2020 - Chan Siu Ming-2

Impact of Housing on Poverty Situation in Hong Kong:
An Examination on Deprivation, Social Exclusion, and Subjective Poverty
CHAN, Siu Ming
A Thesis Submitted in Partial Fulfilment
of the Requirements for the Degree of
Doctor of Philosophy
in
Social Welfare
The Chinese University of Hong Kong
July 2020
博士論文
香港房屋對貧窮狀況的影響︰
生活匱乏、社會排斥與主觀貧窮的探討
陳紹銘
香港中文大學
社會福利哲學博士課程
2020 年 7 月
Thesis Assessment Committee
Professor TO Siu Ming (Chair)
Professor WONG Hung (Thesis Supervisor)
Professor LAM Ching Man (Committee Member)
Professor YIP Ngai Ming (External Examiner)

Abstract
Housing poverty, specifically the poverty situation induced by housing factors, is the core
concern of this PhD thesis. In this study, a mixed-method approach was used to examine the
housing poverty situation in Hong Kong. Firstly, three housing poverty lines were used to
analyse census data from 1996 to 2016 to examine the housing poverty situation. It was
observed that the overall housing poverty was not worsened because of the increasing
number of homeowners who had paid off all their mortgage payments. However, the housing
poverty situation for the residents of private rental housing was seriously aggravated.
Secondly, another Hong Kong population survey revealed that the living area and indoor
housing problems caused a significant impact on the non-income poverty situation, including
deprivation, social exclusion, and subjective poverty. Both housing factors acted as mediating
factors between income and non-income poverty. Of the three non-income poverty situations,
subjective poverty was influenced most by housing factors, whilst social exclusion was the
least affected. Thirdly, qualitative interviews of 20 residents living in subdivided flats showed
that residents fell into poverty owing to various critical life events and because they had
limited housing choice. High costs of housing, small living areas, and poor housing
conditions further exacerbated their poverty. This thesis adds to the empirical description and
theoretical explanation of poverty by including the housing perspective in poverty study. The
findings on the significant impacts of housing factors on the non-income poverty situation
1
prompted this study to propose progressive housing policies to alleviate the non-income
poverty among Hong Kong citizens.
Keywords: poverty, housing, deprivation, social exclusion, subjective poverty, Hong Kong
2
論文摘要
房屋貧窮，或更具體來說，因房屋因素引致的貧窮情況，是本論文的研究重點。
首先，本論文分析 1996 至 2016 年的人口普查數據，計算出三條不同的房屋貧窮線，
探討香港的房屋貧窮情況。結果顯示，一方面整體房屋情況未見惡化，主要由於完成
供款的業主比例一直增加；而另方面，居住在私人租住房屋的房屋貧窮情況卻越見嚴
重。其次，另一個全港調查結果顯示，住屋面積及室內環境問題對匱乏、社會排斥及
主觀貧窮都有顯著影響，此外，該兩類房屋因素亦被驗證為收入及非收入貧窮的中介
因素，在三類非收入貧窮中，主觀貧窮最受房屋因素影響；反之，社會排斥則最不受
影響。然後，本論文透過訪問 20 位現居或曾居於劏房的居民，他們因著不同的人生事
件成為低收入家庭而被限制了房屋選擇，惡劣環境及昂貴租金進一步令貧窮情況惡
化，令他們捲入貧窮旋渦及面對多元房屋貧窮。本論文的主要貢獻在於加強以房屋視
角來對研究貧窮的實證描述與理論解釋。本論文基於房屋因素對於非收入貧窮有關鍵
影響的發現，提出針對房屋問題政策建議以減低香港市民非收入貧窮狀況。
關鍵字︰貧窮、房屋、匱乏、社會排斥、主義貧窮、香港
3
Acknowledgements
The ideas of this thesis emerged from my previous work experiences as a social worker
in Hong Kong. I would like to thank the service users who faced or are fighting against
housing and poverty problems. They worked with and taught me the meaning of poverty and
housing. Special thanks go to all my informants, who provided their valuable time and shared
their precious experiences. I also acknowledge all social and community workers who
referred the cases for my study.
I express my heartfelt appreciation for my supervisor, Professor Hung Wong, for his
guidance and patience in the past years. I sincerely appreciate his care and concern for the
poor. This attitude inspired me to commence my academic research to serve the poor. My
pursuit of a PhD is a challenging but rewarding journey. Without the trust and support of my
supervisor, this study would not have been completed.
I am likewise expressing my appreciation to my thesis committee members, namely,
Professor Siu Ming To, chairman of my doctoral study panel; Professor Ching Man Lam,
member of the same panel; and Professor Ngan Ming Yip, my external examiner, for their
comments and recommendations to improve my thesis. I am also grateful to my previous
thesis committee members, namely, Professor Terry Leung Tse Fong and Professor Yu
Cheung Wong. Their constructive comments in my proposal formulation enabled me to
clarify my research direction and improve the research methodology.
4
Meanwhile, I am grateful for the enlightenment provided by my graduate and post-
graduate teachers, including Professor Steven Ngai Sek Yum, Professor Ji Kang Chen,
Professor Hai Jing Dai, Professor Timothy Leung Yuk Ki, Dr. Yuk King Lau, Professor
Sammy Chiu Wai Sang, and Dr. Ho Lap Fung. Moreover, my host supervisors in my
exchange programme in the University of Bristol, namely, Professors David Gordon and
Misa Izuhara, gave me critical comments and advice for my thesis, as well as shared their
knowledge of housing and poverty. Evidently, I am deeply indebted to their assistance.
Special thanks also go to all my peers and friends, who inspired and supported me in the
anti-poverty movement and my PhD study. Particularly, I am thankful to Fernando Cheung,
Tat Chor Au Yeung, Man Ying Fong, Emay Chan, Lai Shan Sze, and Apple Cheung. I am
likewise grateful for the research grant from the Madam Tan Jen Chiu Fund of the
Department of Social Work, The Chinese University of Hong Kong.
I am likewise expressing my gratitude to my parents and family for their unconditional
love and support for my endeavours in life. Lastly, I am profoundly grateful to my love, Ceci.
Her love, support, and acceptance enabled me to complete my PhD study.
I thank God for all the graces that I have received.
5
Table of contents
Chapter 1 Introduction 15
1.1 Background and Rationale 16

1.1.1 Research background 16
1.1.2 Multidimensional nature of poverty 17
1.1.3 Multifaceted nature of housing 19
1.1.4 Variety of ‘housing poverty’ 20
1.1.5 Hong Kong context 21
1.1.6 Theoretical and empirical research gap 23
1.1.7 Conceptual framework 23
1.2 Research objectives 25
1.3 Research questions 26
1.4 Overview of the thesis 29
Chapter 2 Theoretical Review of Housing and Poverty 30
2.1 Poverty 31
2.1.1 Conceptualization and Definitions of Poverty 31
2.1.2 Different Approaches of Poverty 39
2.1.3 Short summary 57
2.2 Housing 59
2.2.1 Nature of housing 59
2.2.2 Theories or approaches for housing study 66
2.2.3 Political economy of housing 68
2.3 Housing poverty 75

2.3.1 Meanings of ‘housing poverty’ 75
2.3.2 Housing affordability (monetary ‘housing poverty’) 76
2.3.3 Relationship between housing and non-income poverty 91
2.4 Conclusion: Theoretical gaps and significance of research 109
Chapter 3 Empirical Review of Housing and Poverty in Hong Kong 111
3.1 Academic literature review 112
3.2 Poverty Situation in Hong Kong 117
3.3 Housing situation in Hong Kong 129

3.4 Housing poverty situation in Hong Kong 135
6
3.5 Political economy background of Hong Kong 145
3.6 Conclusion: Empirical study gaps and significance of research 154
Chapter 4 Philosophical Foundation, Research Method and Design 156
4.1 Philosophical foundation of social research 157

4.1.1 Comparison of social research paradigms 157
4.1.2 Pragmaticism and critical realism 159
4.1.3 Position of the researcher 164
4.2 Research method 169

4.2.1 Quantitative and qualitative research 169
4.2.2 Mixed method research (MMR) 175
4.2.3 Secondary data analysis 178
4.3 Research design 182

4.3.1 QUAN: Corresponding research questions and hypotheses 182
4.3.2 QUAN: Data collection strategy 184
4.3.3 QUAN: Data analysis strategy 186
4.3.4 QUAL: Corresponding research questions and hypotheses 187
4.3.5 QUAL: Data collection strategy 188
4.3.6 QUAL: Data analysis strategy and research quality 189
Chapter 5 Quantitative Analysis: Part 1 192
5.1 Introduction of quantitative analysis 192
5.2 Housing poverty situation in Hong Kong (Census) 193

5.2.1 Data and sample 193
5.2.2 Operationalization of concepts and measurements 195
5.2.3 Statistical analysis strategy 197
5.2.4 Result and Findings 197
5.3 Discussion for census data analysis 217

5.3.1 How is the income poverty situation in the past 20 years and why? 217
5.3.2 What is news for using housing poverty measurements? 218
5.3.3 Whose ‘housing problems’? Whose ‘housing profit’? 223
5.3.4 Who are the groups that greatly suffer from housing poverty? Is it different from
income poverty analysis? 225
5.3.5 Success and failure of public housing policy 227
5.3.6 Explanation of housing poverty: political economy of housing in Hong Kong 228
7
Chapter 6 Quantitative Analysis: Part 2 236
6.1 The impact of housing factors on non-income poverty 236

6.1.2 Corresponding research questions and hypothesis 237
6.1.3 Operationalisation of concepts and measurements 237
6.1.4 Statistical analytical strategy 242
6.1.5 Descriptive Result 242
6.1.6 Inferential Analysis 245
6.2 The mediating effect of housing factors on non-income poverty 254

6.2.2 Corresponding research questions and hypothesis 255
6.2.3 Operationalisation of concepts and measurements 255
6.2.4 Statistical analytical strategy 257
6.2.5 Descriptive Result 259
6.2.6 CFA 260
6.2.7 Path Analysis and structural equation modelling (SEM) 261
6.3 Findings and Discussion 273

6.3.1 Impact of housing factors on deprivation 273
6.3.2 Impact of housing on social exclusion 276
6.3.3 Impact of housing on subjective poverty 279
6.3.4 Mediating and interacting impact of housing factors 282
6.3.5 Exploration of ‘multidimensional housing poverty’ 285
6.4 Short conclusion of the quantitative analysis 287
Chapter 7 Qualitative Analysis 289
7.1 Background 290

7.1.1 Profile of Informants 290
7.1.2 Data Collection and Analysis 291
7.1.3 Overview of Themes 292
7.2 Three Stories of Housing Poor 294

7.2.1 ‘I want to jump down to the street … I am exhausted and depressed …’ 294
7.2.2 ‘I found houses more than 30 times … I have no choice to move …’ 296
7.2.3 ‘I cannot sleep at my home … I voice out and hope the policies can be
changed…’ 299
8
7.3 ‘To Me, Housing is … Poverty is …’: Multi-Dimensions of Housing and Poverty
302
7.3.1 ‘Housing is…’: 5S of housing need 302
7.3.2 ‘Poverty is …’: Material, social and subjective dimensions 305
7.3.3 ‘Housing and poverty are…’: the relationship between housing and poverty 310
7.3.4 Short discussion 312
7.4 ‘I am Poor, because…’: Critical Life Events 315

7.4.1 Family change, marriage, and broken marriage 315
7.4.2 Migration from China 316
7.4.3 Getting sick, the whole family changed 317
7.4.4 Unstable employment, unstable housing 317
7.5 ‘I am Poor, so…’: The Impact of Poverty on Housing 321

7.5.1 No money, no housing choice 321
7.5.2 Lack of social capital and poor information access 322
7.5.3 Buying a house is not my choice 322
7.6 ‘Housing Makes Me Poor’: Impact of Housing on the Non-Income Poverty

