Influential Factors on Pro-Environmental Behaviors—A Case Study in Tokyo and Seoul

To develop the low-carbon society, in addition to the efforts by industrial and commercial sectors, promotion of people’s pro-environmental behaviors (PEBs) has become one of the key issues. Some PEBs have been investigated in previous studies, however, the targets were limited to particular behaviors (i.e., recycling, water saving, electricity saving). The holistic view of understanding the characteristics of PEBs has remained insufficient. In this study, we selected 58 daily PEBs from various fields and investigated people’s practice rates and attitudes in Seoul and Tokyo. The questionnaire surveys were conducted and 2393 (Seoul) and 2220 (Tokyo) valid responses were analyzed. Most PEBs had significantly different practice rates between Seoul and Tokyo. It can be concluded that the surrounding conditions, such as policy and infrastructure, have some influences on these differences. The positions of the reasons to practice or not to practice PEBs were visualized using multiple correspondence analyses. The results indicated that the monetary reason was the common factor for many PEBs, while some PEBs showed different reasons. The socio-demographic effects were not significantly different between the two regions. Commonly, women and older people showed more activities on PEBs.


Introduction
The promotion of pro-environmental behaviors (PEBs) in people's daily lives is one of the key issues in developing sustainable societies.According to the national census for environmental conservation [1], more than 80% of people in South Korea express great environmental concerns.The Korean Ministry of Environment started the "Green Start Movement" in 2008 and has intensively promoted people's PEBs through campaigns, video clips, and leaflet distributions.In 2009, the Korean government established "Low Carbon Green Growth," and the "Presidential Committee on Green Growth" to achieve a 30% reduction in greenhouse gas emission by 2020.The "Carbon Cashbag" and "Carbon Footprint" programs were started in response.The "Carbon Cashbag" program offers various low-carbon products, and people who buy these products are reimbursed with shopping-points.Similar to programs in many other countries, the "Carbon Footprint" program shows the lifetime carbon dioxide emission of a product.These programs can enhance people's environmental awareness.
Even if people have a high awareness of environmental issues, there are many internal and external barriers to taking actual action [2][3].Although people may intend to practice PEBs, various factors such as traditional values, lifestyle, and surrounding circumstances can influence their behavior.In the field of PEB studies, to understand the PEBs deeply, various cross-national studies have been conducted [4][5][6][7][8][9].
Previous studies have also paid much attention to methods for enhancing PEBs and environmental awareness.Many studies are found in the fields of recycling [10][11][12][13][14], electricity saving [15][16][17][18][19], and water saving behaviors [20][21][22].However, each study mainly focuses on one particular PEB, and studies comprehensively dealing with various PEBs are rarely found.In response to the lack of the holistic understandings of PEBs, this study deals with various PEBs and discusses their characteristics through comparison of Seoul and Tokyo citizens.Seoul is the capital of the Republic of Korea, with an area of 605 km 2 and a population of 10.5 million.
Tokyo is the capital of Japan, and its central 23 wards, with an area of 622 km 2 and a population of 8.7 million, form the target area for this study.Although these cities are geographically close (about 1150 km apart) in the Asian region, we assume that differences in citizens' behavioral patterns arise from different internal and external reasons.This paper aims to reveal the characteristics of various PEBs through a questionnaire survey.It will particularly focus on the practice rate of each PEB, people's reasons to practice or not practice the behavior, and the effect of socio-demographics.

Questionnaire Design
To achieve the research aims, we decided to use an online questionnaire survey.In this type of survey, people who register via a web research company are requested to fill out a questionnaire, and points are given to correspond to their answers.More general respondents can be obtained through this online method, as the respondents with higher intension tend to be willing to mail back in the case of conventional postal questionnaires.The quality of registered respondents is also controlled by the web research company.This online method is suitable for conducting a large-scale questionnaire surveys and for obtaining a quick and high response rate.The questionnaire design and analysis of the obtained raw data attribute to a researcher.
The questionnaire consisted of three parts: 1) practice rates of PEBs, 2) reasons to practice or not to practice each PEB, and 3) socio-demographics.
We selected 58 daily PEBs (Table 1) based on the previous selection by Aoki et al. [23], which was based on the behaviors recommended by the national and local governments and environmental NPOs in Japan.In addition to Aoki et al.'s selection, we added or subtracted several behaviors considering the circumstances of each city.For example, "B49: Carbon Cashbag" is a program that only exists in Korea, and therefore this behavior was removed from the list used in Tokyo.Hence, 56 and 52 PEBs were selected for use in Seoul and Tokyo respectively, with 50 PEBs common to both cities.Each respondent's practice rate was measured using the 4-point Likert Scale: "always," "often," "rarely," and "never."Regarding behaviors that require that the respondents have access to certain equipment, such as "B10: Personal computer," "B13: Bidet," "B55: Dishwasher," and behaviors relating to cars (B34-B38), respondents were first asked whether they had the equipment, and respondents who did own the equipment then answered the questions.In part 2, 11 reasons to practice and 13 reasons not to practice PEBs were presented to the respondent as shown in Table 2. Respondents who answered "always" or "of-ten" in Part 1 were asked the reasons why they practiced the behavior, and respondents who answered "rarely" or "never" were asked the reasons why they did not practice the behavior.The respondents chose their primary and secondary reasons for practicing or not practicing the behavior from the list of reasons for each PEB.
Part 3 asked for details of socio-demographics, such as gender, age, occupation, and income.

