Determinants of household waste recycling behavior the case of ho chi minh city

ABSTRACT This paper investigates determinants of household waste recycling behavior for six different materials: paper, carton, plastic, metal, glass and cloth using data from the survey

UNIVERSITY OF ECONOMICS INSTITUTE OF SOCIAL STUDIES

HO CHI MINH CITY THE HAGUE

VIETNAM THE NETHERLANDS

VIETNAM - NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS

DETERMINANTS OF HOUSEHOLD WASTE

RECYCLING BEHAVIOR THE CASE OF HO CHI MINH CITY

BY

PHAN BÙI KHUÊ ĐÀI

MASTER OF ARTS IN DEVELOPMENT ECONOMICS

HO CHI MINH CITY, NOVEMBER 2015

UNIVERSITY OF ECONOMICS INSTITUTE OF SOCIAL STUDIES

HO CHI MINH CITY THE HAGUE

VIETNAM THE NETHERLANDS

VIETNAM - NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS

DETERMINANTS OF HOUSEHOLD WASTE

RECYCLING BEHAVIOR THE CASE OF HO CHI MINH CITY

A thesis submitted in partial fulfilment of the requirements for the degree of

MASTER OF ARTS IN DEVELOPMENT ECONOMICS

By

PHAN BÙI KHUÊ ĐÀI

Academic Supervisor:

PROF NGUYỄN TRỌNG HOÀI

HO CHI MINH CITY, NOVEMBER 2015

ABSTRACT

This paper investigates determinants of household waste recycling behavior for six different materials: paper, carton, plastic, metal, glass and cloth using data from the survey “Consumption behavior towards green growth in urban area of Viet Nam”, funded by National Foundation for Science and Technology Development (NAFOSTED) and logistic regression My findings reveal that psychological factors towards recycling generally appear to be statistically insignificant Nevertheless, concern about waste, awareness of inheritance for future generation and satisfaction

of waste condition at household’s residency explain their recycling behavior for some materials Another interesting finding is that the household’s belief of money gained from waste recycling will lead to recycling paper, carton and plastic Furthermore, the results disclose that income and age in some cases affect the recycling behavior negatively Moreover, housing characteristics have impact on recycling behavior positively in the case of metal, carton and paper recycling

Key words: household waste recycling, recycling behavior, logistic model

ACKNOWLEDGMENT

Firstly, I would like to express my sincere gratitude to my supervisor Prof Nguyen Trong Hoai who provided me with precious data set from the survey

“Consumption behavior towards green growth in urban area of Viet Nam”, funded

by National Foundation for Science and Technology Development (NAFOSTED) and gave me valuable guidelines, comments and suggestions for the successful completion of this study

I would like to express my special appreciation to Dr Truong Dang Thuy whom I have learned a lot from his enthusiastic guidance, useful recommendations and inspiration Besides, his friendly and inspiring approach has given me a great deal of encouragements to overcome difficulties in the whole research process

I would like to express my thanks to Dr Nguyen Hoang Bao and MA Phung Thanh Binh for introducing the logistic regression model at “Data analysis and forecasting” course on December, 2014, hold by the Faculty of Development Economics of HCMC, University of Economics

I am also thankful to all lecturers and program administrators of the Vietnam – The Netherlands Program for M.A in Development Economics They have given

me wonderful knowledge and helped me kindly during the course

To all my friends in MDE Class 16, 19 and 20, especially MA Nguyen Van Dung (Class 16) and MA Nguyen Quang Huy (Class 19) who gave me emotional encouragement, I would like to express my sincere thanks

Finally, I would like to express my deeply appreciation to my family and Mr Tran Kim Minh for spiritual support and love In particular, I dedicate this thesis to

my beloved farther, Mr Phan Van Bay

HCMC, November 2015 Phan Bui Khue Dai

TABLE OF CONTENTS

LIST OF TABLES v

LIST OF FIGURES vi

Chapter 1: INTRODUCTION 1

1.1 Problem statement 1

1.1.1 Real world problem 1

1.1.2 Scientific problem 1

1.2 Research objectives 2

1.3 Research questions 2

1.4 Research scope and data 2

1.5 The structure of this study 2

Chapter 2: LITERATURE REVIEW 4

2.1 Theoretical Review 4

2.2 Empirical Review 5

2.2.1 Socio-Economic and Demographic characteristics 5

2.2.2 Housing characteristics 9

2.2.3 Psychological factors towards recycling 10

Chapter 3: RESEARCH METHODOLOGY 15

3.1 Conceptual framework and the econometric model 15

3.2 Data source 20

3.3 Methodology 20

Chapter 4: EMPIRICAL RESULTS 22

4.1 Descriptive Statistics 22

4.1.1 Dependent variables 22

4.1.2 Independent variables 22

4.2 Bivariate analysis 24

4.2.1 Metal recycling 24

4.2.2 Carton recycling 27

4.2.3 Paper recycling 29

4.2.4 Plastic recycling 32

4.2.5 Glass recycling 34

4.2.6 Cloth recycling 36

4.3 Regression results 38

Chapter 5: CONCLUSION AND POLICY RECOMMENDATION 43

5.1 Conclusion 43

5.2 Policy recommendation 44

5.3 Research limitation 44

REFERENCES 45

APPENDICE 48

LIST OF TABLES

Table 1 Description of the variables 16

Table 2 Prevalence of recycling toward six materials 22

Table 3 Descriptive statistics of numerical variables 23

Table 4 Descriptive statistics of binary variables 24

Table 5 A comparison between metal recycling and non-recycling in term of income, age and size of housing 24