Situation 325
7.6.1 Housing is one dimension of deprivation that is magnified: Impact of housing on
deprivation 325
7.6.2 Is housing related to social exclusion? It depends. The impact of housing on
social exclusion 327
7.6.3 ‘I do not want to tell others where I live’: Impact of housing on subjective
poverty 329
7.6.4 ‘Poor housing, poor health’: Impact of housing on physical and mental health.
331
7.6.5 Short Discussion 332
7.7 ‘Life is Full of Struggles’: Response, Struggle, and Coping Strategies of Residents
340
7.7.1 Resistance: Useful versus Useless 340
7.7.2 Resilience: To adapt versus Not to adapt 342
7.7.3 Future: Hopeful versus hopeless 344
7.8 Conclusion of Qualitative Analysis 348
9
Chapter 8 Conclusion 350
8.1 Conclusive response to the research questions 352
8.2 Theoretical Contributions and Significance 358
8.3 Empirical and Policy Contributions and Significance 362
8.4 Reflection 365
8.5 Limitations of this study 369
8.6 Recommendation and further research directions 373
8.7 Conclusive Remarks 375
References 377
Appendix 406
Appendix A Operationalisation of key housing variables 406
Appendix B Measurement of residual housing poverty 408
Appendix C Background of individual interview informants 410
Appendix D Consent Form 412
Appendix E Interview Guide 413
10
List of Tables
Table 2-1 affordability classification: income and expenditure 81
Table 2-2 affordability classification: poverty before and after housing costs 81
Table 2-3 relationship among housing factors and poverty situation 93
Table 3-1 Occupied Quarters, Owner Occupiers and Population in Domestic Households by
Year and Type of Housing 129
Table 3-2 Domestic households in Hong Kong by tenure (%) 131
Table 3-3 Poverty household, population and rate by tenure 135
Table 3-4 Tenure distribution with by selected poor household Group, 2018. 136
Table 3-5 Mortgage Payment and Loan Repayment 138
Table 3-6 Median Monthly Domestic Household Rent and Median Rent to Income Ratio, by
Type of Housing and Household Size 139
Table 3-7 Summary of official reports of SDU and residents. 143
Table 4-1 Comparison table of research paradigms, with reference of ideas of scholars 158
Table 4-2 amended framework for social study 167
Table 5-1 Background information of census data set from 1996 to 2016 194
Table 5-2 Demographic information of respondents (Bi-Census 2016) 198
Table 5-3 Housing cost of respondents (by household) 199
Table 5-4 Income and housing poverty situation (by person) (2016) 200
Table 5-5 Income and housing poverty rate (%), by housing type (2016) 201
Table 5-6 Income and housing poverty rate (%), by household size (2016) 203
Table 5-7 Income and housing poverty rate (%), by living area (2016) 204
Table 5-8 Income and housing poverty rate (%), by location (2016) 204
Table 5-9 Income and housing poverty rate (%), by family composition (2016) 205
Table 5-10 Income and housing poverty rate (%), by education (2016) 206
Table 5-11 Income and housing poverty rate (%), by age group (2016) 207
Table 5-12 Income and housing poverty rate (%), by place of birth (2016) 208
Table 5-13 Income and housing poverty rate (%), by duration of residence (2016) 208
Table 5-14 poverty situation from 1996 to 2016 (overall population) 209
Table 5-15 poverty situation from 1996 to 2016 (public rental housing) 211
Table 5-16 poverty situation from 1996 to 2016 (public subsidized housing with mortgage)
212
Table 5-17 poverty situation from 1996 to 2016 (public subsidized housing without
mortgage) 212
11
Table 5-18 poverty situation from 1996 to 2016 (private rental housing) 213
Table 5-19 poverty situation from 1996 to 2016 (private owned housing with mortgage) 214
Table 5-20 poverty situation from 1996 to 2016 (private owned housing without mortgage)
215
Table 5-21 Mortgage payment analysis (2001 to 2016) 216
Table 6-1 Demographic background of respondents 243
Table 6-2 Income and non-income poverty indices and poverty rate 244
Table 6-3 housing characteristics of respondents 244
Table 6-4 Logistic Regression Model 1: (DV: Deprivation) 246
Table 6-5 Logistic Regression Model 2: (DV: Social Exclusion) 248
Table 6-6 Logistic Regression Model 3: (DV: Subjective Poverty) 250
Table 6-7 Comparison of R2 among different regression models 254
Table 6-8 Descriptive statistics of the construct in SEM models. 259
Table 6-9 Correlation among observed variables for construct of social exclusion 260
Table 6-10 Correlation among observed variables for construct of subjective poverty. 260
Table 6-11 Regression Weights of SEM model 1 262
Table 6-16 Hypothesis and acceptance 282
Table 7-1 Profile of informants. 290
Table 8-1 Comparison of poverty lines and their application in Hong Kong 353
12
List of Figures
Figure 1-1 Relationship amongst poverty and housing situation, housing system and force
behind 24
Figure 2-1 A minimal definition of affordability 82
Figure 2-2 Relationship between Housing-induced poverty and Cost Burden 88
Figure 3-1 Poverty lines by household size, 2009-2018 118
Figure 3-2 Poor population and poverty rate after recurrent cash intervention 118
Figure 3-3 Poverty rate and poor population by selected socio-economic group, 2018 119
Figure 3-4 Expenditure patterns of poor and all households by household size, 2015 120
Figure 3-5 No. of domestic household by housing type, 2016. 130
Figure 3-6 Rental indices for Hong Kong property market 131
Figure 3-7 Price indices for Hong Kong property market 132
Figure 3-8 Private Domestics - completion by class 134
Figure 3-9 Completions of public rental housing flats 134
Figure 3-10 Housing characteristics of poor households 135
Figure 3-11 Poverty rate and poor population by housing type, 2018 137
Figure 3-12 Applications for public rental housing 141
Figure 3-13 Average waiting time of public housing 141
Figure 4-1 Positivist or ‘successionist’ view of causation 164
Figure 4-2 Critical realist view of causation 164
Figure 4-3 Continuum of mix methods research 176
Figure 4-4 Revised conceptual framework of relationship amongst housing and poverty
situation for quantitative study 182
Figure 5-1 Income and housing poverty rate (%), by housing type (2016) 202
Figure 5-2 poverty situation from 1996 to 2016 (overall population) 210
Figure 5-3 poverty situation from 1996 to 2016 (public rental housing) 211
Figure 5-4 poverty situation from 1996 to 2016 (private rental housing) 214
Figure 5-5 Mortgage payment analysis (2001 to 2016) 216
Figure 6-1 The independent effect of income and housing factors on non-income poverty 254
Figure 6-2 Housing factors as mediators among income and non-income poverty 255
Figure 6-3 SEM model 1: impact on deprivation 262
Figure 6-4 SEM model 2: impact on deprivation, private rental housing residents 264
Figure 6-5 SEM Model 3: impact on social exclusion 266
Figure 6-6 SEM model 4: impact on subjective poverty 268
13
Figure 6-7 SEM model 5: impact on subjective poverty, private rental housing residents 271
Figure 7-1 Relationship between themes 293
Figure 8-1 The mediating impact of housing factors on non-income poverty 355
Figure 8-2 The common path of housing poverty 358
Figure 8-3 Final framework of significant relationship amongst housing and poverty situation
361
14
Chapter 1 Introduction
‘Here is just around 20 to 30 square feet. I can just put a single bed here. That is all’.
‘When I was living in a subdivided flat, I thought I got depressed. Once, I told my
wife: I want to jump down to the street. I am really in trouble. I am really exhausted’.
‘Now? My living was worse than the cats and dogs, you see? I could not buy what I
want. I just bought the cheapest food to live’.
‘Normally, I use twenty dollars to buy some meat and five dollars to buy vegetable. I
cook and it will serve lunch and dinner. I count every day, I spent twenty-five dollars
at most’.
These quotes are from interviews with residents living in poor housing areas in Hong
Kong. The flat is tiny, the housing condition is dreadful, and rent is unaffordable.
Accordingly, the following questions should be answered: Why do the residents suffer? Why
does this situation happen in Hong Kong, which is an international financial centre? What
can we do to improve the well-being of the poor? I constantly ask these questions. In 2008, I
started working as a social worker, serving residents living in cage home, cubicles, and
subdivided flats. My vision was to enhance the well-being of grassroots people through direct
service, community organisation and policy advocacy. Given that I have practiced in these
methods, I would endeavour to explore new initiatives and horizon and extend my
intervention strategies and methods to include evidence-based and theory-informed academic
15
research. This thesis serves as the start of my academic research journey to serve the poor.
This chapter briefly introduces the background and rationale of this study (Section 1.1),
research objectives (Section 1.2), research questions (Section 1.3) and thesis overview
(Section 1.4).
1.1 Background and Rationale
1.1.1 Research background
We cannot disregard housing when we study poverty. Housing and poverty interact with
each other and are mutually defined. ‘Housing poverty’, which is also as ‘housing-induced
poverty’ or ‘poverty caused by housing factors’, is the core concern of this thesis and also my
concern to solve the poverty problems of the grassroots people in Hong Kong.
This research is prompted by the deteriorating housing and poverty situation faced by the
Hong Kong people, particularly for those living in subdivided flats, with high housing costs
and poor living conditions (Lau & Wei, 2018; Wong, 2018; Wong & Chan, 2019). In recent
years, housing prices and rent have boosted rapidly in Hong Kong (Demographia, 2017).
Housing price increased approximately six times from 2003 to 2019, and rent increased
approximately thrice during the same period. Moreover, the waiting time of public housing
lengthened from approximately two years in 2008 to over five years in 2019. The number of
subdivided flat residents continue surging. Over 200,000 people lived in this tiny inadequate
housing (Census and Statistics Deprtmant, CSD, 2016). The suffering of the poor population
16
is reflected by their low income and its related material livelihood and also the poor housing
environment and small living area where they live and the non-material livelihood. This
situation inspired the researcher to formulate the following questions: How does the housing
circumstance affect the poverty situation of the underprivileged? How is the ‘housing
poverty’ situation in Hong Kong? Which groups are suffering? How do they suffer? How do
we explain the ‘housing poverty’ situation in Hong Kong? Is the situation in Hong Kong
unique or similar to those in other cities? To answer these questions, ‘housing’ and ‘poverty’
are the two main themes that should be extensively reviewed.
1.1.2 Multidimensional nature of poverty
Poverty is multidimensional in nature. When referring to the basic level of living, poverty
refers to a lack of resources to fulfil basic needs and subsistence (Rowntree, 1901; Townsend,
1979b). For example, those with or without adequate income for basic level of food may be
identified as living in poverty (Orshansky, 1969). Accordingly, the monetary approach,
particularly income measure, has been dominant in the conceptualisation or
operationalisation of poverty for many decades (Laderchi, Saith, & Stewart, 2003).
Nevertheless, scholars have criticised that the monetary approach is considerably restricted to
portray the human suffering (Sen, 1999; Townsend, 1979b). Different frameworks or
perspectives have been proposed to enrich our understanding of poverty, including
deprivation, social exclusion, and subjective poverty. These perspectives are categorised as
17
‘non-income poverty’ in this thesis.
Peter Townsend pioneered in proposing the concept of deprivation to describe a lack of
socially perceived necessities. The idea of relative poverty is used as basis to demonstrate
that deprivation reveals a relative disadvantage of individuals to the local community or
wider community (Townsend, 1987). Those who lack certain types of necessary items are
identified as deprived. Deprivation is mainly used to conceptualise the material dimension of
poverty. However, the concept of social exclusion was developed in the 1960s and widely
used in European countries to describe the social dimension of poverty. Social exclusion was
initially used to characterise the poor, who are excluded from the labour market (Silver,
1994). Eventually, social exclusion was applied to depict the lack of social participation and
poor social relationship and social network (Levitas, 2006b). The third main dimension of
poverty is subjective poverty. Studies have argued that poverty is material-based and
subjectively felt as well (Pradhan & Ravallion, 2000). Subjective approach defines poverty
with subjective judgement and feelings of individuals, such as lack of well-being or
happiness (vanPraag & Ferrer-i-carbonell, 2006). Several measurement tools were developed,
such as minimum income assessment (Bradshaw & Finch, 2003; Kapteyn, Kooreman, &
Willemse, 1988) and self-evaluation questions (Kapteyn et al., 1988). Although many
approaches have been developed to conceptualise poverty, the multidimensional nature of
this phenomenon made it an endless topic for discussion or debate. No single approach can
18
fully describe and explain the entire landscape of poverty. Housing poverty is one of the
neglected areas in the study and discussion of poverty.
1.1.3 Multifaceted nature of housing
Housing is a necessity in life and also has an impact on all aspects of our well-being
(Stone, 1993). The simplest understanding of housing is a physical space with walls and ceiling.
Houses provide space for individuals and families to live in and serves as a place to protect
them from physical harm and natural disasters. Apart from the material dimension, the social
dimension of housing is also important. Houses serve as homes for many families and
individuals and enable people to build relationship with one another. However, the meanings
of housing are not limited in living rooms but also the geographical locations, communities and
neighbourhoods (Malpass & Murie, 1999). Neighbourhood relationships, community
resources and facilities are also important for residents to build social network and obtain social
capital.
Housing also has economic and political dimensions, and is commonly treated as a
commodity, with use and exchange values. In terms of exchange value, each flat typically
receives a price in the housing market. Housing prices are influenced by numerous factors,
including supply and demand of housing, housing policy and politics, flow of capital,
monopolisation and culture of speculation (Harvey, 2014). The rapid changes in housing prices
and rent have caused many housing issues or problems, particularly to the underprivileged.
19
Housing is a tool for capital accumulation and political tool for governance. The ideology of
housing and housing policy design has deeply affected the daily livelihood of citizens (King,
2015; Madden & Marcuse, 2016).
Overall, housing can contribute to human well-being by satisfying basic needs and promoting
social relationship, and also makes the largest cost to the poorest population. The nature and
impact of housing are ever-changing based on social context. For example, the housing
development in Hong Kong should be analysed in the context of world city with rapid flow of
capital and with consideration of cultural of speculation and pro-business governance.
1.1.4 Variety of ‘housing poverty’
‘Housing poverty’ is broadly defined as ‘housing induced poverty’ (Kutty, 2005).
Similar with the conceptualisation of poverty, the preceding definition is dominated by the
monetary approach. The extensive literature on this topic has focused on ‘housing affordability’,
which demonstrates the impact of housing cost burden on families (Bramley, 2012; Li, 2014;
Stone, 2006a). However, the nature of poverty is not limited to monetary dimension but also to
the material, social and subjective dimensions. Hence, housing poverty also means ‘housing-
induced deprivation’ (Borg, 2015), ‘housing-induced social exclusion’ (Arthurson & Jacobs,
2003) and ‘housing-induced subjective poverty’ (Clapham, 2010). These themes are the main
study themes of this thesis.
The relationship amongst the three types of non-income poverty is complicated. For
20
housing-induced deprivation, the physical condition of housing is associated with deprivation
(Napiorkowska-Baryla & Witkowska-Dabrowska, 2018). Moreover, housing cost and
location are also correlated with deprivation (Daniel, Baker, & Lester, 2018; Markkanen &
Harrison, 2013), although the mechanism of influence is unclear. In terms of social exclusion,
housing location and community resources have been found to be significantly related with
social exclusion (García-Vélez et al., 2020; Park, Cho, & Chen, 2019). Nevertheless, the
impact of housing tenure is ambiguous. Different housing factors have also been found
associated with subjective poverty, such as housing tenure (Wu, Stephens, Du, & Wang,
2019), community factors (Low et al., 2018) and the living environment (Rudolf & Potter,
2015). However, the effects of housing cost are questionable (Zhang, Zhang, & Hudson,
2018). Additional detailed discussion is provided in the literature review, thereby revealing
the complicated relationship between housing and poverty.
1.1.5 Hong Kong context
In recent years, housing and poverty problems have deteriorated rapidly and
tremendously across all cities in the world. Under globalisation, the increasing influx of
capital has caused a substantial increase in housing prices and rent in cities globally (Sassen,
2011). Moreover, the impact of neoliberalism has catalysed the privatisation and
financialisation of housing across the world (Harvey, 2005). This situation also impacts the
well-being and poverty situation amongst those living in highly globalised cites, particularly
21
in terms of tightening welfare expense, contractualisation of labour and loss of job security,
amongst others. In Hong Kong, which is one of the most open and globalised economies in
the world, housing and poverty problems have become considerably serious (Lee, Wong, &
Law, 2007). Housing prices and rent increase continuously, in which the former is the most
unaffordable amongst all cities in the world in the past years. Various empirical studies have
shown that the different dimensions of poverty, including deprivation and social exclusion, is
also getting worse (Chou & Lee, 2018; Saunders, Wong, & Wong, 2014). Moreover, housing
poverty has also been deteriorating in the past years. An increasing number of subdivided
units and residents. The poor population suffers from high rent, small living area and poor
living condition.
The Hong Kong government recognises the preceding situation, but it has yet to conduct a
comprehensive diagnosis of the problems and effectively address poverty. Although the
government has identified the official poverty line in 2013 and implemented different poverty
alleviation policies (HKSAR government, 2019), poverty rate remained steadily in high
levels even after implementing policy interventions. Furthermore, the official poverty
measurement and anti-poverty reliefs have been dominated by the income approach, which
disregards the other dimensions of poverty. The impact of housing factors on poverty has yet
to be sufficiently studied as well. Although the government has claimed placing housing as
the highest priority in its policy agenda, the housing condition of the poor has failed to
22
improve in recent years.
1.1.6 Theoretical and empirical research gap
The theoretical discussion of poverty has several research gaps that should be filled in,
including the housing perspective in conceptualising poverty, roles of housing factors in
explaining the different dimensions of poverty in the era of globalisation and how poverty
affects the housing factors. Meanwhile, empirical research gaps include the measurement of
after-housing cost poverty, relationship between housing and non-income poverty, housing
pathway and subjective experiences of residents and political-economic analysis of the
existence of housing poverty in Hong Kong.
1.1.7 Conceptual framework
Derived from the literature reviewed, which is focused on the interaction between
housing and poverty (Crisp, Eadson, & While, 2016; JRF, 2015; Stephens & vanSteen, 2011;
Tunstall et al., 2013), Figure 1-1 shows the study framework of this research. In the centre,
the housing factors and poverty situation is inter-related, such as housing cost, living
condition and community, affect poverty, or the income level will affect the housing situation
of residents. Housing policies and systems, such as housing planning and market regulation,
have an impact on housing and poverty. Additionally, structural forces are behind the housing
system, including state ideology, market forces, institutions, and beliefs of individuals, which
fundamentally affect the current circumstance. This research mainly focuses on studying the
23
impact of housing factors on poverty through quantitative and qualitative approach.
Moreover, this research attempts to explain the current situation in a macro approach, with
the analysis of social policy and structural forces behind.