Survey and Analysis
The surveys were conducted by EZ Survey Co. in Seoul from November 2 to 10, 2010, and by Nikkei Research Co. in Tokyo from February 1 to 13, 2011.The questionnaire was completed by men and women aged 20 -50 in Seoul and 10 -70 in Tokyo.The obtained sample numbers were 5546 and 3489, respectively.
When comparing the results of the two cities, the age and gender distributions should fit the distributions of the parent populations.Therefore, respondents aged 20years were randomly extracted from the respondents to fit the distributions recorded in the 2010 National Censuses of both cities.Finally, 2393 and 2220 respondents were used for analysis in Seoul and Tokyo respectively.Data analysis was performed using SPSS 18.0 (IBM co., USA).

Practice Rates of PEBs
The practice rate of each PEB was defined as the percentage of answers that were "always" or "often." A practice rate of more than 50% was observed in out of the 56 PEBs in Seoul and 37 out of the 52 PEBs in Tokyo.In Seoul, the behaviors with high practice rates were "B58: Outdoor incineration" and "B24: Waste disposal by the rules," whereas low rates were observed for "B29: Composting" and "B5: Using stairs."Similarly, in Tokyo, B24 had a high practice rate and B29 had a low rate.While these trends were found in both cities, a difference in the PEB practice rate was also found.For example, "B29: Composting" had similarly low practice rates in Seoul and Tokyo when compared to other PEBs, but the practice rates found in Seoul (21.1%) and in Tokyo (7.0%) were quite different.
To check the significance of the difference between Seoul and Tokyo, the answers were coded on a scale from 4 (for "always") to 1 (for "never"), and statistical tests were conducted.According to the normality test, no behaviors showed normal distributions, and therefore the Mann-Whitney test, one of the non-parametric tests corresponding to the t-test, was adopted.Figure 1 shows the results.In Seoul and Tokyo, 44 out of 50 PEBs showed significantly different practice scores (p < 0.05).

Influence of Surrounding Conditions
Cultural difference is one of the reasons for the difference in practice rates between the two cities.
For instance, the practice rate of "B54: Temperature control off" was significantly higher (p < 0.000) in Tokyo (3.48) than in Seoul (3.02).In Korea, traditional "Ondol" floor-heating is well-established as the system for controlling the temperature of a house.Since this is a central-heating system, people must go to a boiler room to control the room temperature.On the contrary, it is common in Japan for each room to have an individual temperature control system.Therefore, Japanese people can control the room temperature more easily than Korean people.This is reflected in the higher practice rate of B54 in Tokyo than in Seoul.Cultural difference is a fundamental factor and cannot be easily changed by external forces.However, other external factors, such as policies and infrastructure, can be controlled to enhance PEBs.The practice rate of "B26: Recycling second-hand clothes" was significantly higher (p < 0.001) in Seoul (3.20) than in Tokyo (2.19).In Tokyo, when people want to recycle their old clothes, they need bring their clothes to collection places like schools and department stores where local governments or apparel makers collect old clothes.On the other hand, in Seoul, the local government consigns old-clothes collection to private organizations, and collection boxes are placed beside daily curbside waste collection boxes (Figure 2).The higher practice rate of B26 in Seoul can be attributed to this difference.
The effect of policy can also be seen for other PEBs such as "B29: Composting" and "B51: My own bag."Seoul respondents had a significantly higher (p < 0.000) practice score (1.89) for composting behavior (B29) than people in Tokyo (1.29    in landfills used for other combustible waste nationwide.

