 (12)
0.034


H.T. Van et al. / Transportation Research Part A 69 (2014) 36–44

41

found to have a positive and significant coefficient in the behavioral intentions of car and public transit use. The constant for
the car in the Japanese model was also found to be significant, but the magnitude was much smaller than the constant
coefficient for public transit.

4. Discussion
Based on the combined sample from all six countries, we found that all psychological factors were significantly related to
the behavioral intention regarding commuting mode choice. This result confirms the hypothesis that attitudes may be
predictors of behavioral intention, as stated in the theory of planned behavior (Ajzen, 1991). More importantly, every
attitude aspect, that is, symbolic affective, instrumental, and social orderliness, was positively significant for both public transit
use and car use. This implied that in general, Asian university students have a higher probability of using public transit (or
car) when the symbolic affective aspect, instrumental aspect, and social orderliness aspect of attitudes toward public transit
(or car) improve.
Note that we sampled first-year students and investigated their intentions regarding modes for work travel after
graduating and getting jobs. It is known that career choice clearly represents a deliberate and planned behavior (Harren,
1979; Super and Hall, 1978; Arnold et al., 2006). Thus it is assumed that together with job decision, considerations about
commuting mode might be made early by a student. However, it should be conceded that the variation related to the time
factor in the choice process could therefore make the findings of this study less reliable than those of other possible studies
that investigate the intention just prior to the respondents making the actual choices. Nevertheless, the intention reported
here should relate to their future intention at least to some extent. The results therefore simply imply that in general,
attitude factors may be useful for understanding, predicting, and changing the choice of commuting mode in Asian countries.
Future research should verify this hypothesis based on a sample representative of the whole population.
From the combined model, we also found that the contributions of attitude variables in explaining the intention to use a
car were larger than those for public transit. This implies that the behavioral intention to use a car compared to public transit
was more strongly related to attitudes. In addition, the fact that the constant for the car was larger than that for public transit
demonstrates a common tendency to prefer the car in Asian countries, as previously noted (Action Plans for Reducing Vehicle
Emissions, 2002). In contrast, the significant and large negative constant for the car in the Japanese sample as opposed to the
positive constant for public transit implies that everything else being equal, public transit would be the preferred mode only
in Japan.
The result shows that no attitude factors affected behavioral intention regarding the commuting mode choice in the Philippines, but some attitude factors were significant for the other five countries. This may have resulted because the constant
for the car was positively significant only in the case of the Philippines and was the largest among the six countries, indicating that the Filipino participants had a strong positive bias toward car use. A report by ACNielsen (2005) revealed that the car

ownership aspiration index, which based on ‘‘current ownership’’ levels and ‘‘future intention’’ to buy a private car in the
next 12 months, in the Philippines was among the highest in the Asian region. Moreover, unlike other surveyed cities, in
terms of regulation, it is very easy to possess a car in Metro Manila as long as one has money, and a general belief among
various households is that cars, including used ones, suiting any budget will be available (Rubite, 2004). This could foster
a kind social norm for the younger generation in the Philippines to think that car ownership is requisite for expressing
freedom and status. This possible social norm of respondents in the Filipino sample may have overshadowed the role of
the other psychological factors on the intention of future mode choice.
Apart from the sample from the Philippines, the country-specific models produced some interesting results. First, the
social orderliness aspect of public transit was a common concern of the respondents in developing countries (i.e., Thailand,
China, Vietnam, and Indonesia) in terms of the behavioral intention of selecting public transit for commuting. This implies
the desirability of improving public transit service in terms of ‘‘social orderliness’’. For example, moderating the destructive,
risky, and aggressive image of public transit while spreading commendable beliefs about the modesty and altruism of public
transport use and improving safety and environmental friendliness would be highly important for promoting public transit
in developing countries, although such improvements would not be as important in a developed country like Japan. This
result may have occurred because the performance of the current public transit vehicle fleets in developing countries is still
generally at a low level (The Korea Transport Institute, 2000; Van and Fujii, 2011).
In a different vein, the behavioral intention to commute by public transit in Japan was related to the instrumental aspect
of the attitude toward public transit. The explanation for this could be that the social orderliness aspect of public transit in
Japan is already considerably higher than in developing countries, permitting focus on other aspects of public transit, i.e., the
instrumental aspect. In other words, the cognitive resources (cf., Payne et al., 1993) for behavioral intention regarding the
commuting mode choice by the Japanese participants may be capable of being directed toward the instrumental aspect
of public transit rather than toward social orderliness, or the level of chaos, which might otherwise have had a high cognitive
impact. This explanation seems to be consistent with the fact that Vietnam, whose social orderliness score was as high as
Japan’s (see Table 3), also had a significant positive instrumental coefficient for public transit. The fact that the model for
Vietnam had significant coefficients for both the instrumental and social orderliness aspects of attitudes toward public
transit might have occurred because the objective social orderliness of public transit in Vietnam may lie somewhere between
that of Japan and the other countries.