Figure 1-1 Relationship amongst poverty and housing situation, housing system and force
behind
24
1.2 Research objectives
This research is conducted to examine the housing poverty situation to study the
interaction between housing and poverty situation of residents in Hong Kong, and to
understand the living experience of residents lived in poor housing. Mixed research methods,
mainly quantitative and supplemented by qualitative techniques, are adopted to achieve the
research purposes. Particularly, the objectives of this research are as follows:
1. To enrich the theories and concepts of poverty with a housing poverty perspective in a
globalised context,
2. To examine the housing poverty situation and interaction between the housing conditions
and poverty situation in Hong Kong,
3. To understand the living experiences of people living in poverty and poor housing
condition in Hong Kong and
4. To explain the current housing and poverty situation in Hong Kong.
25
1.3 Research questions
The housing and poverty situation of residents in Hong Kong has raised the research
interest of the researcher and led to the initial research questions. The proposed research
objectives narrowed down the research scope. Additional concrete research questions are
formulated derived from the theoretical and empirical gaps. This research attempts to answer
several major research questions and sub-questions:
■ RQ1. What is the housing poverty situation in Hong Kong?
◼ RQ1.1. What is the existing situation of housing poverty in Hong Kong?
◼ RQ1.2 What is the historical development of housing poverty in Hong Kong?
■ RQ2. How are the interactions between the housing situations and poverty in Hong
Kong?
◼ RQ2.1: To what extent do housing factors affect the different dimensions of
poverty?
◼ RQ2.2: Which type of non-income poverty is the most affected by the housing
factors?
◼ RQ2.3 Are housing factors the mediators of the impact of income poverty on non-
income poverty?
◼ RQ2.4 What are the impacts of housing factors on poverty and vice versa?
■ RQ3. Why does housing poverty exist, persist, or change in Hong Kong? What are the
26
factors or forces affecting the formation of housing circumstance?
■ RQ4. What is the experience of people living in poverty and housing poverty situation?
◼ RQ4.1 How do housing factors affect the poverty situation and daily lives of the
Hong Kong residents?
◼ RQ4.2 How do they understand the relationship between housing and poverty
situation?
◼ RQ4.3 How do they react, deal with, or attempt to change the existing situation?
Mixed methods are applied to answer these research questions. Quantitatively, RQ1 (and
RQ1.1 and 1.2) is answered by using the census population data sets in Hong Kong from
1996 to 2016. Analysis of the largest population data sets in Hong Kong facilitated in
examining the development of housing poverty situation in the past years. A discussion of the
analysis result follows and the applicability of housing poverty measurement in Hong Kong.
To answer RQ 2 (i.e. RQ 2.1, 2.2 and 2.3), a cross-sectional random sample data set in
Hong Kong is used. By using regression modelling, the impact of housing factors on non-
income poverty situation are studied. Moreover, the mediating effects of housing factors
between income and non-income poverty are investigated by using structural equation
modelling (SEM).
Discussion and dialogue with the literature followed the quantitative analysis attempts to
answer RQ 3 and explain the long-term housing poverty situation through an analysis of the
27
political economy of housing in Hong Kong.
A qualitative approach is applied to answer RQ2.4 and RQ4 (i.e. RQ4.1, 4.2. and 4.3).
The living experience of residents are revealed through in-depth individual interviews.
However, several questions remain unanswered after the quantitative analysis. The interview
directions and questions are based on the literature reviewed and the initial result of the
quantitative analysis.
28
1.4 Overview of the thesis
This thesis has eight chapters. Chapter 1 introduces the research background, proposed
research objectives and research questions. Chapter 2 presents the theoretical literature
review on the concepts of ‘poverty’, ‘housing’ and ‘housing poverty’. Chapter 3 provides an
empirical review of the Hong Kong situation. Statistics on the housing and poverty situation
are examined through a study of the political-economic background of the housing and social
welfare in Hong Kong. Chapter 4 discusses the philosophical foundation of this research and
provides the rationales for using the mixed methods model in this thesis. Research design,
including data collection and analysis strategy, are presented in this chapter as well. Chapter 5
presents the result of the quantitative analysis of census data, followed by a discussion on the
housing poverty situation in Hong Kong. Chapter 6 presents the quantitative result of the
SPPR data set with the use of regression and SEM. A discussion follows to investigate the
impact of the housing factors on the poverty situation in Hong Kong. Chapter 7 focuses on
the qualitative analysis of the individual interviews to explore and understand the living
experiences of people facing housing poverty. Chapter 8 concludes the theoretical and
empirical significance of this research and presents the personal reflection of the author. The
Appendix includes the details of the data sets and interview guide and followed by the
reference list.
29
Chapter 2 Theoretical Review of Housing and Poverty
This session reviews the theoretical discussion of three main concepts used in this
proposal, ‘poverty’ (session 2.1), ‘housing’ (session 2.2) and ‘housing poverty’ (session 2.3).
It aims to enrich and broaden the understanding of them, including their definitions,
meanings, causes, explanations.
Providing theoretical background for the study is crucial and this will identify the
theoretical research gaps which this study intended to fill. Theories generally means
repositories of general knowledge (Ritzer, 2004). Criteria of theories include definitions of
concepts or variables, domain or scope of application and relationships among terms or
variables. Theories normally provide explanations of existing social situation or social change
(O’Brien & Penna, 1998; Wacker, 1998). Overall, there are several key features for social
theories, 1) Description, conceptualization and abstraction of terms or social phenomenon; 2)
Explanation/ interpretation, provide analysis and explanation for the existing situation or
relationship among variables; 3) Prediction, based on analysis, offers forecasting for the
future; 4) Understanding, provides in-depth understanding for human behaviour; 5) Change,
criticize the current situation or propose ideal future, with suggestions of ways to achieve. In
this theoretical review of poverty, housing, and housing poverty, in limited words, it will
focus on description and explanation.
30
2.1 Poverty
2.1.1 Conceptualization and Definitions of Poverty
There are different definitions and conceptualizations of poverty, and accordingly
distinct types of measurements of poverty. Scholars has developed variety of categories and
classifications for poverty analysis, such as monetary, economic material, deprivation, social
exclusion, subjective poverty, capability, and social capital.
a) Brief overview of poverty definition
Before going into the discussion, it worth first look at some important definitions
proposed by famous scholars of poverty or stated by international organizations concerning
poverty.
First, poverty is related to basic need and covers various aspects of human well-being.
Peter Townsend (1979a) reviewed the literature for definitions of poverty and proposed one
of the most important definitions in the early year was given by Seebomn Rowntree (1901),
‘total earnings are insufficient to obtain the minimum necessaries for the maintenance of
merely physical efficiency as being in primary poverty’. It provided the ground for the
poverty research in the later decades (Townsend, 1979a). Townsend himself defined poverty
as ‘Individuals, families and groups in the population can be said to be in poverty when they
lack the resources to obtain the types of diet, participate in the activities and have the living
conditions and amenities which are customary, or are at least widely encouraged or approved,
31
in the societies to which they belong.’ (Townsend, 1979, p.31). Jonathan Bradshaw added
that, ‘Poverty (if it means anything) is a categorical need – one that must be met for human
beings to function’ (Bradshaw & Finch, 2003, p.513). And the working definition of poverty
by Joseph Rowntree Foundation was, ‘When a person’s resources (mainly their material
resources) are not sufficient to meet their minimum needs (including social participation)’
(Goulden & D’Arcy, 2014). Amartya Sen used the term ‘capabilities’ in ‘Development as
freedom’ and suggested, ‘poverty must be seen as the deprivation of basic capabilities rather
than merely as lowness of incomes, which is the standard criterion of identification of
poverty’ (Sen, 1999, p.86), or the World Health Organization’s definition, ‘Poverty is
associated with the undermining of a range of key human attributes, including health. The
poor are exposed to greater personal and environmental health risks, are less well nourished,
have less information and are less able to access health care’ (WHO, 2017)
Second, poverty is not with a single meaning but with multidimensional concepts and
meanings. David Gordon mentioned, ‘poverty’ could be considered to have a cluster of
different overlapping meanings depending on what subject area or discourse is being
examined’. For example, health can be a moral as well as a scientific concept. But normally
there was a problem for measuring poverty that confused both concepts. He proposed, the
best definition of poverty (line) needs to maximize the between the groups of ‘poor’ and
‘non-poor’ and minimize the difference within two groups. (Gordon, 1998, p.5). In ‘Voices of
32
the poor: Crying out for change’, Deepa Narayan and other scholars mentioned, ‘The
evidence suggests that poverty is a multidimensional social phenomenon. Definitions of
poverty and its causes vary by gender, age, culture, and other social and economic contexts’
(Narayan, et al. , 2000, p.26)
Third, the definition of poverty should be understood under its social context and value.
Lister (2004, p.3) highlighted, ‘There is no single concept of poverty that stands outside
history and culture. It is a construction of specific societies.’ Mollie Orshansky (1969, p.37)
mentioned the value involvement of definition, ‘Poverty is a value judgment; it is not
something one can verify or demonstrate, except by inference and suggestion, even with a
measure of error. To say who is poor is to use all sorts of value judgments.’
Forth, there are structural factors causing poverty and they can be changed. Many
definitions of poverty of INGOs highlighted these pointes. Oxfam (2017) denoted, ‘Poverty
means little income, too few assets, lack of access to basic services and opportunities, deep
inequalities, ongoing insecurity and little opportunity for development. Poverty is rooted in
inequality, and in human action or inaction. It can be worsened by natural disasters, human
violence, oppression and environmental damage, and maintained by institutions and
economic means.’. United Nations (2017) defined poverty with similar meaning ,‘Poverty is
more than the lack of income and resources to ensure a sustainable livelihood. Its
manifestations include hunger and malnutrition, limited access to education and other basic
33
services, social discrimination, and exclusion as well as the lack of participation in decision-
making. Economic growth must be inclusive to provide sustainable jobs and promote
equality’.
From the above definitions, poverty is multidimensional, and the meanings of poverty
are related to inadequacy of resource, human need, well-being, and freedom. The concrete
definitions depend on value judgement and different with specific social context.
b) Brief history of poverty studies
There has been a long history for discussion and research on poverty. For example, the
establishment of English Poor Law can be traced back to 16th centuries and the New Poor
Law was passed in 1834, which had significant effect on the poverty relief policies and
welfare system(Garland, 2016). But it was impossible to trace the history of poverty
comprehensively within the limited words in this thesis, instead, the researches and
discourses in recent decades are be briefly reviewed.
Charles Booth (1887) started the study of poverty in London in 1887; He used monetary
approach and divided the population into eight classes, which four of them have varied
degree of poverty. Another influential and pioneering work was conducted by Seebohm
Rowntree in 1901 studying the situation of York (Rowntree, 1901). It estimated that the
minimum level of income to fulfil basic need in nutrition, clothing and rent and defined it as a
primary poverty line. Those who were seen by informants to be in obvious squalor may
34
counted as secondary poor (Laderchi, Saith, & Stewart, 2003). Later, Rowntree conducted the
poverty research in York of UK in 1930s and updated the minimum standard of basic need.
These researches are influential for the later development of social policy and academic
researches. The method of Rowntree was adopted by the US government for developing the
poverty index based on food cost in 1940, with the nutrition approach developed by Mollie
Orshansky (1969). In UK, Peter Townsend had conducted a large survey about poverty in
Britain, using the concept of deprivation. With his literature review and analysis of poverty,
he publish the seminal work ‘Poverty in the United Kingdom’ in 1979 (Townsend, 1979b).
Oscar Lewis published his anthological researches of poverty since 1950s and the
discourse of cultural of poverty was wide spreading since 1960s. On the other hand, the
capacity approach proposed by Amartya Sen (1985, 1999) was also influential and this
affected the poverty measurement and research of World Bank and its anti-poverty policies.
Besides, the researches of social exclusion, spreading by scholars like Hilary Silver (Silver,
1994) and Anthony Atkinson (1998), also had significant impact on poverty and anti-social
exclusion policies in European Union(EU). There was also increasing of researches on
subjective poverty since 1980s (Colasanto, Kapteyn, & van derGaag, 1984; Kapteyn et al.,
1988; vanPraag & Ferrer-i-carbonell, 2006).
There were significant changes in describing the people in poverty, for example,
Welshman(2013) traced the history of terminology of the ‘underclass’ and found the change:
35
social residuum in 1880s, unemployable in early 1900s, social problem groups in 1930s,
problem families in 1950s, culture of poverty in 1960s, cycle of deprivation in 1970s,
underclass debate in 1980s and 1990s, social exclusion in 2000s (Crossley, 2016; Shildrick &
Rucell, 2015; Welshman, 2013).
c) Classifications and typology of poverty
To describe poverty, scholars have developed various categories for analysis. There is no
consensus on unique or universal definition of poverty (Kwadzo, 2015; Laderchi et al., 2003),
instead, multidimensional approaches have been adopted (Duclos, Sahn, & Younger, 2006).
Here are some examples for the classification or typology of poverty:
■ Monetary poverty, social exclusion and capability poverty (Kwadzo, 2015);
■ Monetary, capability, social exclusion and participatory (Laderchi et al., 2003)
■ Economic well-being, capability and social exclusion (Wagle, 2002);
■ Income with budget standard, income below a threshold, lack of items of necessities,
feeling poor, social exclusion (Bradshaw & Finch, 2003);
■ Direct and indirect poverty (Ringen, 1988);
Hangenaars & de Vos (1988) suggested the all definitions of poverty can be fit in one of
the three classification: 1) having less than an absolute and objectively defined minimum
standard; 2) having less than other people; 3) feeling not having enough to get along From
this classification, they proposed at least eight approaches for measuring poverty, including
36
1a) Basic need; 1b) Food expenditure to income ratio; 1c) Fixed cost to income ratio; 1d)
Total expenditure to income ratio; 2) Relative deprivation of commodities; 3a) Subjective
minimum income; 3b) Subjective minimum consumption; and 3c) Official minimum income.
d) Considerations for poverty definitions
Why do we need to define poverty? One simple reason is the study of different
definitions of poverty help to enrich the understanding of poverty (Wagle, 2002). But what
are the implications or meanings for a deeper understanding? Number of scholars suggested
practical reasons for defining poverty. Orshansky (1969, p.37) suggested ‘There is no
particular reason to count the poor unless you are going to do something about them’. The
choice or study of the definitions not only enhance our understanding of nature and causes of
poverty, but also promotes ideas to reduce or eradicate poverty (Kakwani & Silber, 2008;
Sen, 2000). The definitions may help in describing and identifying the people in deprivation
and lead to public action.
Sen proposed ‘diagnosis precedes policy choice’(1992, p.107) that the policy making
needed to depend on the definition of poverty. The definitions of poverty have crucial
consequences to the poor population, for example, the setting of anti-poverty policies
(Hagenaars & deVos, 1988; Kwadzo, 2015). The measurement of poverty need to related to
the construction of policy (Townsend & Gordon, 2002). There are also social and political
concerns for choosing definition. Alkire (2005) proposed the choose of approaches need to be
37
user-friendly for better spreading of the idea. For an irrational or arbitrary measurement of
poverty , it is difficult to design appropriate policy for eradication of poverty (Townsend,
2002).
There are other considerations for defining poverty. One is the scope of concerns,
whether it includes only material aspect, or also psychological, political, and cultural
perspectives. The second one is the universality of definition, whether the definition can be
used in distinct types of society, for example, in both developed and developing area. Or the
definitions need to be specific and fit in certain society (Laderchi et al., 2003). The third
concern is the debate of objective or subjective measurement. The camp of subjectivity
argued that the definition of poverty should involve value judgement and socially
constructed, but it is also suggested that the measurements needed to be scientific and
objective for explicitly usage. Fourthly, the unit of analysis, for example, measure in unit of
family or individual, a family member can be materially poor in a ‘non-poor’ family if the
distribution of family income is uneven within the family.
Lastly, there is political or ethical issue, questioning the power and justification in
defining who poor and non-poor is, especially using an arbitrary poverty line. The choice of
poverty definition is scientific and also political (Laderchi et al., 2003). Moreover, to choose
among poverty definitions, Ruth Lister (2004) suggested it need to decide whether the
definition should be broad or narrow, income or living standard, income or capabilities,
38
absolute or relative.
2.1.2 Different Approaches of Poverty
Major approaches to poverty definition are be introduced in this session, including
monetary approach, deprivation, social exclusion, subjective approach, capability, and social
capital.
a) Monetary/ Economic resource approach
Among number of dimensions, the monetary/ economic resource approach is the most
dominant or commonly used approach in academic researches and policy implementations
(Laderchi et al., 2003).
Definition and nature. This approach assumes the welfare of a person is mostly
depends on economic resource owned, amount of income, the monetary expenditure or the
utility created by consumption (Laderchi et al., 2003). Wagle (2002) suggested there are
mainly three types of measures for economic well-being: income, consumption and welfare.
This can be measured in absolute, relative, or subjective approach. For example, a
person or a family categorized as poverty when their income is less than a certain threshold.
There are different types of threshold setting. For example, in early years, Orshansky (1969)
developed two thresholds, the ‘poor’ and ‘near poor’, from the food consumption surveys of
Department of Agricultural in 1948 and 1955. The thresholds are defined by the minimum
income to purchase basic level of food needed, in absolute approach. In that time, families
39
spent around one-third of income on food in average.
The World Bank measure extreme poverty and set the threshold as US1.9 per day, those
who cannot have enough income for such spending are classified as poverty. More commonly
used, in relatively approach, the threshold can be set of certain percentage of median
household income of a place, for example, EU used 60% of median household income as the
poverty line.
Discussion. One significant advantage of this approach is that it provides a clear
threshold for measurement. Money or income is most used for purchasing necessities, as to
fulfil basic need. Moreover, it can be easily calculated and used for comparison of living