Table 6 A comparison between metal recycling and non-recycling in term of gender, type of housing and education 25

Table 7 A comparison between carton recycling and non-recycling in term of income, age and size of housing 27

Table 8 A comparison between carton recycling and non-recycling in term of gender, type of housing and education 27

Table 9 A comparison between paper recycling and non-recycling in term of income, age and size of housing 29

Table 10 A comparison between paper recycling and non-recycling in term of gender, type of housing and education 30

Table 11 A comparison between plastic recycling and non-recycling in term of income, age and size of housing 32

Table 12 A comparison between plastic recycling and non-recycling in term of gender, type of housing and education 32

Table 13 A comparison between glass recycling and non-recycling in term of income, age and size of housing 34

Table 14 A comparison between glass recycling and non-recycling in term of gender, type of housing and education 34

Table 15 A comparison between cloth recycling and non-recycling in term of income, age and size of housing 36

Table 16 A comparison between cloth recycling and non-recycling in term of gender, type of housing and education 36

Table 17 Parameter estimates for the logit models 39

LIST OF FIGURES

Figure 1 Conceptual framework 15 Figure 2 A comparison between metal recycling and non-recycling in term of some selected psychological factors 26 Figure 3 A comparison between carton recycling and non-recycling in term of

some selected psychological factors 28 Figure 4 A comparison between paper recycling and non-recycling in term of some selected psychological factors 31 Figure 5 A comparison between plastic recycling and non-recycling in term of

some selected psychological factors 33 Figure 6 A comparison between glass recycling and non-recycling in term of some selected psychological factors 35 Figure 7 A comparison between cloth recycling and non-recycling in term of some selected psychological factors 37

Chapter 1: INTRODUCTION

1.1 Problem statement

1.1.1 Real world problem

In Vietnam, the household waste is a pressing environment issue in recent years Ministry of Natural Resources and Environment (2014) indicated that total solid waste from household was 12.8 million tons per year and would reach to 22 million per year by 2020 Increasing amounts of household waste may lead to the risk of heavy environmental pollution and the seriously impact on public health In the general context, Ho Chi Minh City is facing many challenges in waste management Every day, the city has more than 7,000 tons of garbage and costs each year up to

235 billion VND to handle Thus, in order to recommend government in the design

of effective policies to minimize waste, it is necessary to understand household waste recycling behavior The study analyzes the determinants of household waste

recycle behavior with the scope of research in Ho Chi Minh City

1.1.2 Scientific problem

Currently, there are few studies on the issue of household behavior in Vietnam, mainly towards green consumption (water use; energy use; recycling and transport choice) Of which, there is only one study of Luu Bao Doan and Nguyen Trong Hoai (2015) on household waste recycling behavior Using structural equation modeling, this study indicated that recycling related to the attitude of the household towards recycling Further, general concern and knowledge of environment do not have direct relations with the behavior of interest to recycle

Therefore, the study with logit models may contribute to academic knowledge on recycling behavior in the context of a developing country like Vietnam Furthermore, this study can provide policy makers with a new perspective on the nature of recycling behavior, the relationship between this behavior and such factors

as psychological factors of the environment in general and waste recycling in particular, so that the authority can consider appropriate tools to adjust citizen's behavior towards recycling more

1.2 Research objectives

The objective of the study is to identify factors affecting household’s waste recycling behavior in Ho Chi Minh City towards six different materials: metal, carton, paper, plastic, glass and cloth

1.4 Research scope and data

The study will investigate determinants of household waste recycling behavior for 6 different materials: paper, carton, plastic, metal, glass and cloth by using data from the survey “Consumption behavior towards green growth in urban area of Viet Nam”, funded by National Foundation for Science and Technology Development (NAFOSTED) The survey was conducted in Ho Chi Minh City on April and May

2014, including 200 households from District 1, 3, 4, 9, Binh Thanh, Go Vap, Phu Nhuan and Thu Duc To collect data, investigators directly contacted the household head for an interview, clearly explained the questions and choices, then recorded the respondents' feedback

1.5 The structure of this study

Five chapters will be constructed in this study as follows:

Chapter 1: Introduction This is the beginning section of thesis, which consists of

the research topic and problem statement The research objectives, research questions and the research scope and data are also presented in this chapter The final section will provide the structure of the research