42

H.T. Van et al. / Transportation Research Part A 69 (2014) 36–44

In addition, the results from country-specific models indicated that the symbolic affective aspect of attitudes toward
public transit did not have any positive effect in relation to the behavioral intention to commute by public transit. This
implies that the symbolic affective aspect of public transit is as important as that of the car (discussed below).
No attitude variables were significantly related to the behavioral intention to commute by car in Indonesia and Thailand,
as well as in the Philippines. Note that these countries are the top three in the present study in terms of the frequency of
participants who stated their intention to commute by car. This strong tendency toward the car might overshadow the role
of attitude variables for the car on the behavioral intention of driving to work.
Except for Thailand, Indonesia, and the Philippines, both the instrumental and symbolic aspects of attitudes toward the
car had significant or marginally significant positive coefficients. This implies that students will intend to commute by car if
they believe that the car is a mode with high symbolic affective attributes (e.g., richness, luxuriousness, superiority, coolness)
and/or with high instrumental features (e.g., usefulness, convenience, simplicity, friendliness).
A difference between Japan and China or Vietnam with respect to the coefficients for attitude variables for the car
appeared in the social orderliness aspect of the attitude toward the car; this was not significant for Japan but was positively
significant for China and Vietnam. This may reflect the large difference in the mean values for social orderliness of the car
between Japan and the other countries. As Table 3 shows, the car seems to be considered inappropriate in terms of social
orderliness only in Japan. Therefore, the model estimation result may imply that a lower level of appropriateness in terms
of social orderliness did not play a large role in explaining the behavioral intention to use the car as indicated in the Japanese
model, whereas a higher level of appropriateness in terms of social orderliness did have a significant contribution, as Table 3
shows for the Chinese and Vietnamese models.
5. Conclusions
In Van and Fujii (2011), we found symbolic affective, instrumental and social orderliness when investigating the attitudes
toward car and public transport in six Asian countries. This verified the existence of the three basic aspects of travel modes
i.e. symbolic, affective and instrument, found by several researchers on this field in Europe (e.g. Gatersleben, 2004; Steg,
2003, 2004 etc.), and also repeated the finding by Steg (2001, 2005) that there is no clear distinction between symbolic
and affective factors. Social orderliness in that study was a newly-found factor that made it distinctive from the previous studies. Inheriting such an exceptional result, this study moves to a further step, showing that attitude factors toward the car and
public transit provide valuable insights into the determinants of behavioral intention regarding the commuting mode choice
in Asian countries. The main findings in this study are as follows.

1. In general, attitude variables (i.e., symbolic affective, instrumental, and social orderliness) for the car and public transit are
all significant determinants for the samples from six Asian countries.
2. The social orderliness aspect of attitude toward public transit was a common concern of respondents in developing countries but not in Japan. This means that the improvement of the public transit image in the aspect of social orderliness is
highly important for promoting it as a commuting mode. Policy makers should develop education programs as parts of
transit promotion campaigns. Specific intervention programs need to be done to improve people behavior when using
public transport modes. This will actually improve the image of the public transport modes which may have profound
effects on attracting people to use public transit. In addition to that, in order to reduce social dilemmas caused by defective behaviors which consequently destruct the image of public transport, education measures should be basically and
systematically applied to children in developing countries to help them understand and practice cooperative behaviors.
This is important not only in building a sustainable transportation system but also in developing country in a whole
meaning.
3. In countries like Japan, China, and Vietnam, where the intention or desire to use the car is not high, instrumental,
symbolic, and affective aspects of attitudes toward the car are significant determinants of the behavioral intention to
commute by car. In other countries like Thailand, Indonesia, and the Philippines, however, where the intention is high,
the models do not explain the variance in the behavioral intention to drive to work.
All of these findings might have contributed to understanding the possible role of modifying attitudes toward travel
modes in managing transportation demand for the countries included in this study. Based on our findings, we speculate that
the effectiveness of such approaches, however, may be limited in countries such as the Philippines, Indonesia, and Thailand,
where the roles of the attitude variables were limited. Our primary findings, if verified by further studies using a sample of
current commuters, would be useful in designing appropriate transport measures to manage mobility or transportation
demand. Such management measures would be important for improving the quality of life in Asian countries, as well as contributing to mitigating the effects of motor vehicles on global warming.
Acknowledgments
The survey in this study was supported by the research group for the ‘‘Applicability of Ecological Transportation Systems
in Developing Countries’’ of the International Association of Traffic and Safety Science. We thank all those who worked with