standard of people among cities and nations.
However, there are limitations of this approach. For example, the nutrition approach,
Townsend (1962) criticized the setting of income requirement to purchase minimum nutrition
was a hazardous exercise as the calculation of nutrition requirement was with wide range of
error, even suggest by nutritionists. It is also hard to define how much is the minimum
standard of subsistence or requirement for basic living (Alcock, 1997).
For income approach, it put too much focus on monetary aspect but misses the
heterogeneity of human nature, such as subjective feeling, dignity. In 1962, Townsend
already mentioned there were other aspects of poverty other than income perspective, for
example, housing, education, medical and welfare service (Townsend, 1962). Moreover, the
40
capital asset of individuals or families are usually overlook for the commonly used income
approach (Orshansky, 1969).
For expenditure approach, there are at least three limitations for using expenditure as
measurement of poverty. First, the level of ratio is necessarily arbitrary, for example, 30% or
40% spending on food means suffered in poverty. Second, the expenditure of a family may
depend on tastes rather than necessity. Third, people may limit their expenditure because of
low income, for example, they need three meals a day but choose to eat two because of
poverty. (Hagenaars & deVos, 1988)
Household income is usually used for measurement. However, the power in distribution
within families may be differed, for example, a working family member may have more
power for owning money than the housewife, a member can be suffered from poverty in an
non-income poor family (Laderchi et al., 2003). Moreover, the patterns of consumption of
family members may vary, equivalence scales need to be used to adjust for different
household composition.
This approach is criticized by its focus on individual well-being but overlook the
interaction and interdependence with the community. Additionally, it is normally measured
by scholars externally, without involvement of the poor people themselves.
b) Deprivation
Definition and notion. Peter Townsend was the pioneering and one of the most
41
influencing scholars proposing the ideas of deprivation. The academic literatures of
deprivation in recent decades mostly cited Townsend’s work as foundation. It worth first
viewing the ordinal ideas and words of Townsend to understand the idea of deprivation.
Peter Townsend had proposed the idea of ‘relative poverty’ in earlier year, ‘My main
thesis is that both 'poverty' and 'subsistence' are relative concepts and that they can only be
defined in relation to the material and emotional resources available at a particular time to the
members either of a particular society or different societies’ (Townsend, 1962, p.210). ‘The
vague concept of 'subsistence' is an inadequate and misleading criterion of poverty’ and
‘Poverty is not an absolute state. It is relative deprivation.’ (Townsend, 1962, p.224). At least
since 1962, Townsend already clearly stated the limitations of absolute poverty.
Later, in the first paragraph of Townsend’s book ‘Poverty in the United Kingdom’, he
denoted ‘Poverty can be defined objectively and applied consistently only in terms of the
concept of relative deprivation. That is the theme of this book. The term is understood
objectively rather than subjectively’ (Townsend, 1979a, p.31). Townsend proposed three
forms of deprivations. They are ‘objective deprivation’, ‘conventionally acknowledged or
normative deprivation’, and ‘individual subjective group deprivation’. The first one
represents majority of value in the society. The second and the third one focus on the views of
minority groups and they are different in a matter of degree (Townsend, 1979a).
Townsend proposed, ‘Deprivation may be defined as a state of observable and
42
demonstrable disadvantage relative to the local community or the wider society or nation to
which an individual, family or group belongs’ (Townsend, 1987, p.125). He suggested the
concepts of deprivation cannot be viewed as only a scientific scale, instead, ‘They are
socially 'structured'—through a process of familiarity and indifference, advocacy and
repetition in social experience and discourse’ (Townsend, 1987, p.127). Deprivation has both
dimensions of objective and subjective (socially perceived). It is not only scientific but also
applicable for social policy to allocate resources.
For simpler meaning, deprivation can mean ‘a lack of socially perceived necessities’
(Bradshaw & Finch, 2003, p.515). But Townsend suggested three reminders for defining
deprivation. First, the concept of deprivation needs to be distinguished from the concept of
poverty. Second, the operational definition needs to be comprehensive and try to include all
material and social aspect of life of people. Third, it also need marking difference between
material and social deprivation (Townsend, 1987). Townsend proposed using the conception
of ‘style of life’ instead of ‘consumption’ and use of ‘resource’ to replace ‘income’. The
formers are wider concepts to represent the lives and behaviours of people and the later are
just the ways of expressions (Townsend, 1979a)
There were large number of reports and literatures about the measurement of deprivation
since 1980s (Gordon & Pantazis, 1997; Mack & Lansley, 1985). Even the UK government
applied the ideas of deprivation in measurement in 1980s, for example, the department of
43
Environment developed eight indicators of deprivation in 1983, including unemployment,
overcrowding, single parenting and pensioners (Townsend, 1987). Townsend himself had
developed a pioneering list of deprivation for a survey of poverty and labour market
conducted in 1985 to 1986, 77 indicators were proposed. The ‘material deprivation’ includes
deprivation of diet, clothing, housing, home facilities, environment, location and at work. The
‘social deprivation’ includes deprivation of recreation, family activity and education, and lack
of rights in employment, integration into community, formal participation in social
institutions. There were some indicator weighting more and the maximum score is 94
(Townsend, 1987). In Hong Kong, Saunders, Wong and Wong (2014a) applied the
deprivation approach to measure the poverty situation and reported an 18.2% deprivation rate
in 2014. Amongst several factors, income poverty was commonly associated with deprivation
(Saunders et al., 2014a).
Discussion. The deprivation approach broadens the measurement of well-being,
compared with income approach. People may be found deprived in some aspect without
simply classified as poor or not. The researches of deprivation approach are much more
detailed and multidimensional with in-depth investigation in the daily lives of people.
However, the list of necessity items is normally defined by scholars or experts externally,
without enough participation of the public. This may overlook some important and necessary
items of families. Moreover, how many items lacked by a family can be classified as
44
deprivation? The definition of threshold is necessarily arbitrary. For measurement
consideration, research in deprivation approach is costly and the items of necessities may
change overtime and not easy to make comparison.
c) Social Exclusion
Definition and notion. The concept of social exclusion is now widely used in European
countries and international agencies like United Nations and the World Bank. Some scholars
suggested the use of the concept of social exclusion started to describe the social situation of
France in 1960s (Silver, 1994), about large scale of unemployment, increasing inequality and
disruption of social cohesion (Kwadzo, 2015). Levitas (2006b) claimed the terminology of
social exclusion was originated with the publication ‘Les exclus’ by Rene Lenoir in France on
1974, concerning the people excluded from social protection. She proposed the use of the
concept in Britain was influenced by the work of Peter Townsend in 1979, which widen the
meaning of poverty to participation in society or social inclusion. Townsend (2002) proposed
that there were widely usage of the concept of social exclusion in France in 1980s, with
concern of long-term unemployment or lack of social insurance. It illustrates the relative
deprivation and process of marginalization or becoming poor, especially in the advanced
industrialized areas (Kwadzo, 2015; Laderchi et al., 2003).
Atkinsons (1998) argued social exclusion was widely used but without clear definition.
The only common point was that it is impossible to define ‘excluded’ by one unique criterion.
45
Nevertheless, scholars had tried to define social exclusions with different framework.
Atkinsons proposed three elements of social exclusion. First, relativity, meaning people are
excluded from certain place and time relatively. The judgement of social exclusion cannot be
isolated from the social circumstances. Second, agency, people are usually excluded from
objects, for example, excluded from a company, from a bank, so they cannot get a job or
borrow money. Third, dynamics, exclusion is not only describing the current situation but
also the prospects of future (Atkinson, 1998).
Hilary argued it was difficult to exactly define social exclusion, nevertheless, she
suggested three paradigms of social exclusion, namely solidarity, specialization, monopoly.
Each paradigm attributes based on different conception of integration and political
philosophy: republicanism, liberalism, and social democracy. For example, solidarity
paradigm focuses on cultural boundaries and moral exclusion; Specialization highlights the
importance of liberty and problems of discrimination; and Monopoly paradigm stressed the
inequality caused by monopolies.
Levitas (2006b) proposed that there are three discourses of social exclusion. First, the
redistributive discourse (RED), represents the poor people lack resource, including money, to
access the public services. The second one is social integration (SID), concerns the long-term
unemployment problems and labour market exclusion. The third is the discourse of moral
underclass (MUD), which focuses on the moral deficiencies and behaviours of the groups
46
which viewed as problematic. Levitas demoted that it is difficult to make distinction between
social exclusion and poverty. She reviewed many reports of government, academic and
organization in studying social exclusion and concluded there was no clear definition of
social exclusion and how it differed from poverty in all the reports. Sometimes it can be
generally call ‘poverty and social exclusion (PSE)’ in use (Levitas, 2006b).
There are different types of exclusion such as excluding from labour market, political
rights, citizenship, cultural activities (Wagle, 2002) or from social service, social relations,
social activities, social networks and civic participation (Levitas, 2006b). Empirical
researched shown that different dimensions of social exclusion can be highly correlated, one
set of social exclusion may lead to other types of exclusion. (Kwadzo, 2015; Laderchi et al.,
2003)
As the concept of social exclusion is complicated, it is different to make general and
universal indicators for measurement. Nevertheless, there were attempts in developing sets of
indicators. For example, in Europe, the Social Protection Committee of EU endorsed a set of
18 harmonized indicators of social exclusion in 2001. The set divided into part primary and
secondary. The primary set included ‘persistence of low income’, ‘long-term unemployment
rate’ and ‘life expectancy at birth’. The second set included ‘Gini coefficient’, ‘person with
low educational attainment’. Besides EU, there were different indicators set proposed by
other organizations, such as Institute for Public Policy Research (IPPR) and Centre for the
47
Analysis of Social Exclusion (CASE) (Levitas, 2006b).
Discussion. As the concept of social exclusion is socially defined, it put more focus on
the characteristics of specific social groups instead of individual base. However, with its
relative and multidimensional nature, it is not easy to make a clear or unique definition and
measurement of social exclusion (Laderchi et al., 2003). Many scholars tried to define social
exclusion in different approaches and with different dimensions or paradigms, but there was
still no one clear definition that can be commonly used. On the contrary, although there are
several limitations of income approach, it is a clearer definition and measurement which can
be used worldwide.
On the other hand, social exclusion can make a difference from the concept of poverty.
There are some distinctions between social exclusion and poverty proposed by scholars. First,
it is argued that poverty is a narrower concept which focus on lack of economic resource and
social exclusion is a broader concept concerning the social aspect of people (Gordon, 1998).
Second, some argued poverty usually used to describe a static situation, instead, social
exclusion studies on the dynamic process of exclusion (Room, 1995). But Gordon denoted
the definition of social exclusion of EU in 1990s was much narrower, which mainly focus on
the situation of unemployment (Gordon, 1998). Additionally, there are contrasting studies
suggesting that social exclusion was the cause of poverty and the other studies proposed the
opposite which poverty was the cause of social exclusion. It is questionable that which
48
concept is more comprehensive (Wagle, 2002).
Given the complex nature and definition of social exclusion, universal indicators for
measuring this type of exclusion cannot be easily developed. Amongst the different
definitions of social exclusion, social support and social network are most commonly used for
operationalised measurements. Given the availability of data, this study focuses on the social
dimension of social exclusion.
d) Subjective poverty approach
Definition and notion. Poverty is not only an objective measurement but also a
subjective feeling. For example, the view of basic nutrition requirement is not fully objective,
but with social specificity and subjectivity of the expert (Pradhan & Ravallion, 2000). Leu,
Chen & Chen (2016) suggested there are numerous limitation for using objective approach,
for example, the objective criteria was not socially neutral but with bias. Although it seeks
generalization and standardization, the social difference of culture is usually overlooked. The
tradition approach focusing on income and consumption was challenged by those with
broader consideration including concerns of capability and basic need. Subjective well-being
or subjective perceived welfare was one of the important aspects of human being which
worth for more attention (Kingdon & Knight, 2006). To overcome the limitation of
mainstream objective measurement, scholars developed alternative approach for measuring
subjective poverty (Mahmood, Yu, & Klasen, 2018).
49
The subjective approach focuses on the subjective judgement and feeling to define poverty.
Some scholars defined poverty is ‘lack of happiness’ or ‘lack of well-being’ and this is
psychological loaded. The level of poverty is depending on the subjective interpretation of
people (vanPraag & Ferrer-i-carbonell, 2006). The subjective approach proposed that the
human well-being is related to experience which is essentially subjective. It recognized the
authorities of human in defining and accessing their own well-being. Moreover, the level of
well-being can be declared by the person (Rojas, 2008). Compare with measuring snapshot
income and expenditure, the subjective measurements can relatively capture longer-term of
well-being, for example, with consideration of the wealth and asset of the respondents (Posel
& Rogan, 2016).
Subjective poverty can be measured in different ways. One common measure is self-
evaluation of living standard (Siposné Nándori, 2014). A simple measurement is asking the
direct feeling of person, based on the idea of ‘Those who say that they feel poor represent
subjective poverty’ (Bradshaw & Finch, 2003, p.516). But there are also measurement of
subjective poverty using quantitative method suggested in 1970s (Goedhart, Halberstadt,
Kapteyn, & vanPraag, 1977) and developed into various approaches in later years (Colasanto
et al., 1984). There were two main approaches in measuring subjective poverty, the subjective
poverty line (SPL) and the Leyden poverty line (LPL). The setting of SPL based on survey
questions asking the minimum income question (MIQ), for example, the amount of absolute
50
minimal for making end meet per month (Kapteyn et al., 1988), or asking the level of income
for buying necessities per week (Bradshaw & Finch, 2003). The benchmark with consensus
may use as the threshold to define who are the people in poverty. The LPL is bases on income
evaluation question (IEQ), which asking the people to judge their circumstance, by answering
‘very good’, ‘good’, ‘sufficient’, ‘insufficient’, ‘bad’ or ‘very bad’ (Kapteyn et al., 1988).
Subjective approach is increasingly applied for measuring quality of life and evaluating
social progress and impact of social policy (Crettaz & Suter, 2013). Moreover, the subjective
poverty approaches were well applied in researches across countries, including European
countries, United Stated and developing world, in the past decades (Gustafsson, Shi, & Sato,
2004).
For determinates of subjective poverty, the household background, including age,
household composition and consumption were found as crucial determinates of subjective
poverty (Mahmood et al., 2018). Other important determining factors of subjective welling
includes sex, age, marital status, health condition, education, (Helliwell, 2003; Kingdon &
Knight, 2006). Food, exclusion from society, social inequality are also determinants of
subjective poverty (Ibrahima, 2013). In addition, geographic factors are also important as
people normally compare with others living in the same or nearby district. The assessment of
poverty can be very different among districts and cites. Furthermore, employment status also
affect the self-assessment, for example, unemployed person may think their families deserve
51
less support than other families (Gustafsson et al., 2004)
Discussion. The subjective approaches highlighted the subjectivity of the people which
is normally ignored or overlook by other approaches. Not only it challenged the traditional
material approach in defining poverty, but also reveals the subjective and psychological
nature of human well-being. It also recognizes the judgement and assessment of people.
However, it was criticized that this approach is not objective and hard to make
comparison among cities and nations (Kwadzo, 2015). The assessment of people themselves
can be bias in their social conditions and neglect some objective situation. For example, a
member of middle-class family may be feeling poor compared with the rich communities he
lives in but neglecting the high amount income or capital he owned compared with the whole
society.
e) Capabilities approach
Definition and notion. Amartya Sen was the pioneering and the most influential scholar
in proposing the capabilities approach. He criticized the income approach with an example of
two cases. One person has lower income and the other has higher income but has kidney
problem, who is poorer? As the man with health problem may suffer from high medical
expense and has economic difficulties, moreover, has worse functioning, so in some sense he
may be poorer than the one with lower income. Thus, besides income approach, other set of
measurement is needed (Sen, 1992). Sen also questioned the meaning behinds income and
52
economic resource. They only had instrumental meaning to human but not intrinsic values.
He tried to propose an approach that highlighted the importance of human capability to
achieve individual freedom, instead of utility measured by money or income.
In Sen’s influential book ‘Development as freedom’ , he illustrated the ideas of capability
approach more comprehensively (Sen, 1999). There are five distinct types instrumental
freedoms, including (1) political freedoms, (2) economic facilities, (3) social opportunities,
(4) transparency guarantees and (5) protective security. Each of these types of rights and
opportunities can help to advance the general capability of a person to live more freely (Sen,
1999, p.38). In chapter four Sen suggested ‘poverty as capability deprivation’. He highlighted
this approach concentrate on deprivation of ‘intrinsically important’, instead of income or
money which is instrumentally significant. He mentioned ‘Capability is thus a kind of
freedom: the substantive freedom to achieve alternative functioning combinations’ and
‘While a combination of person’s functioning reflects her actual achievements, the capability
set represents the freedom to achieve’ (Sen, 1999, p.75). He suggested freedom-based would
be the best approach, ‘the freedom-based perspective can take note of, inter alia,
utilitarianism’s interest in human well-being, libertarianism’s involvement with processes of
choice and the freedom to act and Rawlsian theory’s focus on individual liberty and on the
resources needed for substantive freedoms’ (Sen, 1999, p.86)
Another influential scholar promoting capability approach is Martha Nussbaum. In her
53