Chapter 2: Literature review This chapter provides with theoretical and empirical

reviews related to household waste recycling behavior The first section will review

a related theory explaining recycling behavior, which is Random Utility Theory The empirical researches indicate that there are three main groups of factors that affect individual’s behavior of recycling They are socio-economic and demographic characteristics, housing characteristics and psychological factors towards recycling Furthermore, these factors impact positively or negatively on recycling behavior depending on the circumstances and the material as well

Chapter 3: Methodology This chapter presents research methods and conceptual

framework

Chapter 4: Empirical results This chapter starts with descriptive statistic and then

provides a bivariate relationship between household recycling behavior and some important determinants Finally, the regression results and interpretation are presented

Chapter 5: Conclusions This chapter summarizes the findings and concludes with

some policy implication and research limitations

Chapter 2: LITERATURE REVIEW

This chapter provides with theoretical and empirical reviews related to household waste recycling behavior The first section will review random utility theory The empirical researches indicate that there are three main groups of factors that affect individual’s behavior of recycling They are socio-economic and demographic characteristics, housing characteristics and psychological factors towards recycling Furthermore, these factors impact positively or negatively on recycling behavior depending on the circumstances and the material as well

2.1 Theoretical Review

A related theory explaining recycling behavior is Random Utility Theory by Marschak (1960) This theory indicated that an individual utility could be defined

by two components: deterministic components and random components So, the

total utility of a household i recycling waste are the sum of the two utility

components:

U1i = V1i + e1i Where, V1i can be approximated by a linear function of recycling in the vector of Xi and the population utility weights for each attribute in the vector i : V1i = 1iX1i In additional, e1i is random utility component

Similarly, the total utility of a household i with non – recycling:

U0i = V0i + e0i Where, V1i can be approximated by a linear function of recycling in the vector of Xi and the population utility weights for each attribute in the vector i : V0i = 0iX0i In additional, e0i is random utility component

The probability that a household recycles can be expressed as the probability that the utility associated with recycling is higher than the utility of non – recycling:

Pr(recycling) = Pr (U1i > U0i)

Or Pr(recycling) = Pr (V1i + e1i > V0i + e0i) = Pr (e1i - e0i > V0i - V1i)

Or Pr(recycling) = Pr (e1i - e0i > 0iX0i- 1iX1i)

We assume that error terms of alternatives do not correlate with each other and they have the same variance and follow logistic distribution In this case the probability that a household chooses to recycle is a logit probability:

e recycling

household waste recycling behavior, namely: (i) Socio-economic and demographic

characteristics, (ii) Housing characteristics and (iii) Psychological factors towards

recycling (Ferrara and Missios, 2012; Fiorillo, 2013; Nixon et al., 2009; Hage et al.,

2009; Afroz et al., 2011; Dalen et al., 2011)

2.2.1 Socio-Economic and Demographic characteristics

Household Income

Income which is studied in most of researches relating to recycling behavior plays a crucial role in affecting household recycling behavior of both household and individual Many studies have shown a significant positive relationship between higher household income and recycling behavior (Halvorsen, 2008; Lee et al., 2011;

Ferrara and Missios, 2012; Fiorillo, 2013; Halvorsen, 2012; Hui Zhao et al., 2013) Halvorsen (2008 and 2012) surveyed 1.162 and 10.251 households respectively from Norway and 10 various OEDC countries in 2008 and indicated that the higher income respondents have, the more likely they are to recycle household waste In

2011, Lee et al showed that higher income creates a higher incentive for households to participate in both food separation and recycling through a survey including 196 responses conducted in Seoul, Korea Ferrara and Missios (2012) used the sample of 10.251 respondents from 10 OEDC countries to conclude that very richer households or individuals are more likely to take part in glass recycling

as well as recycle at a larger ratio of glass, plastic and aluminum One year later, Fiorllo (2013) conducted an empirical study of 47.643 households and found out that people with high income tend to recycle all materials included paper, glass, plastic and aluminum excepting for food waste In the meanwhile, Hui Zhao et al (2013) researched a data of 500 questionnaires collected in Qingdao to prove the positive correlation between higher income and recycling behavior Moreover, analyzing a data of 402 households in Dhaka City, they summarized that the frequency of solid waste recycling behavior is influenced positively by the middle

of income group These results implied that households on low and middle income tend to take advantage of all materials to minimize the cost of buying new things Several other researchers have suggested a negative or insignificant correlation between income and household waste recycling behavior (Hage et al., 2008; Shaufique et al., 2010) Using a sample of 2800 household members in 4 different Swedish municipalities in 2006, Hage et al (2009) evaluated that income is not a determinant of household recycling behavior of packaging waste which is similar to the finding of Nixon et al (2009) that examined households’ attitudes towards recycling On the other hand, Rafia et al (2011) did not see the correlation between income and recycling behavior in the direct manner More interestingly, Shaufique et al (2010) demonstrated that income has a negative impact on residential recycling rate per annum Based on a data set of 774 respondents representing 86 counties in Minnesota, the study of Shaufique et al (2010) calculated that if annual income per capita increase 1000 dollar then there will be a