43

H.T. Van et al. / Transportation Research Part A 69 (2014) 36–44

us in collecting data in the different countries. We also acknowledge grants from the Japan Society for the Promotion of Science, which made it possible for Kasem Choocharukul to visit the Tokyo Institute of Technology to collaborate in this

research and acknowledge grants by Japanese Government to support scholarship for Hong Tan Van for his PhD study in
Tokyo Institute of Technology to make this research.
Appendix A.
Rotated factor loadings of beliefs on travel modes by principal components analysis (PCA)
Method

PCA, rotation method: varimax with Kaiser normalization

Belief

Factor
Symbolic affective

Poor–Rich
Austere–Luxurious
Inferior–Superior
Uncool–Cool
Cheap–Expensive
Vulgar–Aristocratic
Traditional–Advanced
Dirty–Clean
Unattractive–Attractive
Uncomfortable–Comfortable
Outdated–Fashionable
Public–Personal
Bored–Excited
Unpleasant–Pleasant
Non-esteemed–Esteemed
Slow–Fast
Stressful–Relaxed

Uncontrollable–Controllable
Not free–Free
Modest–Arrogant
Negative–Positive
Useless–Useful
Inconvenient–Convenient
Complicated–Simple
Unfriendly–Friendly
Environmental damaging–Environmental friendly
Risky–Safe
Egoistic–Altruistic
Destructive–Constructive
Aggressive–Nonaggressive
Boisterous–Quiet
Variance contribution (%)

0.813
0.790
0.788
0.786
0.781
0.745
0.738
0.734
0.728
0.728
0.701
0.691
0.682
0.621

0.592
0.584
0.565
0.547
0.530
0.494
0.451

0.477
36.9

Instrumental

Social orderliness

0.492
0.438

0.461
-0.428
0.402
0.648
0.625
0.532
0.511

10.9

0.727
0.653

0.549
0.539
0.522
0.492
4.8

Note: Only factor loadings |P0.4| are noted.
Three variables generated by PCA were subsequently used in the MNL.

References
ACNielsen., 2005. Consumers in Asia Pacific—car ownership and our purchase intentions. a report on global consumers. Available at www2.acnielsen.com/reports/documents/2005_automotive.pdf>.
Action Plans for Reducing Vehicle Emissions (RETA 5937), 2002. Synthesis and recommendations: reducing vehicle emissions. In: Regional Workshop:
Transport Planning, Demand Management & Air Quality, Manila, the Philippines.
Ajzen, I., 1991. The theory of planned behavior. Organ. Behav. Hum. Decis. Process. 50 (2), 179–211.
Ampt, E., Rooney, A., 1999. Reducing the impact of the car—a sustainable approach. Paper presented at Travel Smart Adelaide, 23rd Australian Transport
Forum, Perth, Australia.
Anable, J., Gatersleben, B., 2005. All work and no play? The role of instrumental and affective factors in work and leisure journeys by different travel modes.
Transp. Res. Part A 39 (2–3), 163–181.
Arnold, J., Loan-Clarke, J., Coombs, C., Wilkinson, A., Park, J., Preston, D., 2006. How well can the theory of planned behavior account for occupational
intentions? J. Vocat. Behav. 69 (3), 374–390.
Brög, W., 1998. Individualised Marketing: Implications for TDM. CD-ROM of the Proceedings of the 77th Annual Meeting of the Transportation Research
Board.


44

H.T. Van et al. / Transportation Research Part A 69 (2014) 36–44

Department for Transport, UK, 2004. Smarter choices: changing the way we travel. London, UK.