seminal book ‘Women and human development : the capabilities approach’ (Nussbaum,
2000). She proposed a list of ‘central human functional capabilities’ and emphasized that
‘The list represents the result of years of cross-cultural discussion, and comparisons between
earlier and later versions will show that the input of other voices has shaped its content in
many ways. Thus it already represents what it proposes: a type of overlapping consensus on
the part of people with otherwise very different views of human life.’ (Nussbaum, 2000,
p.76). Nussbaum (2011) highlighted that the central capabilities have irreducible
heterogeneity and all elements are distinctive. But she added that ‘the list is a proposal’
(Nussbaum, 2011, p.37) and open to be contested that some items may not be so central. She
also suggested there are some social conditions which it may be impossible to fulfil all the
capability of everyone, for example, in the extreme poverty area, people need to give up
certain capabilities to survive.
Discussion. It is an absolute approach in a relative position of societies (Wagle, 2002).
‘Relative deprivation in the space of incomes can yield absolute deprivation in the space of
capabilities’ (Sen, 1992, p.115), Sen made an example, in a rich country, more income is
needed to purchase health service and education, to achieve same amount of functioning. The
capability approach provides put a deeper understanding of human life, not only focus on
physical needs but with dimension of human freedom (Wagle, 2002). On the other hand,
scholars like Alkire (2005) suggested that the non-specific nature of capability approach can
54
provide more space and choice for the diverse nature of different society. He suggested there
need not a unique list of capabilities, instead, to apply in different level with different
problem, the operationalization of capabilities approach can be varied. It may ‘more art than
science’.
f) Social capital approach
Definition and notion. There are several concepts of capital developed by scholars in
the past decades, including physical capital, human capital, economic capital, and social
capital. For example, the concept of human capital was first developed in 1960s by
economists, Theodore Schultz (1961) and Gary Becker (1964), for measuring the value of
skills of workers. On the other hand, the concept of physical capital, including material,
machines, and tools, had extended in the educational field as human capital since 1960s
(Coleman, 1988). And economic capital, in modern society, was development to describe the
accumulation of money, or houses, stocks (Field, 2008).
Acknowledged Glen Loury’s concern of social inequality and intergeneration poverty
(Loury, 1977, 1981), Coleman put attention to the effect of social capital in creation of human
capital in next generation (Coleman, 1988). Coleman defined three forms of social capital: 1)
obligations, expectations, and trustworthiness, 2) information channels, and 3) Norms and
effective sanctions. His research showed lacking social capital may result in dropping out of
school (Coleman, 1988). Coleman linked up individuals with social structure and proposed
55
some social structures are more likely to facilitate persons to make better choice, for instance,
privileged family background provided more information for their children. (Coleman, 1988;
Gray & Simpson, 1998).
Robert Putnam used social capital as analytical framework to research the social
relationship in societies and made influential impact not only in academic field but also in
government policy setting (Putnam, 1993). He used social capital to conceptualize
characteristics of social life including trust, norms and networks, to analyse the social
relationship in the society (Field, 2008; Putnam, 1995). Putnam suggested there are different
forms of social capital, such as bonding and bridging, bonding capital refers to socialize
people with similar characteristics, bridging capital refers to connect people with other
strengths and capital.(Putnam, 2000).
Discussion. Although social capital was widely used, there are several risks in the
definition process and for application in the society. One core problem of using social capital
is that the concept is too board and not precise (Oyen, 2002). There were too many
definitions of social capital, developed by different scholars or institutions, but there was not
consensus for a concrete definition. Moreover, it was criticized that the definition process of
social capital was usually without gender perspective (Field, 2008; Morrow, 1999). As the
definitions can be varied, it may be difficult to distinguish the nature of social capital as cause
or effect of poverty, or as a means or an end (Portes, 1998).
56
On the other hand, social capital can be bad for the poor. Not all social networks are
beneficial for the poor. Sometimes poor people may tie together too tightly so they cannot
build other networks. It may call ‘negative social capital’ of the poor people (Royce, 2015).
There are several negative effects of social capital, for instance, exclusion of outsiders,
limitation of individual freedom in community, excess use of resource of other members,
maintaining unfavourable norms, prohibiting creativities (Portes, 1998). In macro level, Oyen
(2002) suggested the reappear of the concept of social capital matched the agenda of reducing
the role of government and public expenditure. Using the concept of social capital can be
risky if it promotes viewing poverty as individual deficit (Warren, Thompson, & Saegert,
2001).
2.1.3 Short summary
Poverty is in nature multidimensional. Scholars attempted to create different frameworks
to broaden the understanding of poverty, but neither one approach can fully describe the
poverty phenomenon. The explanation of poverty bases on these approaches are also
diversified, not only because of the multidimensional nature of poverty, but also based of
different ideologies. Similarly, there is no single theory can give a comprehensive
explanation, or they can explain distinctive part of the poverty phenomenon. Generally, they
can be divided into individual and structural dimensions for easier understanding. It needs
research to answer whether which one dimension has the greater impact on poverty situation.
57
Monetary approach is the most dominating. One of the reasons is that it is easy to
measure and to compare. However, it normally neglects the social and subjective dimensions
and other material inadequacy. The concept of deprivation attempted to tackle these
problems, by measuring the items of material and social deprived. On the other hand, the
concept of social exclusion highlighted the poverty process instead of the static view of
poverty. The subjective approach emphasized the importance of subjective view in defining
well-being and subjectivity of human. These three approaches also provide well established
measurement scales. The social capital approach provides a broad framework in
understanding poverty. The capability approach in core asking the ultimate well-being of
people. However, these two approaches were too vague in definition and difficult to measure.
To limit the study scope of this thesis, this study will focus on the approaches of deprivation,
social exclusion, and subjective poverty to investigate the poverty situation and to extend the
theories of poverty.
The definition and explanation of poverty is a continuous process, which it needs to
refer to the corresponding social and economic context. But the previous discussion neglects
one crucial factor which is influential in conceptualization and explanation of poverty. It is
housing. There are some elements of housing mentioned in those theories, but they cannot
reveal the importance of the impact of housing factors on different dimensions of poverty. In
the following sessions, the concept of housing and housing poverty will be introduced.
58
2.2 Housing
2.2.1 Nature of housing
‘Housing is not only a necessity of life; it has a pervasive impact on all aspects of our
existence’ (Stone, 1993, p.1). Michael Stone highlighted the complicated nature of housing
which related all aspects of human life. Housing, with the simplest understanding, is a
physical space with walls, which aiming for people to live in. However, there are diversified
natures and meanings of housing, for instance, for meeting basic needs, building social
relationship, as a commodity or collective consumption, as culture or ideology.
a) Housing as physical living space and satisfying basic needs
Much of the human basic needs, such as sleeping, building family relationship, daily
eating and drinking, need a venue to be satisfied, and housing provide such place basic living
(Stone, 1993). Physically, housing can provide human being for security and privacy from
disturbance. Most people in the world, across various civilization and culture, need a housing
for basic living. The physical design of housing can be varied, with different size, facilities,
and settings, but generally, housings are physically separated from outside with walls, roofs,
and doors, protecting human being from natural disasters and contingencies.
On the other hand, housing is not only a physical building, it also provides a home for
individuals and families. Home means different with and more than housing, especially in a
social dimension. For example, a home normally refer to a human relationship inside a house,
59
or a home provide basic necessities to individuals for daily living and also served as storing
our memories (King, 2015; Stone, 1993). King (2015) proposed that the word ‘house’ is cold
and empty but the term ‘home’ contain much emotion and possibility.
b) Housing as social relationship and identity
Housing can be more than a home. It provides resources and opportunity for human
development and social life and influences the access to school, employment, social service
and community facilities (Malpass & Murie, 1999). For example, people normally find
schools or jobs nearby their living place for convenience and easier transportation. The
neighbourhood of a person is mainly determined by the area they are living, and this highly
affects their social network and relationship.
Moreover, the type of tenure can also reflect the social status and influence the
development of personal identity (Malpass & Murie, 1999) and ‘Home is the principal locus
of family life and social reproduction’ (Stone, 1993, p.22). For instance, normally the social
economic status of a person is not easily known from their appearance, but their living place
can reflect it. The people living in poverty are hard to live in the luxury houses or the high-
class residential regions. On the other hand, people living in public, social or subsidized
housing may normally with lower income or asset. The identity of a person, whether he or
she is rich or poor, upper, or lower, not only determined by their income, but also by their
living place. However, it is reminded that the information can be limited, there are still
60
variation within each type of tenure (Malpass & Murie, 1999).
c) Housing as commodity
Housing can be viewed as a commodity. For a commodity, there are two values within,
the use-value and exchange value. For use-value, people treated housing as a living space, to
satisfy their basic need for dwelling. The physical dimension focus on the use-value of
housing, which can be reflected by the subjective views of the household, for example,
satisfaction of living, or some objective criteria of living standard. For exchange value,
normally housing get a price, especially for private housing. There are many factors affecting
the housing price, such as the demand from public, the supply of private housing, the housing
policy and politics, the speculative culture, and the flow of capital in housing market.
However, there are contradictions among these two dimensions (Harvey, 2014). If housing is
only for use but not for exchange, there will be not so many housing problems or issues. The
exchange value of housing may rise or drop sharply because of economic or political reason,
when the price of housing rises too high, the used value of housing with same price as before
will be affected. For example, the used value of a house priced one million dollars in 10 years
ago can be much higher than the house with same price in the present.
Housing is different from normal commodity. Housing can be a long-term investment
asset, rather than one-time consumption. Most probably, it is the most expensive item a
household purchased in the whole life. In addition, the cost of this item can be varied and no
61
one can actually know the final cost of it (King, 2015). The price can be changed vigorously
by numerous uncontrolled factors including interest rates, the economic situation of other
countries and believes of housing value of other people. The housing market can be failure
for regulate the housing price with several reasons. One is about the monopoly of housing
development. When the supply of housing is controlled by few parties, the price can be
manipulated since the housing choice in the market is limited. The other reason is the culture
of speculative. When substantial amount of people magnified the exchange value of housing
and believing continue rising of housing price, housing can be treated as an investment tool
rather than space of living. The housing price may rise sharply not depending of the supply
and demand, but the believe in investment value.
d) Housing as collective consumption
Michael Ball (1986) proposed the consumption perspective for housing study.
Development of housing involves the process of production and consumption. As the
production process and material are monopolized by the developers and company, most
citizens cannot build houses themselves. The dwellings become almost necessarily
commodities for citizens and people need to act as consumers for buying or renting houses
(Ball, 1986; Malpass & Murie, 1999). As most people need houses to satisfying their basic
needs, housing serves as a collective consumption in different cities.
The collective consumption of housing not only involves the physical material of
62
housings. It was proposed that there are different types of government subsidies for housing,
for example, financial support of mortgage, building of infrastructure, facilitation of
urbanization and direct provision of social housing. These subsidizations lead generation of
huge consumption in the society (Smart & Lee, 2003b). There are other economic
considerations of housings. For example, the housing cost of households directly affect their
consumption, with high level of mortgage or rent, the residents are not able to spend much on
other goods and services (King, 2015).
e) Housing as capital (accumulation)
Housing involves huge flow and accumulation of capital. As housing is the largest part
of consumption for most families, it necessarily relates to vast amount of money, for
example, the rent or mortgage for every resident, and the construction of housing normally
generates substantial amount of labour works and business. The business of housing not only
involve the buyers and the sellers or the landlords and renter, but with numerous sectors, such
as the real estate business, the real estate agencies, the banking system for mortgage, the
lawyers responsible for trading (King, 2015). The capital flow not only limited in the trading
of housing. The pursuit of housing simultaneously drives other housing consumption, for
example, internal decoration, home furnishing, and housing insurance.
As the rise of housing price implying the capital gain for the homeowners, some people
buying houses not for living but for trading. Historically, the rate of capital return by housing
63
was higher than economic growth across nations (Piketty, 2014). Thus, housing served as a
tool of capital accumulation instead of living. It may be ridiculous that there is ‘lack of
housing’ but with housing vacancy in the same city in a certain period. One of the reasons is
the problem of hoarding of houses. Real estate companies or the rich can buy and hoard
housing units for speculative investment. Even they do not rent out the vacant houses for
money, the rise of housing price can generate capital for them. It was suggested that there are
two types of market, space market and the asset market. Economists normally concerns the
development of space market previously, however, it was found that the asset market is much
more influential in the past years (Lee, 2014). On the other hand, construction of housing
units needs use of land and the land sales usually contribute substantial revenue to the
government
f) Housing as ideology
The housing issues can reveal many ideological debates, or the meanings of housing and
allocation of housing resource can be very varied based on different ideology. For example,
whether the residents have rights to live in the cities, the ratio of private and public sector of
housing, the weighting of use-value or exchange value of housing. The ideologies of housing
not only affect the definitions of housing nature, but also the implementation of housing
policy and hence the housing situation of residents (King, 2015).
With left-wing or social democratic ideas, housing is a basic human right to all citizens.
64
Government is responsible for providing adequate housing, with reasonable size and
affordable price, to all residents (Hartman, 1998; Yung, 2008). With prominent level of state
intervention and market regulation, the government targets at controlling the housing price
from rising vigorously or provides enough public housing for the grassroots or those who
cannot afford housing in the market. On the contrast, the dominating ideology of neo-
liberalism or right-wing since 1980s highly affect the production and consumption of
housing. With the doctrine of ‘free market’, the ideological power urges for minimum
government intervention in housing market. The believes of house owners may also reinforce
the right-wing ideology as they are the most benefited by the rise of housing price.
g) Housing as governance and politics
‘Housing is politics’ (Madden & Marcuse, 2016). The housing issue involves many
parties to strive for their own interest and there are necessarily political struggles. For
example, the upper-class homeowners desire rising of housing price for capital gain, but the
lower-class renters wish the opposite for less spending on housing, or whether the land and
housing should be distributed for public use or privately sell is always controversial.
Moreover, housing activities also refer to legislation, financing and regulation policies (King,
2015). All these are related to politics.
On the other hand, the state also makes use of housing as a tool for governance. The
high level of home ownerships may enhance the stability of society since the homeowners’
65
fear of housing price drop in restless political environment. In addition, the provision of
public housing also helps in relieving the social pressure from grassroots. However, in
capitalist society, the government not only concerns the welfare of citizens for building
legitimacy, but also protect the business of private company for enhancing economic growth.
Thus, there is continuous political combat for housing issues.
2.2.2 Theories or approaches for housing study
There was long discussion of the housing theories or using theories for housing analysis,
for example, the Jim Kemeny’s seminal work ‘Housing and Social Theory’ (Kemeny, 1992),
the housing class discussion (C.Bell, 1977; Rex, 1971) and critique for class analysis
(Saunders, 1984), comparative research approaches from convergence to divergence
(Kemeny & Lowe, 1998) principles and nature of housing (King, 2015, 2017) and placing
housing in political economy analysis (Aalbers & Christophers, 2014; Ansell, 2014). Housing
scholars, like Peter King (King, 2009) and Jim Kenemy (1992, 2001) insisted that the
importance of theory building in the field of housing, rather than borrowing sociological or
political theories for analysis of housing. However, It is argued in the academic field, housing
studies are still lack of its own theory and methodologies (Clapham, 2009; O’Neill, 2008).
Instead of well-established theories, there are several approaches proposed by scholars for
housing study.
David Clapham (2005) summarized various housing study approaches. The most
66
common approach focusses of analysis of government housing policy, legislation, and
institutional process. Another approach put attention on economic analysis of simply supply
and demand of housing and the impact on housing price, normally with neo-liberal
assumption of free market. The geographical perspective studies the human behaviour and the
interaction with the spatial factors, for example, the choice of housing district and the
mobility of households. Housing studies with sociological approach apply social theories in
analysis of housing situation, with focus on resource distribution, class analysis and social
inequality (Clapham, 2005). Besides the dominant positivism approaches in housing studies,
the subjective approach focusses on the perceptions, actions and attitudes of residents and the
influence of discourses. For example, the study of decision making for moving home, the
factors of choosing houses, the assessment of housing quality, thought of community and