0.2 percentage point decline in the rate of recycling It was also an average income group that raises the likelihood of reaching the aim of 150kg/capita residual

household waste of each municipality (Gellynck et al., 2011)

Gender

There are a great number of studies citing gender as a determinant of household waste recycling behavior (Sterner and Bartelings, 1999; Saphores et al., 2012; Fiorillo, 2012; Ferrara and Missios, 2012) Sterner and Bartelings (1999) analyzed nearly 600 samples in Tvaaker which were gathered in 1994 to assess respondents’ willingness to pay for the attention of the waste and recycling issue The variable gender has a significant indication, which means that women are more probable to pay and pay more than men Similarly, Saphores et al (2011) concerned that women are easier to engage in e-waste recycling and be willing to recycle e-waste at drop-off center than men after testing a data of 2136 households from a 2006 national survey of US Noticeably, Ferrara and Missios (2012) and Fiorillo (2013) proved the significant positive sign between gender and household recycling behavior of five following materials: glass, plastic, paper, food and aluminum waste However, whilst Ferrara and Missios suggested that men are willing to recycle and recycle more aluminum than women, Fiorillo found out that women are ready to recycle all materials than men These findings implied that the gender’s impact on recycling trend maybe dependent on kind of waste materials as well as

survey area

In contrast, basing on a survey of 10.000 households from 10 various OECD countries, Dalen and Halvorsen (2011) indicated that women do not intend to embark on recycling activities Besides, the gender factor has also been examined in

a lot of researches, but the correlation is insignificant or very slight (Hage et al., 2009; Lee et al., 2011; Pakpour et al., 2013; Ayalon et al., 2013; Huffman et al.,

2014)

Education

Many sudies showed that the role of education in recycling behavior is very diferrent Some scientific researches recognized the positive effect of education level on practicing recycling behavior (Amy W and Gosselin, 2005; Shaufique,

V.Joshi, Lupi, 2010; Dwivedy and Mittal, 2013; Pakpour et al., 2013; Lange et al., 2014) Ando (2005) used the probit and double-censored tobit model to examine recycling rates of 214 multifamily dwellings in Urbana, Illinois before concluding that number of years of education present a positive correlation with container recycling rate Significantly, Anderson et al (2013) applied the logistic regression model in order to analyze a sample from the 2003 - 2006 and suggested education level of household head plays a crucial role in making the decision to recycle Education also showed a negative impact on recycling behavior for monetary reasons

On the other hand, Sterner and Bartelings (1999) found a negative correlation of age with household’s willingness to pay for caring for recycling issues which imply that people with less education are willing to pay more Similarly, Hage et al (2009) showed that education has a negative influence on paper recycling However, some empirical studies investigated that there is no impacts of education on recycling behavior (Nixon et al., 2009; Lee et al., 2011; Ferrara and Missios, 2012; Byrne and O’regan, 2014; Hui Zhao et al., 2013)

Age

Almost recent empirical studies have presented a positive correlation between age and household waste recycling behavior (Ando et al., 2005; Shaufique et al., 2010; Saphores et al., 2012; Pakpour et al., 2013; Hui Zhao et al., 2013; Lange et al., 2014) As for Hage et al (2009), age was deemed to be a positive determinant for all packaging materials recycled by individuals and households In the same pattern, Ayalon, Sharon and Shechter (2013), through a survey of 12.000 households in

2011, proved age influence positively on household separation and recycling behavior except for food and garden Whilst Lee et al (2011) presented the importance of individuals being older in the Korean household recycling behavior, Dalen et al (2011) mentioned older women are easier to intend to recycle Afroz et

al (2011) and Fiorillo (2013) also found that respondent from 25 to 35 or from 51

to 60 years old will tend to engage in recycling activities for all materials

However, a few recent studies showed negative relationship between age and household waste recycling According to Ferrara and Missios (2012), the younger

are less likely to recycle except plastic waste Anderson et al (2013) stated that age

of non – African household head has a negative effect on their recycling behavior While Sterner and Bartelings (1999) also demonstrated an inverse relationship between age and recycling, there are many researches not finding any connection of two above factors (Nixon et al.,2009; Byrne and O’regan, 2014)

2.2.2 Housing characteristics

House type

The relationship between home type with the intention of household recycling behavior has been examined Ando and Gosselin (2005) indicated that multifamily dwellings living in the apartment having adequate storage space will achieve higher recycling rate According to Barr (2007), home type was deemed to be positively impact on recycling intention through applying the standardized regression model to analyze a data of 673 respondents of Exeter Dalen and Halvorsen (2011) realized that women living in detached house will be more likely to recycle In 2012, Halvorsen also found the similar impact of detached house on the number of materials recycled In another recent study, Dwivedy and Mittal (2013) illustrated that among 150 respondents from India who living in an apartment will be more willing to take part in e-waste recycling