Behavior
Based on this policy, Korean people have had to separate their food waste from other waste and dispose of it in charged waste bags.This condition creates motivation for people to separate and reduce their food waste.On the other hand, the Tokyo metropolitan government collects and treats food waste along with other combustible waste.Hence, Seoul residents have the habit of separating their food waste and some residents try to reduce the food waste themselves through composting.The policy effect on the behavior practice rate of "B51: My own bag" seems to be large.The Korean government has charged for plastic shopping bags nationwide since 1999, while free plastic shopping bags are distributed in most regions in Tokyo.This is reflected in the significantly (p < 0.001) higher practice rate of B51 in Seoul (3.13) than in Tokyo (2.81).The effects of charging for shopping bags have also been discussed in previous studies.Ueta and Koizumi [24] compared the packing waste systems of Freiburg in Germany and Neyagawa in Japan and found that the charging system in Freiburg contributed to behaviors that reduce the use of disposable plastic shopping bags.Kurisu and Bortoleto [25] compared three mega prefectures (Aichi, Osaka, Tokyo) in Japan and found that the city that implemented a charging system for shopping bags showed significantly higher practice rates of people using their own bags.Hence, the policy of implementing a charging system can have a significant effect on the PEB practice rate.

Influence of Personal Judgment
As discussed above, some external conditions can affect PEB practice rates.However, personal reasons can also influence people's behaviors.To understand the characteristics of PEBs based on personal reasons, we used multiple correspondence analysis (MCA).MCA is one of the statistical techniques used to reveal the optimal quantification that describes the relationships between the categorical scores of each variable as well as the relationships between the variables themselves.The reasons selected from the list shown in Table 2 for practicing or not practicing behaviors were used as the column variables, while 50 PEBs were set as the row variables to give the input matrices for MCA.
The resulting scores were plotted as shown in Figures 3-6.Reasons and behaviors that are placed closer together are more closely connected.
Figures 3 and 4 show the reasons for practicing PEBs in Seoul and Tokyo respectively.Many PEBs were plotted close to the reason "Saving" in both Seoul and Tokyo.Monetary saving is considered one of the important fac-tors for PEBs [3,26].Abrahams et al. [26] reviewed various previous studies related to monetary intervention for household energy conservation and also showed its effectiveness."Saving" is the main reason for practicing most PEBs in the two target cities.If we look at the positions of the 11 reasons, the related reasons of "Habit" and "Convenience" are found close to "Saving."The positions reveal that the respondents in Tokyo perceive "Convenience" and "Habit" as closely connected to "Saving," and that these three reasons are their main reasons for practicing most PEBs, while the respondents in Seoul consider these reasons distinct from "Saving."The reasons that are particularly distinct are "Rule" and "Healthy.""Rule" is closely related to external forces like laws and regulations and so is different from other reasons for practicing PEBs.The behaviors close to the reason "Rule" are "B24: Waste (rule)" and "B25: Recycle (newspaper)" in both cities.These are behaviors practiced following the waste disposal rules set by the local governments."B30: Food waste (dehydration)" is also close to "Rule" in Seoul, because food waste disposal is regulated by the law in Korea, as mentioned in 3.2.1.On the other hand, B30 is more closely related to "Moral" rather than "Rule" in Tokyo, where no regulations about food waste disposal exist.The PEBs closely related to "Healthy" are "B5: Using stairs," "B39: Bicycle or walking," and "B47: Organic goods," all of which are related to health enhancement.In the case of "B6: Filter cleaning," Seoul respondents recognize the health advantage of this behavior, while Tokyo respondents practice this behavior as a "Habit.""Environment friendly" is relatively close to "Moral" in Seoul, while this reason is closely connected to "Following others" and "Recommendation" in Tokyo.Social norms, such as descriptive and subjective norms (i.e., the influence of others), are one of the most important factors in PEBs [7,8,10,[27][28][29]. Ando et al. [7,8] have pointed out that the Japanese are more influenced by others, namely by subjective norms, when practicing PEBs than people in other countries.
Figures 5 and 6 show the reasons for not practicing PEBs in Seoul and Tokyo respectively.The reason that appears most distinct from the others is "Cost."The behaviors that involve purchasing Eco-products, such as "B47: Organic goods," "B50: Eco-car," "B41: Eco-appliance," and "B53: LED," are closely related to this reason in both cities.Eco-products are generally more expensive than regular products, and so "Cost" can be the main barrier to practicing these behaviors.
Unlike Figures 3 and 4, the PEBs are not clustered close to any particular reason, but clusters do form near multiple reasons.This indicates that the internal barriers for PEBs are more diverse than the incentives.Multiple barriers can affect PEBs.This creates difficulties in removing the barriers for the PEBs.