neighbourhood, the views of residents for housing and the experience of living. The
subjective tradition normally corelated with postmodernism, symbolic interactionism and
social constructionism (Clapham, 2005).
Another commonly used approach for housing study is regulationist theory (Forrest,
2003; Lee, 1999, 2014; Smart & Lee, 2003b). Regulation approach assumed the capitalism
itself lacks principles of regulation and may generate social crises in ways of economic
development, production and consumption, power between capitalist and labour. The crises
need external social and political forces and state intervention for regulation to promote
67
continuation of capital accumulation. (Aglietta, 1979, 1998; Dunford, 1990; Smart & Lee,
2003b). For regulationist, housing can maintain the stability of economic growth and enhance
capital accumulation. Since economic growth in capitalist society need support from
consumption, the provision of public housing not only generate employment opportunities for
construction, but also helping residents for saving with less rent and promote other
consumption (Castells, Goh, & Kwok, 1990; Lee & Yip, 2006; Smart & Lee, 2003b).
Regulation theory suggests there may not have a unique way for promoting capital
accumulation, but there can be various answers to tackle the crises. (Smart & Lee, 2003b)
In the following, the macro context of housing issue will be mainly analysed by political
economy approach since the nature of housing is highly related to politics and economic
discipline. The micro analysis of human interaction and subjective perception of housing are
suggested using social constructionism for study and will be discussed in another segment.
2.2.3 Political economy of housing
a) Global political economy
O’Brien & Penna (1998) suggested there were three focus in contemporary political
economy, namely globalization, work and welfare, new socio-political division. The
complicated process of globalization involves huge change in mode of production and
consumption, technology level, transportation, communication, division of labour, economic
structure, and mode of financialization. Under globalization, no cities or countries can live
68
without the connection with other parts of the world, instead, the economy and political
situation are highly affected by the global trend. For example, the work and welfare, the past
Keynesian welfare states were more depends on managing the national economy and job
provision, but they were challenged by the new mode of political economy under
globalization. The state finances and labour supply were no longer highly controlled by the
government. Moreover, there was growing gap from the rich and the poor, not only regional,
by national and global (O’Brien & Penna, 1998).
One feature on the contemporary political economy is the shift from industrial to post-
industrial age. The discussion of post-industrialism mainly started since 1970s, for example,
Bell (1973) suggested the economic structure had shifted from manufacturing to tertiary
service and the change in science and information technology was the main engine for the
transformation. Touraine (1971) also highlighted the role of technology for social change but
put more focus on the social division caused by the transformation and resulting in more
social conflicts (O’Brien & Penna, 1998). Another similar feature was the shift from Fordist
to Post-Fordism. It mainly studied the change of Fordist organizations, which focus on
massive production and consumption, and national economy with Keynesian approach. The
new mode of Post-Fordist suggested a regime with more flexibility and segmentation, in
production, consumption and the role of state. For example, the jobs became less secure and
with more in part-time mode, and the state aimed at solving social crisis rather than social
69
planning (Jessop, 2002).
There was a global turn to neo-liberalism, at least started at the late 1970s, and widely
spread across nations since the governing of Ronald Regan in US and Margaret Thatcher in
UK. It was suggested that the rise of neo-liberalism was a reaction to the rapid expansion of
social welfare and economic crisis in 1960s and 1970s. It was also a revival of upper class for
regaining their power since World War II. With the assist of right-wing think tanks, media and
International organizations like IMF and WTO, the neo-liberalism diffusely proliferated and
became dominant over the world (Harvey, 2005). The idea of neo-liberalism suggested
minimum intervention from the government and promoting privatization of public service
and commodification of public goods. Under this idea, market is the best institution for
satisfying human needs and distribution of social resource. Inequality is a natural outcome
and impossible to be eliminated. (Lavalette & Pratt, 2006; Lister, 2010).
b) Political economy of housing
There was a long tradition of discussing housing issue in political economy approach. It
can be traced back to Friedrich Engels’s work ‘The Housing Question’ in 1872, criticizing the
economy logic of capitalism for causing rapid increase of rent in the city (Engels, 1975
[1872]). David Ricardo’s work ‘On the Principles of Political Economy and Taxation’ in the
early 19th century (Ricardo, 2001 [1817]) was also crucial for contributing ideas for political
economy of housing and land, such as the theory of rent, study of rent-seeking behaviour and
70
using land as the main element of production.
Generally, the political economy perspective challenges the classical economics theory
which focus on demand and supply of housing but neglecting other political and social
factors. It is criticized that the classical economics theory cannot provide comprehensive
explanations for the volatile housing market and housing deprivation. The housing market
fails to regulate the housing price and resource. The housing problems are not only related to
economy but also interests of different parties and the social structure. Investigation about
housing problems not only involves study of the social, economic and political dimensions of
housing, but also the interaction among these aspects (Boyer, 2000b; Smart & Lee, 2003b).
Scholars suggested it is essential for placing housing in political economy analysis in
capitalist society since housing in capitalism is with various roles such as circulation of
capital, constituting social relation and as ideological institution (Aalbers & Christophers,
2014; Clarke & Ginsburg, 1975). The housing problems cannot be simply explained by
policy or market approaches but need to be analysed with political consideration and
dynamics of social relations. Moreover, the housing problems are hugely affected by
ideological change, for example, the neo-liberal turn of housing policy deeply affects housing
situation and price, like in Britain, the policy implemented by Margaret Thatcher since 1980s
is one of the main factors for high housing price nowadays.
In recent years, under the influence of expanding flow of capital across nations, there
71
was growing discussion of residential capitalism (Schwartz & Seabrooke, 2008) and
financialization of housing (Fernandez & Aalbers, 2016; Smart & Lee, 2003a), cantering
housing of more crucial roles for capital accumulation, economy growth and state financing.
For example, when the housing is much more commodified, the large capital flow from
China may influential for the rising of housing price in London and this phenomenon was not
limited in certain cities but worldwide (Minton, 2014).
d) Role of state, housing policy and welfare state
It is impossible to neglect discussion of role of government and policy for housing study.
‘Housing policy can be defined in terms of measures designed to modify the quality, quantity,
price and ownership and control of housing’ (Malpass & Murie, 1999, p.7). Policy design and
implementation influences the supply, finance, management, and culture of housing, and are
crucial for affecting the livelihoods of every resident. There are different types of housing
policy, such as legislation, like the Housing Act in 1988, Rent Act in 1957, Housing Finance
Act in 1972 in Britain, taxation policy and rental policy (Malpass & Murie, 1999). For
example, by controlling the supply and demand of housing or land, the housing market, price,
and rent can be regulated in certain extent. Or with legislation for minimum living standard,
the housing quality can be improved. The Influence of housing policy to other social areas is
also critical, for example, the high housing price and small living area may result in later
marriage and fewer children, the increase of cases of divorce or separation may raise housing
72
demand (Forrest, 2003). Moreover, the promotion of home-ownership may enhance wealth
accumulation but also widen the gap between the rich and the poor (Forrest, 2003).
Moreover, the housing issues in nature is political, the role of state in the capitalist
society is usually ambiguous. The ideology behinds may directly the rationale of policies. For
example, for the view of neoliberalism, it is not the housing problems require state
intervention, but the government intervention creates the problems (Malpass & Murie, 1999).
It was also argued that the role of government in housing became less crucial in the past
decades since the trend of privatization initialled by globalization over the world, the ‘right to
buy’ housing policy in Britain since 1997 is one of the examples (Clapham, 2005). Thus, for
housing policy study, it is essential for investigating the political environment or ideological
development in the area.
On the other hand, housing accounts not only the largest budget for households, it also
one of the main pillars of welfare state (Kemeny, 2001). It is debatable whether the
government is responsible for provide housing for all the residents or to what extent should
the government intervene the housing market. Kemeny (2001) attempted to compare the
different of nature among housing and other welfare pillars including health, education and
social security. For example, social security mainly distributed cash transfer and more
targeted on the socially disadvantaged, the spending on education and health care service
mainly used for salaries of professionals such as teachers, doctors, and nurse. The main
73
difference of housing from others is the nature of capital intensive, not only includes the
capital investment on housing stock, but also the housing finance and taxation. The role of
housing in welfare state and welfare system is not clear, however, need more studies
(Kemeny, 2001). A study showed the poverty reduction effect of the housing system can be
independent of the welfare regime in the same country (Stephens & vanSteen, 2011). More
researches are needed for studying the relationship between housing system and welfare
regime, or whether there are housing regimes across countries (Doling & Ronald, 2014;
Kemeny, 2006).
74
2.3 Housing poverty
This part discusses the concept of ‘housing poverty’ and the closely related term
‘housing affordability’. Various definitions and their limitations will be explored. In addition,
the theories of relationship between housing factors and poverty situation are also illustrated
in this session.
2.3.1 Meanings of ‘housing poverty’
The term ‘housing poverty’ is not frequently used in academic literature, instead, there
was substantial amount of literature discussing ‘housing affordability’ (Bramley, 2012; Li,
2014; Stone, 2006b; Whitehead, Monk, Clarke, Holmans, & Markkanen, 2008). Generally,
the ‘housing poverty’ means ‘housing-induced poverty’ (Kutty, 2005), or ‘poverty caused by
housing factors’ (Stephens & vanSteen, 2011). Some scholars used this term with similar
meanings with ‘living in unaffordable housing’ (Davies, 2008), or the housing cost exceed
certain threshold (Sato, 2006). Other scholars used other similar term ‘house poverty’ (Kutty,
2002; McConnell, 2012; Thalmann, 2003) or ‘shelter poverty’ (Stone, 1993), meaning the
income after housing cost is insufficient for daily living. Broadly speaking, When discussing
‘housing poverty’, it also refer to the relationship between housing factors and poverty
situation (He, Liu, Yang, & Wang, 2017; Stephens & Leishman, 2017; Wallace, 2016) , or
reducing poverty situation by housing means (Clarke, Morris, & Williams, 2014; Gilbert,
2014).
75
The term ‘housing poverty’ used in this thesis refers to the broader meanings ‘housing
induced poverty’ or ‘housing related poverty’. As discussed in the previous session, there are
many dimensions of poverty, implying various perspectives of ‘housing poverty’, including
housing induced income poverty (Fusco, 2012; Maestri, 2015; Saunders, 2017), housing
induced deprivation (Borg, 2015; Guio & Maquet, 2007), housing induced social exclusion
(Arthurson & Jacobs, 2003; Somerville, 1998). ‘Housing affordability’ usually refers to
housing induced monetary poverty, but ‘housing poverty’ need not limited in monetary
approach. A broader definition of housing poverty also includes studies of relationship among
housing quality and income (Lelkes & Zólyomi, 2009) or mental health (Evans, Wells, Chan,
Saltzman, & Saltzmand, 2000), living environment and quality of life (Ng, Zhang, Ng, Wong,
& Lee, 2018), housing factors and employment (JRF, 2015).
Nevertheless, like the researches in poverty field, the economic or monetary approach is the
most dominant perspective in literature. The most researched area of ‘housing poverty’ also
focus on housing related monetary poverty. The concept ‘housing affordability’ was used
mostly.
2.3.2 Housing affordability (monetary ‘housing poverty’)
Housing affordability was one of the most dominating issue in housing research in the
past decades (Bramley, 2012; Linneman & Megbolugbe, 1992). Under the concept of
‘housing poverty’, ‘housing affordability’ is the study of monetary ‘housing poverty’. The
76
most common idea of housing affordability is the relationship between household income and
housing cost. But there was continuous debate for the conceptualization and
operationalization of ‘housing affordability’.
a) Conceptual definitions of housing affordability
Before asking the question ‘What is housing affordability?’ It is more fundamental to
ask, ‘Why should we concern housing affordability’? One of the core reasons is that it relates
or affects the livelihood of people. When we talk about housing is expensive or cheap, it
needs objects to refer to, for example, affordable for whom. Michael Stone, in his seminal
work ‘Shelter Poverty: the new idea of housing affordability’, proposed ‘affordability
expresses the linkage between the well-being of individual families and the mechanisms of
housing provision and income determination.’ (Stone, 1993, p.1). He continued the concern
livelihood of people in defining affordability, ‘it is an expression of the social and material
experiences of people, constituted as households, in relation to their individual housing
situations. Affordability expresses the challenge each household faces in balancing the cost of
its actual or potential housing, on the one hand, and its non-housing expenditures, on the
other, within the constraints of its income.’ (Stone, 2006, p.152). Or in the recent report, it
suggested ‘affordability (and lack of affordability) is not an inherent characteristic of a
housing unit—it is a relationship between housing and people. For some people, all housing
is affordable, no matter how expensive; for others, no housing is affordable unless it is free.’
77
(Stone, Burke, & Ralston, 2011, p.2). All these definitions highlighted the importance of
residents’ situation in the conceptualization.
The other concern is the relationship with housing cost, normally with some normative
standards. It highlights the relationship of household income and housing cost. Maclennan &
William (1990, p.9) made a general definition, ‘Affordability is concerned with securing
some given standard of housing (or different standards) at a price or a rent which does not
impose, in the eyes of some third party (usually government) an unreasonable burden on
household incomes.’ Moreover, it not only refers to the actual cost paid by the households,
but the opportunity cost of housing, like Hancock’s (1993, p.129) idea, ‘concerned with the
notion of the opportunity cost of housing, and clearly this is the essence of the concept of
affordability…also concerned with the standard of housing consumption’. However, whether
we should use the actual cost, or the opportunity cost of housing revealed a great debate of
‘ratio-income’ and ‘residual income’ approach in measurement, which will be discussed later.
Moreover, housing affordability concerns the housing quality issue, like Arnold &
Skaburskis (1989, p.502) discuss in the early years , ‘Housing affordability … can be defined
with reference to the nature and quality of the housing delivered at the price households are
able to pay.’ The consideration of housing affordability is hard to be separated from housing
quality or housing standard, since household may live in very small area or housing with
inadequate facilities to squeeze their housing cost, their housing cost maybe low but this does
78
not mean they are living well (Stone et al., 2011). The discussion of housing affordability also
encompasses many other issues, such as interaction of land costs, land use pattern, housing
finance, repeated sale and resale, housing market and poverty (Stone, 1993; Yip, 1995).
Overall, Stone, Burke & Ralston (2011) summarized six approaches for defining
housing affordability. The first type is ‘categorical’, or maybe tautological, denote
affordability is ‘the ability of people to secure housing based on their ability to pay’. The
second on is ‘relative’ approach, usually based on housing cost, accessing the ability of
potential homebuyers for payment. Median multiple measure is one of the examples. The
third one is ‘subject’ approach. It assumes all the households are rational and should
maximize their utility. All of them will pay for housing which just affordable for them. It has
no general meaning. The fourth ‘family budget’ approach considers both normative standard
and actual spending of household. It assumes there was a socially and historically determined
minimum or adequate standard of living. It used the budget standard methodology to specify
the physical standard of housing, but without monetary calculation. It is the base of the ‘ratio’
approach. The fifth and sixth approach are ‘ratio’ and ‘residual’ which receive widest
recognition and with most discussion for accessing affordability. These two approaches will
be discussed in another session. The concept of housing affordability was commonly used in
society. David Hulchanski (1995) identified six common usage of the concept of housing
affordability, including description of housing cost, social trend analysis, defining eligibility
79
of housing policy and housing need, prediction of household ability for paying rent or
mortgage, and selection criteria for housing decision. For example, with the approach of
housing-induced poverty, rental subsidies are effective for reducing poverty, especially for the
renters near the official poverty line (Kutty, 2005).
b) Measurements of housing affordability
In the past decades, scholars had developed various measurements of housing
affordability, with lengthy debate and discussion of the merits and limitations and typology of
measurement. Yip (1995) defined three main approaches in measuring housing affordability,
normative, behavioural and subjective. The ‘normative approach’ is the most used method. It
defines a threshold or norm for affordable or non-affordable, for example, ratio of housing
cost to income, or level of income after housing cost. Housing quality or core housing need
measurement is two of the variations in this approach, with normative standard of adequate
housing. The ‘behavioural approach’ focus on the housing decision of and exhibit problems
of households, for example, measuring the consumption of good and service against housing
cost to determine a threshold of affordability, like S-curve of measuring poverty. The
‘subjective approach’ put attention on the subject experience of households on how difficult
they feel for paying housing cost, Likert scale can be used for measurement (For more detail,
see Yip, 1995, session 3.4).
For normative approach, Yip (1995) used three dimensions, housing expenditure, non-
80
housing expenditure and housing income, to classify whether the situation is affordable. For
example, for a household with high level in these three categories, it is classified as
affordable. Or for a household with low income with low non-housing expenditure, it was
count as poverty. But if the low-income household has high level of non-housing expenditure
but with low housing cost, it was unclassified because the housing cost maybe squeezed
(Table 2-1). On the other hand, Yip (1995) also made a table for comparing poverty before
and after housing costs. When the households are poor in these 2 dimensions, they are count
as ‘poverty’. However, if they are not poor before housing cost, but are poor after housing
cost, it counts as ‘unaffordable’ (Table 2-2).
Table 2-1 affordability classification: income and expenditure
Housing Non-housing Household Classification