In contrast, using ordered probit analysis, Hage et al (2009) proposed that household living in an apartment tend to recycle less metal waste than others Moreover, Ferrara and Missios (2012) indicated that household living in a detached

or semi-detached is less probable to recycle There are a few of the studies not showing any relationship between two above variables (Nixon et al.,2009; Byrne and O’regan, 2014)

out that size of residence will impact positively on household recycling and waste prevention The more rooms household owns, the more probable it is to engage in recycling all materials except for plastic and aluminum waste Afterwards, Fiorillo (2013) suggested that people living in a house having from 1 to 5 rooms promote the likelihood of recycling all materials except for food waste However, another empirical investigation did not propose any relationship between home size and

recycling behavior (Afroz et al., 2011)

2.2.3 Psychological factors towards recycling

With reference to empirical investigations, there is a great deal of studies depicting the effect of psychological factors towards recycling behavior The first and foremost one would be Sterner and Bartelings (1999) suggesting that people showing their concern toward the importance of waste are easier to pay for taking care of household waste, whereas, respondents reporting difficulty of recycling different materials are less likely to pay attention to recycling problem together with recycling behavior

Regarding Bruvoll et al (2002), it was clear that there is an intense positive relationship between sorting behavior and households’ attitudes toward sorting After examining psychological factors of 1162 interviewees in 1999, namely: Perceiving sorting as mandatory, environmental considerations and moral requirements, the authors pointed out that individuals concerning more about natural habitats as their own and civic responsibility will definitely contribute to classifying behavior

In reference to Barr (2007), the most striking finding was that individuals who considered recycling to be the norm of the society are more probable to recycle than others It was also assumed that recycling willingness and intensity are able to be stimulated by providing households convenient recycling facilities For instance, if respondents perceive the presence of drop-off locations or feel easy to access curbside recycling services, it is certain that residential recycling rate per annum will increase substantially (Shaufique et al., 2010, X Gellynck et al., 2011) Furthermore, the habit of recycling organic waste encouraged municipalities to minimize their domestic waste

In 2008, Halvorsen applied ordinary least square estimation (OLS) to evaluate the number of fractions recycled by the household and pointed out that belief of recycling is the strongest factor In another later study, Halvorsen and Dalen (2011) used the same dependent variable is the number of materials recycled by the household This study described that women’s recycling efforts are more encouraged by their concern about waste generation, whilst men respond more to the beliefs of better environment Furthermore, Halvorsen (2012) illustrated that people in an environmental organizations will tend markedly to recycle more than others In addition, respondent concerned about waste generation, water pollution and believing of recycling as their civic duty is more probable to recycle However,

in 2012, Halvorsen also indicated that people who have concern about climatic change will concentrate their efforts on other green-friendly activities rather than recycling

According to Nixon et al (2009), respondents will recycle more if it is believed that recycling behavior is the determinant of reducing the use of landfills significantly and conserving natural resources effectively Moreover, households’ recycling participation was also proven to arise from environmental benefits rather than economic benefits Noticeably, this study demonstrated that people concurring with how internal values and morals affecting pro-environmental behavior or feeling a moral obligation to recycle are willing to recycle and recycle more by 7.2 times than those dissenting with the above statement Interestingly, the results of the study also showed a positive relationship between information sources and how information influences the intention to recycle, namely: respondents who receive recycling information from print and family/friends, print and work/school are more likely to recycle by 7.3 and 5.0 times respectively Also, they implied that the more information resources households receive, the more likely they are to engage in recycling behavior

In 2009, Hage et al showed a highly positive correlation between moral obligation and recycling behavior which means that households who are aware of their personal responsibility will do recycle all materials, particularly paper, glass and metal It is assumed that the perception of others’ recycling effort, concern about the

environmental impacts of waste disposal, beliefs of their recycling efforts are crucial incentives for respondents to engage in recycling, too In other words, individuals who perceive the negative externality arising from desist from recycling will be willing to recycle more their packing waste, especially plastic materials Conversely, households who are not aware of the implication of their throwing away packaging waste with the environment seem not to engage in recycling efforts

Another recent study, Lee et al (2011) applied the new environmental paradigm (NEP) index to analyze the respondent’s environmental attitudes and behavior According to that, people showing higher concerns about the environment implied higher degrees of recycling participation Moreover, household’s attitude, higher agreement and interest in waste management had a significant positive impact on food classifying and recycling behavior

Afterwards, Afroz et al (2011) showed a positive and correlation between the household’s attitude and recycling behavior Specifically, environmental awareness, willingness to separate and minimize the household waste and respondent’s belief toward solid waste management practices are the determinants encouraging them to produce less waste and recycle more

Besides that, Bao (2011) analyzed a sample of 1523 students in Turku through Pearson chi-square statistic and suggested positive, relevant implication of recycling and psychological factors, namely: respondents reported higher concern about sustainable development and believing Recycling assists to conserve the environment will show higher degrees of desire toward recycling Additionally, individuals accepting moral norms as well as considering recycling as their responsibility are more willing to take part in recycling Finally, recycling behavior

is able to be stimulated by giving individuals more information on waste separation, making recycling more convenient and guiding them the destination of separated waste