Effects of Demographics
One of the factors that cause practice rate difference is socio-demographics.Gender and age are potentially important factors in determining PEBs practice rates.To understand the effect of gender and age, non-parametric statistical tests such as the Mann-Whitney test and Kruskal-Wallis (KW) test were applied.We selected these non-parametric tests because the original data did not show the normal distribution.However, the parametric test corresponding to the KW test, namely one-way ANOVA (analysis of variance), which is considered to give robust results even without the normal distribution requirement, was also applied.
A post hoc (multiple comparison) test was also conducted to discover which attributes differ in practice rate.Tukey-Kramer Hochberg's GT2 was selected for the behaviors showing homogeneity of variance, while Dunnett's T3 method was adapted for the others.
Tables 3 and 4 show the result of the effects of gender and age on the practice rates in the two cities.In terms of gender, 36 and 50 PEBs show significantly different practice rates between men and women in Seoul and Tokyo respectively.The practice rate of females is statistically higher than that of males in 33 out of 36 PEBs in Seoul and 37 out of 50 PEBs in Tokyo.The higher practice rates of PEBs in women have been reported in many previous studies [5,30,31].It is easily understood that higher practice rates for the female respondents were observed in many homemaking PEBs, such as "B14: Ironing," "B21: Laundry," "B22: Used cooking oil," "B23: Little detergent," "B24: Waste (rule)," and "B25: Recycle (newspaper)." In terms of age, 41 PEBs in Seoul and 32 PEBs in Tokyo show significantly different practice rates.The trend is similar in the two cities.Elder respondents tend to practice PEBs more than younger people.The same trend was also observed in waste prevention behaviors [25,30].As Whitmarsh [32] pointed out, age can be a ignificant variable for several PEBs.The elder respon-s  Note: Significant p value was indicated in bold.a One-tail; b HV: Homogeneity-of variance; c Be reference; d Post-Hoc method, GT2: Tukey-Kramer Hochberg's GT2, T3: Dunnett's T3; e m < k: m is significantly better than k (p < 0.01) m, j < k: There is no significantly difference between m and j, but m and j are significantly better than k (p < 0.01).
dents show higher practice rates in saving energy related to the use of lights, whereas the younger people show relatively higher scores for mobility.Our results also show that younger respondents have higher practice rates for "B40: Public transportation" than elder people in both cities.The previous cross-country studies like Hunter et al. [33] and Aoyagi-Usui et al. [5] imply that the influences of gender and age on behaviors can be different in different countries, but insignificant differences in socio-demographics were observed between Seoul and Tokyo in our study.Comparison within the Asian region appears to yield quite similar trends in basic variables like socio-demographics for the PEBs.

Conclusion
We selected various PEBs and discussed the practice rates and influential factors using Seoul and Tokyo as the case study areas.Some external forces like policy and infrastructure influence the practice rate differences in some PEBs.Relationships between the PEBs and internal incentives and barriers are visualized using MCA.The positions of the reasons to practice or not to practice the PEBs provide us with implications about the common and different perceptions of reasons for practicing/not practicing PEBs in the two cities.Monetary saving is considered to be the most common influential factor for many PEBs.Health is a different type of incentive from other factors and can be an effective factor for some PEBs relating to health improvement.The reasons for some PEBs are quite different for the two cities, indicating that different reasons can increase or suppress these PEBs.Unlike the results of the MCA, the sociodemographic effects on the practice rates show little difference between the two cities.Our study can contribute to the holistic understanding of various PEBs and can be used to consider effective measures to enhance people's PEBs by promoting incentives or removing the barriers for each PEB.

Figure 1 .
Figure 1.Comparison of practice rates between two cities.

Figure 2 .
Figure 2. Old cloth collection boxes in Korea.

Table 1 . Targeted pro-environmental behaviors. Behavior number Pro-environmental behavior Short description
＊*Only in Seoul; + Only in Japan.

Table 2 . Conduct/not conduct reasons.
).In 1998, the Korean government put the "Basic Plan for Recycling Food Waste (1998-2002)" into force and banned the disposal of food waste

Table 3 . The effects of gender and age in Seoul.
Significant p value was indicated in bold.a One-tail; b HV: Homogeneity-of Variance; c Be reference; d Post-Hoc method, GT2: Tukey-Kramer Hochberg's GT2, T3 : Dunnett's T3; e m < k: m is significantly better than k (p < 0.01) m, j < k: There is no significantly difference between m and j, but m and j are significantly better than k (p < 0.01).