Expenditure Expenditure Income
1 High High High Affordable
2 Low Low Low Poverty
3 High Low Low Poverty
4 High Low High Unaffordable
5 Low High Low Unclassified
6 Low High High Affordable
Note: Adapted from Yip (1995, p.82)
Table 2-2 affordability classification: poverty before and after housing costs
Poverty Before Housing Poverty After Housing Costs

Costs
Poor Not Poor
Poor Poverty Unclassified
Not Poor Unaffordable Affordable
Note: Adapted from Yip (1995, p.83)
81
Another scholar Hancock (1993) proposed other 4 operational measurement for housing
affordability: 1) ‘Minimalist’ definition. Housing is unaffordable if the ‘housing price (H)’
AND ‘non-housing consumption (Y)’ are both less ‘the socially desirable minimum standard’
(H* and Y*) (Area A in Figure 2-1); 2) ‘official’ definition. Housing is unaffordable if the
‘income after housing cost (M - H)’ is less than the ‘socially desirable minimum standard of
non-housing consumption (Y*)’, where there is an upper limit of housing cost H < Hmax; 3)
‘residual income’ definition. Count the income after housing cost, with a minimum standard
of housing cost. 4) ‘generous’ definition. Count the income after housing cost, with a lower
limit of non-housing cost which set at 140% of certain socially acceptable income amount
(Figure 2-1).
Figure 2-1 A minimal definition of affordability
Note: Adopted from Hancock (1993, p.129)
There are numerous approaches used in studying housing affordability. Li (2014)
reviewed the academic discussion of housing affordability from 1990 to 2013 with 112
journal papers in top-tier journals. He found there were three main methodologies used in the
82
past studies. They were ‘ratio income approach’, ‘residual income approach’ and ‘composite
approach’. He proposed the past studies largely focus on the normative measurement using
these three approaches, but with more attention was put on composite methods in recent
years.
c) Ratio income approach
The ratio income approach is one of the most common approaches in measuring housing
affordability (Cai & Lu, 2015; Hancock, 1993; Stone, 2006a). Generally, it measures the ratio
of household income to housing cost. A threshold of ratio is set for measuring whether the
housing cost is unaffordable. Traditionally, there was a ‘rule of thumb’ principle originated
from 19th century which using 25% as the threshold in measuring affordable, meaning ‘‘a
week's wages for a month's rent’ (Arnold & Skaburskis, 1989). For example, when the
housing cost exceed 25% of the household income, it is said the house is unaffordable to the
family. However, there were variation of thresholds of countries in different period. For
instance, thresholds of 25%, 30%, 40% and 50% had been used in America in different time
(Kutty, 2005), or Australia researchers used 30/40 rule, meaning housing cost should not
exceed 30% the lower 40% housing income group (Cai & Lu, 2015) and using 25% as social
housing rent threshold
One of the merits in using ratio income approach is the simple nature in measurement
(Cai & Lu, 2015). It is easy to understand by the public. The calculation is not complicated
83
and normally requires the statistics of income and housing cost, which is commonly available
in different cities and make it easier for comparison. There are common applications of ratio
income approach. For example, the housing price to income ratio usually used to calculate the
affordability of homeowners. Or the international organization Demographia used ‘median
house prices to median household incomes’ in comparing housing affordability across
different cities in the world (Demographia, 2017).
d) Criticism to Ratio income approach and response
There are several shortcomings for using ratio income approach. One core issue is about
housing quality. Some scholars suggested that high ratio of housing cost to income may due
to high quality of housing of individual preference, for example, a high-income family
choose a luxury housing may also result in high ratio of housing affordability (Lerman &
Reeder, 1987). Or on the other way, some households living in poverty squeeze their housing
cost, for example, living in housing with very small living space and low quality, may result
low level of housing cost-to-income ratio. But this cannot reflect their housing hardship
(Thalmann, 1999, 2003).
Nevertheless, this problem of measurement can be managed. For example, upper limits
to income and housing cost can be set, to eliminate extreme cases or those cases living in
luxury conditions, which can be assumed as not facing problems of housing affordability
voluntarily (Kutty, 2005). The other alternative was proposing a measurement with cost of
84
renting an adequate housing, with certain basic housing characteristic. For example, a house
with basic living standard rent $3000, the ratio of this housing cost to the income of
household will be counted as housing affordable ratio (Lerman & Reeder, 1987). However, it
was questioned the housing cost with similar housing features can be ranged with various
factors such as location or other market considerations. Moreover, there may be also not
enough adequate housings for meeting residential demand (Kutty, 2005).
Philippe Thalmann (1999) instead proposed other three dimensions in advancing the
measurement of housing affordability. The first dimension measures the ratio of housing
consumption with other appropriate consumption (high or low). The second was the ratio-to-
income ratio (RIR) of current housing with quality consideration. The third was the RIR for
appropriate housing in the market. Three dimensions formed 8 categories in a housing
consumption metrics. The table can help to indicate which groups of households are facing or
at risk of housing affordability problems. For example, in the table, ‘group bl’ was currently
without affordability problems (RIR <25%), but they cannot afford appropriate housings if
they moved out and it was supposed that they were under-consumption and squeeze their
housing cost. Another example, ‘group ch’ was supposed to be overconsumed, they were able
to lower their housing cost for appropriate housings, and it imply no reason to assist this
group (Thalmann, 1999).
This method was advantageous for finding those households with need of housing
85
assistance. However, Thalmann (1999) also admitted number of limitations with this
approach, such as dependence on the definition of ‘appropriate housing’, statistical limitation
of hedonic model and the unavailability of average price housing in market.
In addition, there are other limitations for using ratio income approach, for example, it
makes no absolute reference to the consumption of housing and non-housing, for example, a
household with income $10,000 and spend $3,000 on housing is different with a household
with income $100,000 and spend $30,000 on housing, as the latter case has $70,000 leave for
non-housing consumption (Chaplin & Freeman, 1999). Moreover, it is sometimes not easy to
reflect to ability of household to pay, especially for retired people, wealth is more important
than income (Thalmann, 2003). On the other hand, A clear threshold is difficult to set for all
households, for example, 30% maybe high enough for low income families but it may make
less burden for rich households
The most critical challenge to ratio income approach is that it cannot accurately reflect the
livelihood of households after housing cost. Like what Michael Stone (1993, p.35)
highlighted, ‘What most households actually pay is not what they realistically can
afford…Since a housing affordability standard is intended to measure whether a household
has sufficient resources left to meet its non-shelter needs after paying for housing’. The
residual income approach attempted to tackle this problem.
e) Residual income approach
86
The residual income approach concerns the available resource after the households paid
their housing cost. Why housing cost matter? One of the core problems of high housing costs
is the families need to squeeze their expenditure in other aspects, such as food, health,
education, necessities and other daily activities (Cai & Lu, 2015; Chaplin & Freeman, 1999;
Kutty, 2005). Stone (1993, p.34) used ‘shelter poverty’ for this approach and highlighted the
burden of housing cost on non-housing needs, ‘Households paying more than they can afford
on this standard are shelter-poor, the squeeze between their housing costs and incomes
leaving them unable to meet their non-shelter needs at a minimum level of adequacy. That is,
shelter poverty is a form of poverty that results from the burden of housing costs rather than
just limited incomes.’. This approach also recognized that housing usually makes the largest
cost and is least flexible for households (Stone et al., 2011). Thus, it used non-housing cost
instead of non-food cost or non-health cost in measurement.
Nandinee Kutty defined ‘a household that cannot afford the poverty basket of non-
housing goods after paying for housing’ as ‘housing-induced poverty’ (Kutty, 2005, p.114),
which was similar to the idea of ‘shelter poverty’ of Michael Stone (Stone, 1993). She agreed
the normative framework of Stone but questioned the base for calculating basic non-housing
cost was outdated. She used data to illustrate the difference between housing-induced poverty
and cost burden approach, for example, in 1999, a person of income $6,000 are categorized
as housing-induce poverty with 11% of income spending on housing. This method is
87
advantageous in finding those ‘housing poverty’ households, which located above the
housing-induced poverty line and below the cost burden line (red circle in Figure 2-2) (Kutty,
2005)
Figure 2-2 Relationship between Housing-induced poverty and Cost Burden
Note. Adapted from Kutty (2005, p.120)
Comparing with the concept of shelter poverty, Kutty (2005) proposed the measurement
of housing-induced poverty are consistent with the official poverty line, by using disposable
income. Hence, the later one may reflect a lower consumption level. She used US national
data to show the families living above the official poverty threshold may also have poor
living standard after housing cost. For example, the housing-induced poverty in US was 2.7%
higher than official poverty rate in 1999 (Kutty, 2005). Compare with a fixed threshold of
‘ratio income approach’, ‘residual income approach’ is a sliding scale of housing affordability
(Stone et al., 2011)
There are increasing trend using residual income approach in housing studies (Stone et
88
al., 2011). From the past empirical studies, normally, those with larger family size and with
lower household income had higher risk of unaffordability, as usually they had higher portion
of non-housing cost to income (Cai & Lu, 2015; Stone, 2006a; Stone et al., 2011).
f) Composite approach and other considerations
The debate between these two approaches had not stopped. They are with different
focuses and features, as Bramley (2012, p.134) highlighted, ‘While normative affordability
ratios reflect common sense, consumer demand theory, and many practical rent subsidy
systems, the residual income approach reflects poverty definition and measurement traditions
and the use of equivalisation scales’ to reflect household size and composition.’. And there
was an increasing trend for using composite methods (Li, 2014). Moreover, Bramley
(Bramley, 2012) used empirical data to validate housing measurement and suggested
composition of ratio and subjective problem measure were good at measuring housing
problems. The three main conditions for affordability proposed by Thalmann (2003) can
make a simple summary for composite approach, housing is affordable if 1) income minus
housing cost is not less than standard non-housing cost; or 2) housing cost does not exceed
certain portion of income; or 3) housing cost does not exceed a certain level.
To handle the difficulties in measuring housing factors, some scholars had suggested
other framework. For example, Cai & Lu (2015) used the term ‘housing appropriateness’,
with consideration of affordability, accessibility, amenity and adequacy. Housing affordability
89
counts the household income and housing cost. Housing accessibility refers to time and
money spending on transportation to essential facilities such as workplace and school.
Housing amenity measures the housing condition with availability of toilets, kitchen and
running water. Housing adequacy evaluate the level of over-crowding and living instability.
On the other hand, there are still many considerations need to consider. Besides the
housing quality issue, Hancock (1993) added other factors such as the unit of measurement
(whether using individual, household, or tax unit), size of household (whether weighting is
needed) and meaning of ‘socially desirable’ and ‘maximum standard of housing cost. The
measurement household income and housing cost are also difficult to decide. For household
income, it is questioned whether permanent or transitory income was used, or whether
welfare income or housing benefit were included. For housing cost, beside rent and mortgage,
there are other expense like electricity fee, water payment or maintaining fee. Including all
the expense may lead to more complicated measurement (Linneman & Megbolugbe, 1992).
Moreover, it is more complicated for the cases of homeowners. The mortgage payment
will change overtime with interest rate, and there can be secondary or more level of
mortgage. It was debatable that the housing cost for homeowners can be viewed as wealth by
capital accumulation. Gan & Hill (2009) distinguished purchase and repayment affordability
for homebuyers, which the former reflects the ability for borrowing money to buy home and
the latter denotes the burden of households for repaying mortgage. Moreover, there are
90
different concepts of home price, transaction price, estimated values and standardized price
(Linneman & Megbolugbe, 1992).
g) Cause of housing affordability problems
To discuss the cause of housing unaffordability, it should first back to the composition of
the concept. Housing affordability involves several key factors, including income, housing
quality, housing cost and other factors such household size and family characteristics. Hence,
those social situations relating to these factors can be the causes of affordability problem, for
example, the rising of housing price and rent, housing supply, housing welfare and planning,
criteria of living condition, the stability of employment and income inequality situation (Li,
2014). Since there is theoretical discussion of housing and poverty separately in the previous
sessions, the later part will focus on studying the relationship among housing and poverty
2.3.3 Relationship between housing and non-income poverty
a) Housing and poverty in general
Literatures and reports showed that housing and poverty are highly corelated and
interacting (Crisp et al., 2016; JRF, 2015; Madden & Marcuse, 2016; Saunders, 2017;
Stephens & Leishman, 2017; Tunstall et al., 2013; Wallace, 2016). Generally, there are two
directions of study, one is ‘the impact of housing on poverty’, the other is the ‘impact of
poverty on housing’. Tunstall et al. (2013) reviewed the evidence for both ways and found the
later one had much support, for example, families with less income generally have less
91
desirable housing circumstance. However, the ‘impact of housing on poverty’ showed much
complicated relationship and revealed more questions and space for study.
Some housing factors have direct effect on poverty situation of residents. For example,
the available of affordable housing provision can lower the market rent, prevent homeless,
provide more stable livelihood for low income families. Better housing location facilitates
employment of low income workers by providing better transportation, more job
opportunities and incentive to work, building houses also can create jobs for workers (Crisp
et al., 2016; Tunstall et al., 2013). There are other housing factors with relatively indirect
effect on poverty. For example, bad housing condition have negatively effect on the child
development and health of adults (Tunstall et al., 2013). The housing situation such as
housing price also affect the prospect and future planning of youth (Terry, 2011).
The impact of some housing factors on poverty is complicated, such as homeownership.
Generally, home owners without mortgage can save housing cost, benefited from rise of
housing price and have risk of poverty (JRF, 2015). However, study found that homeowners
with mortgage payment may suffered in financial crisis and the housing wealth may help the
household get out of poverty mainly for the older outright owner but not the lower income
group, who may face more risks of financial distress (Wallace, 2016). The effect of housing
benefit is also uncertain. The housing benefit provided by the government can directly reduce
the housing cost of residents and more resource is left for non-housing consumption (Crisp et
92
al., 2016), however, some studies supported that housing benefit may create poverty trap
reduce the incentive of employment (Tunstall et al., 2013). Besides, the macro housing
factors also have significant impact on the current housing and poverty situation. They
include social planning of housing, housing monetary policy, housing law on housing
standard and regulation of housing market (Crisp et al., 2016). It is important to study the
forces behind the housing system and analysed the factors affecting the housing policies and
hence the current poverty situation. The framework below (
Table 2-3), taken reference from the past studies (Crisp et al., 2016; Kemeny, 2001, 2006;
Stephens & Leishman, 2017; Stephens & vanSteen, 2011), attempt to summarize the
relationship among the forces behinds system, housing system, housing situation and multi-
dimensional poverty. Four dimensions of housing factors and poverty situation are interactive
with each other:
Table 2-3 relationship among housing factors and poverty situation
Forces behinds Housing Housing situation / Poverty situation