In 2011, the report of OECD about green household behavior showed availability of convenient recycling service and the characteristics of recycling collection services also impact positively on recycling behavior Curbside recycling facility and drop-

off system increase aluminum recycling rate by 34 percent This scientific source also proposed that recycling intention and participation for glass, plastic and aluminum are promoted markedly by individuals’ environmental attitudes Recycling programs are also cited as positive factors impacting on the likelihood of household recycling decision, particularly aluminum waste

One year later, Saphores et al pointed out moral beliefs are considered as the most statistically significant variables toward recycling behavior followed by e-waste recycling convenience and mandatory recycling at work or school respectively In other words, households perceiving their responsibility as voluntary or mandatory and being aware of ease of recycling will tend to recycle e-waste more considerably than others Furthermore, recycling participation can be encouraged by providing households knowledge about the potential of danger of e-waste

In addition, Ferrara and Missios (2012) found evidences proving that respondents reported higher concern about environmental problems are at higher levels of participation of glass, plastic, aluminum and food waste recycling Besides, the individuals’ environmental attitude variables are assumed to be the determinants promoting glass, plastic and aluminum waste recycling Whether and what extent to which recycling is considered to be beneficial for the environment need taking into account as positive factors toward recycling behavior

In 2013, Fiorillo indicated that respondents stating no dirtiness problems at the residence will increase the probability of recycling glass, paper and plastic waste If

a household claims that there is no pollution around area they live, his/her likelihood of recycling is lower than others, especially for paper, plastic and aluminum waste Conversely, if an individual perceives the habitat being clean day

by day, he/she then tries to keep and encourage recycling behavior continually Moreover, Pakpour et al (2013) used data from 1782 households of 8 urban health centers in the Qazvin city to emphasize the importance of individuals’ attitude, subjective rules, perceived behavioral control, moral obligation and self-cognition towards waste recycling behavior According to that, households who reported higher concern to these above variables will stimulate recycling considerably

Ayalon et al (2013) analyzed the survey of OECD and suggested that households reported higher concern about environmental issues are less likely to refuse to participate in waste recycling This study stressed the positive impacts of environmental motivations toward recycling in all countries surveyed except for Israel citizen Obviously, being aware of the recycling service available like drop-off center will encourage individuals to recycle more

Moreover, Anderson et al (2013) investigated the decision to recycle among urban South Africans and demonstrated that households who see littering as a community issue are more willing to recycle than others do not The two main factors, including greater awareness of environmental concern and greater access to recycling facilities play an important role in mitigating household waste and rising waste recycling rate Hui Zhao et al (2013) also stated that environmental concern had a positive effect on recycling behavior

In 2014, Byrne and O’regan have constructed statements to evaluate the reason why

or why not recycling was considered as a daily routine Statistics showed that almost respondents had positive recycling habits and individual’s recycling efforts would create a significant incentive for others to recycle In some cases, households suggested that they will tend to recycle if received more information and knowledge about recycling facilities supplied by waste collectors In another latest study, Huffman et al (2014) supported the view that there is a strict interaction between social norms and recycling attitudes as well as self-reported recycling behavior In other words, individuals who perceive a high social effect are more likely to announce that they will engage in waste recycling

The appendix 1 provides the summary of above empirical studies on household waste recycling behavior in a more visual way

Chapter 3: RESEARCH METHODOLOGY

This chapter presents research methods and conceptual framework

3.1 Conceptual framework and the econometric model

Conceptual framework

Based on the theoretical and empirical reviews, a conceptual framework is built as the Figure 1 following The empirical researches indicate that there are three main groups of factors that affect individual’s behavior of recycling They are socio-economic and demographic characteristics, housing characteristics and psychological factors towards recycling

Figure 1 Conceptual framework

The econometric model

Using Stata, six logistic regression models are estimated for household waste recycling behavior towards six different materials: metal, carton, paper, plastic, glass and cloth Household recycling behavior is investigated through the question:

“Which of the following materials does your family usually recycle or collect for vendors?” The possible answer to each material is “yes” or “no” The response for

 Concern about waste

 Household’s awareness of impacts

that material is coded into a binary variable which means 1 in case of “yes” and 0 otherwise

Table 1 Description of the variables

Dummy variable, = 1 if household recycles containers and utensils made of iron, steel, stainless steel, aluminum, = 0 otherwise

recycles glass objects, = 0 otherwise

Dummy variable, = 1 if household recycles paper and newspaper, = 0 otherwise

recycles carton, = 0 otherwise

recycles old clothes, = 0 otherwise

Independent variables

Socio-Economic and Demographic Characteristics

female)

3 education

Education of the respondent (1= completed vocational school, college, university; 0 = no education, completed elementary school, secondary school, high

school)

month (in million dongs)