Department of Transport, Western Australia, 2000. TravelSmart: a cost-effective contribution to transport infrastructure. Perth, Australia.
Dittmar, H., 1992. The Social Psychology of Material Possessions: To Have is To Be. Hemel Hempstead, UK.
Fujii, S., Gärling, T., 2003. Development of script-based travel mode choice after forced change. Transp. Res. Part F 6 (2), 117–124.
Fujii, S., Taniguchi, A., 2005. Reducing family car use by providing travel advice or requesting behavioral plans: an experimental analysis of travel feedback
programs. Transp. Res. Part D 10 (5), 385–393.
Fujii, S., Taniguchi, A., 2006. Determinants of the effectiveness of travel feedback programs—a review of communicative mobility management measures for
changing travel behaviour in Japan. Transp. Policy 13 (5), 339–348.
Gatersleben, B., 2004. Affective, social and instrumental aspects of the commute to work: comparing perceptions of drivers, public transport users, walkers
and cyclists. Paper presented at the 3rd International Conference on Traffic & Transport Psychology, Nottingham, UK.
Gatersleben, B., 2007. Affective and symbolic aspect of car use: a review. In: Gärling, T., Steg, L. (Eds.), Threats to the Quality of Urban Life from Car Traffic:
Problems, Causes, and Solutions. Elsevier, Amsterdam.
Harren, V.A., 1979. A model of career decision making for college students. J. Vocat. Behav. 14 (2), 119–133.
Hayashi, Y., Doi, K., Yagishita, M., Kuwata, M., 2004. Urban transport sustainability: Asian trends, problems and policy practices. Eur. J. Transp. Infrast. Res. 4
(1), 27–45.
Hiscock, R., Macintyre, S., Kearns, A., Ellaway, A., 2002. Means of transport and ontological security: do cars provide psycho-social benefits to their users?
Transp. Res. Part D 7 (2), 119–135.
Institute for Global Environmental Strategies, 2004. Urban energy use and greenhouse gas emissions in Asian Mega-cities—policies for a sustainable future.
In: Urban Environmental Management Project. Available at < />Kuppam, A.R., Pendyala, R.M., Rahman, S., 1999. Analysis of the role of traveler attitudes and perceptions in explaining mode-choice behavior. Transp. Res.
Rec. 1676, 69–76.
Litman, T., 2004. Mobility management. Sustainable Transport: A Sourcebook for Policy-makers in Developing Countries eds Deutsche Gesellschaft für
Technische Zusammenarbeit (GTZ), Available at < />Mészáros, P., 2000. Towards sustainable mobility—conflicts impacts, principles, conditions, aspect of CEECs. Periodica Polytech. Transp. Eng. 28 (1–2),
65–75.
Morichi, S., 2005. Long-term strategy for transport system in Asian megacities. J. East. Asian Soc. Transp. Stud. 6, 1–22.
Payne, J.W., Bettman, J.R., Johnson, E.J., 1993. The Adaptive Decision Maker. Cambridge University Press, New York.
Poverty and Development Division (The United Nations Economic and Social Commission for Asia and the Pacific-UNESCAP), 2006. Enhancing Regional
Cooperation in Infrastructure Development Including that Related to Disaster Management.
Recker, W.W., Golob, T.F., 1976. An attitudinal modal choice model. Transp. Res. 10 (5), 299–310.
Rubite, C.P., 2004. Analysis of car ownership in Metro Manila. A research report submitted to the Japan Society for the Promotion of Science (JSPS). Available
at < />Schwanen, T., Mokhtarian, P.L., 2005. What affects commute mode choice: neighborhood physical structure or preferences toward neighborhoods? J.
Transp. Geogr. 13 (1), 83–99.

Steg, L., 2003. Can public transport compete with the private car? IATSS Res. 27 (2), 27–35.
Steg, L., 2004. Instrumental, social and affective values of car use. Paper presented at the 3rd International Conference on Traffic & Transport Psychology,
Nottingham, UK.
Steg, L., 2005. Car use: lust and must. Instrumental, symbolic and affective motives for car use. Transp. Res. Part A 39, 147–162.
Steg, L., Gifford, R., 2005. Sustainable transportation and quality of life. J. Transp. Geogr. 13 (1), 59–69.
Steg, L., Vlek, C., Slotegraaf, G., 2001. Instrumental-reasoned and symbolic affective motives for using a motor car. Transp. Res. Part F 4, 151–169.
Stradling, S.G., Meadows, M.L., Beatty, S., 1999. Factors Affecting Car Use Choices. Transport Research Institute: Napier University.
Super, D.E., Hall, D.T., 1978. Career development: exploration and planning. Annu. Rev. Psychol. 29, 333–372.
The Korea Transport Institute, 2000. Development and Dissemination of a Best Practices Manual to Address Urban Transport Problems in the APEC Region.
Final report of the project ‘Development and Dissemination of a Best Practices Manual to Address Urban Transport Problems in the APEC Region’,
prepared for the APEC Transportation Working Group.
Van, H.T., Fujii, S., 2011. A cross Asian country analysis in attitudes toward car and public transport. J. East. Asia Soc. Transp. Stud. 9, 411–421.
Victoria Transport Institute, 2005. Prestige and pleasure, mobility as a prestige good and as a pleasurable activity – Implications for transport planning. In
Online TDM Encyclopedia.
Vredin Johansson, M., Heldt, T., Johansson, P., 2006. The effects of attitudes and personality traits on mode choice. Transp. Res. Part A 40 (6), 507–525.