Housing characteristics

semi-detached house; 0 = apartment)

Psychological factors towards Recycling

“Concerned” or “Very Concerned”

Dummy variable, = 0 if the respondent’s awareness degree of environmental impacts on human life “No idea” or

“Disagree” or “Fairly Agree”; = 1 if the respondent’s awareness degree of environmental impacts on human life

“Agree” or “Strongly Agree”

9 inheritance

Dummy variable, = 0 if the respondent’s awareness degree of environmental impacts on future generation “No idea” or

“Disagree” or “Fairly Agree”; = 1 if the respondent’s awareness degree of environmental impacts on future generation “Agree” or “Strongly Agree”

awareness degree of environmental

impacts on longevity “No idea” or

“Disagree” or “Fairly Agree”; = 1 if the respondent’s awareness degree of environmental impacts on longevity

“Agree” or “Strongly Agree”

11 tradeoff

Dummy variable, = 0 if the respondent’s willingness degree to trade-off to protect the environment “No idea” or “Disagree”

or “Fairly Agree”; = 1 if the respondent’s willingness degree to trade-off to to protect the environment “Agree” or

“Disagree” or “Fairly Agree”; = 1 if the respondent’s willingness degree to save energy to protect the environment “Agree”

or “Strongly Agree”

13 water

Dummy variable, = 0 if the respondent’s willingness degree to save water to protect the environment “No idea” or “Disagree”

or “Fairly Agree”; = 1 if the respondent’s willingness degree to save water to protect the environment “Agree” or “Strongly Agree”

14 waste

Dummy variable, = 0 if the respondent’s willingness degree to treat waste to protect the environment “No idea” or “Disagree”

or “Fairly Agree”; = 1 if the respondent’s

willingness degree to treat waste to protect the environment “Agree” or “Strongly Agree”

15 transportation

Dummy variable, = 0 if the respondent’s willingness degree to limited use of personal vehicles to protect the environment “No idea” or “Disagree” or

“Fairly Agree”; = 1 if the respondent’s willingness degree to limited use of personal vehicles to protect the environment “Agree” or “Strongly Agree”

16 waste_condition

Dummy variable, = 0 if the respondent’s satisfaction degree of waste condition at residency “No idea” or “Dissatisfied” or

“Fairly Satisfied”; = 1 if the respondent’s the respondent’s satisfaction degree of waste condition at residency “Satisfied” or

“Strongly Satisfied”

Dummy variable, = 0 if the respondent’s awareness degree of economic benefits from recycling “No idea” or “Disagree” or

“Fairly Agree”; = 1 if the respondent’s awareness degree of economic benefits from recycling “Agree” or “Strongly Agree”

With the variables and measurements above, the proposed functional form is:

1

i i

P P

 = α + iSEDCi + iHCi+  PFTRi + ui

In particular, Pi is probability of recycling; SEDC is the vector of explanatory variables indicating Socio-Economic and Demographic Characteristics including Age, Gender, Education, Household Income; HC is vector of explanatory variables indicating Housing characteristics including House size and House type; PFTR is the vector of explanatory variables referring to Psychological factors towards Recycling including household’s concern about waste; awareness of impacts on environment; willingness to protect environment; satisfaction of waste condition at residency; belief of money gained from waste recycling

3.2 Data source

Data used in this study is obtained from the survey “consumption behavior towards green growth in urban area of Viet Nam”, funded by National Foundation for Science and Technology Development (NAFOSTED) The survey was conducted in

Ho Chi Minh City on April and May 2014, including 200 households from District

1, 3, 4, 9, Binh Thanh, Go Vap, Phu Nhuan and Thu Duc To collect data, investigators directly contacted the household for an interview, clearly explained the

questions and choices, then recorded the respondents' feedback

3.3 Methodology

I am concerning about what factors affecting behavior of a household to recycle or not recycle toward six different materials Hence, dependent variable is a binary outcome Two standard models are often used are the probit model and the logit model Many papers choose logit model because it has mathematical simplicity (Gujarati, 2003) In the thesis, logit model is employed

The logit model:

In the logit model, ui has logistic distribution The probability density function of ui

i i

i

u u

e





Maximum Likelihood Estimation (MLE) is using to estimate the model It can be

verified the cumulative distribution function (CDF) of ui is  

1

i i

i

u u

Chapter 4: EMPIRICAL RESULTS

This chapter starts with descriptive statistic and then provides a bivariate relationship between household waste recycling behaviors toward six different materials: metal, carton, paper, plastic, glass and cloth and some important determinants Finally, the regression results and interpretation are presented

4.1 Descriptive Statistics

4.1.1 Dependent variables

Table 2 shows the household percentage of recycling toward six materials of metal, carton, paper, plastic, glass and cloth based on 200 observations surveyed in Ho Chi Minh City In which, paper and plastic are the most recycled with the percentage of 73.5% and 70.5% respectively The next is metal recycling at 56% of the households In addition, it can be seen that a moderate percentage of households is also interested in carton recycling This proportion obtains 46.5% The least likely recycled materials are glass and cloth with only 23.5% and 32.5% respectively

Table 2 Prevalence of recycling toward six materials Dependent variables Observations Frequencies

includes age (in years), gender (male coded =1; female coded = 0), education (high school or lower coded = 0, higher education coded = 1) and ln_income implying

age, gender, education and income of households (in million dongs) Housing

characteristics consists of house_type which arguing detached and semi-detached house (coded = 1) or apartment (coded = 0) and house_size which expresses size of

house in squared meter In addition, psychological factors cover 11 dummy

variables including concern (meaning household’s concern about waste); life,

inheritance, longevity (referring to household’s awareness of impacts on

environment); tradeoff, energy, water, waste, transportation (implying household’s willingness to protect environment); waste_condition (referring to household’s satisfaction of waste condition at residency); money (indicating household’s belief

of money gained from waste recycling) In which, artificial variables take on the value 1 indicating the presence of the attribute and 0 indicating absence of the attribute

Table 3 Descriptive statistics of numerical variables Variables Mean Std Dev Min Max Observations

Source: Author’s calculation using Stata from NAFOSTED’s Data

The Table 3 and Table 4 summarize the results of the survey There is only a few missing values Overall, the percentage of households presenting their concern about waste, awareness of impacts on environment, willingness to protect environment and belief of money gained from waste recycling obtains over 60% Moreover, the average age of the people surveyed is nearly 44 years and their average income gets nearly 16 million dongs In which, 41.5% of household heads are male and 41.71% of them have finished high school In addition, around 77% of households live in detached or semi-detached house and the mean of housing size is about 55 in squared meter

Table 4 Descriptive statistics of binary variables Variables Frequencies

Source: Author’s calculation using Stata from NAFOSTED’s Data

The next section provides the bivariate relationship between recycling behavior of household toward the six kinds of waste above and its important determinants

4.2 Bivariate analysis

4.2.1 Metal recycling

Table 5 A comparison between metal recycling and non-recycling in

term of income, age and size of housing Variables Recycling Non-Recycling P-value of difference in mean

Source: Author’s calculation

The t-test results of Table 5 shows that there is statistically significant difference in house size and age between the recycling and non-recycling groups The bivariate

relationship is statistically significant at 1% and 10% respectively (P-value of difference in mean = 0.000; 0.096 respectively) However, there is not much difference in mean of income between people recycling and non-recycling The bivariate relationship is not statistically significant at 10% (P-value of difference in mean = 0.227)

Table 6 A comparison between metal recycling and non-recycling in

term of gender, type of housing and education

(%)

Non-Recycling (%)

P-value of square test

high school or lower 58.04 58.62

Source: Author’s calculation

The Table 6 results show that household gender is not associated with metal recycling Among 112 people recycling metal, 41.07% are male and 58.93% are female Likewise, in the remaining group, 42.05% are male and 57.95% are female This bivariate relationship is not statistically significant at 10% (Pearson chi2 = 0.019, p-value = 0.89) Housing characteristics are not correlated with recycling behavior Among the group of recycling, there is 77.27% people living in detached and semi-detached and 22.73% living in apartment Likewise, in non-recycling group, 76.14% of people is living in detached and semi-detached and 23.86% living

in apartment This bivariate relationship is not statistically significant at 10% (Pearson chi2 = 0.035, p-value = 0.851) The chi-square test results also indicates that academic level is not associated with metal recycling Between the two group

of recycling and non-recycling, household with higher education are about 41% This bivariate relationship is not statistically significant at 10% (Pearson chi2 = 0.006, p-value = 0.934)

Figure 2 A comparison between metal recycling and non-recycling in

term of some selected psychological factors

Figure 2 shows a comparison between metal recycling and non-recycling in term of some selected psychological factors There is not much disparity between the two groups about concern about waste and awareness of impacts of environment on life However, household have the tendency to recycle metal if their awareness of inheritance or awareness of money they gained from recycling is presented

4.2.2 Carton recycling

Table 7 A comparison between carton recycling and non-recycling in

term of income, age and size of housing Variables Recycling Non-Recycling P-value of difference

in Mean

Source: Author’s calculation

The t-test results of Table 7 show that there is statistically significant difference in house size and age between the two recycling and non-recycling groups The bivariate relationship is statistically significant at 1%, 10% respectively (P-value of difference in Mean = 0.000, 0.074 respectively) However, it is not statistically significant difference in mean of income between people recycling and non-recycling The bivariate relationship is not statistically significant at 10% (P-value

of difference in mean = 0.963)

Table 8 A comparison between carton recycling and non-recycling in

term of gender, type of housing and education

(%)

Non-Recycling (%)

P-value of square test

high school or lower 58.06 58.49

Source: Author’s calculation

The Table 8 results show that household gender is associated with carton recycling Among 93 people recycling carton, 43.41% are male and 65.59% are female Likewise, in the remaining group the number are 47.66% male and 52.34% female