housing system system/policy circumstance
/policy
- state ideology - supply and demand - affordability - Income
(housing regime) of housing - living condition / - deprivation
- market forces - tax and regulation quality - social exclusion
(globalization, real - urban and housing - surrounding / - social capital
estate) planning location: - capabilities
- roles of - housing policy neighbourhood, - subjective well
organizations - housing benefit community, social being
- beliefs of service
individuals
93
b) Housing and deprivation
The relationship between housing and poverty is complicated. To investigate the impact
of housing on non-income dimensions of poverty, this part reviews the previous studies about
the relationship between housing and deprivation, social exclusion, and subjective poverty.
Generally, various housing factors are associated with and lead to deprivation. First,
living space and housing condition are crucial factors causing deprivation. Physically,
housing is a place provides household with safety and privacy. The size and facilities of
house directly affect the deprivation level of households. The physical structure of housing
such as doors, wall, roofs, and windows, protects households from disease and disaster.
Moreover, the structure also enhances the sense of belonging and social support to all
household members. Housing serves as a crucial space for people to build interpersonal
relationship (King, 2015). Poor housing conditions such as inadequate fixtures, lack of living
space and bad building structure are found as crucial factors to affect material deprivation
(Napiorkowska-Baryla & Witkowska-Dabrowska, 2018). Research using EU statistics on
income and living conditions showed a high association between housing condition and
material deprivation. Poor housing conditions such as leaking roofs, damp wall, lack of
lighting and available of independent toilet are associated with deprivation (Guio & Maquet,
2007).
Second, the housing cost also plays crucial role on deprivation. Housing cost is usually
94
correlated with housing quality since the lower housing cost may imply poorer quality in
private housing market. Besides the physical structure of housing, there are other dimensions
of housing need to take into account, including housing cost and housing assistance (Shuey,
Leventhal, & Coley, 2016). Research in UK showed that decent quality and affordable
housing in areas with active job market can help families to increase income and get rid of
deprivation (Tunstall et al., 2013). Daniel, Baker, and Lester (2018) used cohort data in
Australia to examine the relationship between housing affordability and material deprivation.
They found there were similar population experience only one of the problems while smaller
number faced both problems at the same time. Further analysis showed that material
deprivation was significantly associated with housing affordability problem. Housing
affordability stress was one key precondition of material deprivation. They suggested it was
important to consider both housing and deprivation together in policy design (Daniel et al.,
2018). In US, a research, studied 3,311 low-income families in urban cities, supports that
housing affordability is associated with material hardship. The increase of individual rent and
decrease of affordable housing supply in a city are associated with increase of odds of
material hardship. Particularly, it was found change in rent is associated with 25% increase in
odds of material hardship (Warren, 2018). In East Asian context, there is a widespread
commodification of housing and turns housing as a platform for wealth accumulation. The
long-term inflation in housing price and rent causes critical impact on families (Izuhara &
95
Forrest, 2013).
Third, housing location is also found associated with deprivation. A research in UK
found that the level of deprivation was higher in the areas of inner city and those clusters far
away from city centre (Markkanen & Harrison, 2013). Housing location is also associating
with other housing factors such as housing cost. For example, In Hong Kong, the distance
from central business district (CBD) was found negatively associated with housing price. It
means houses closer to CBD generally with higher price (Hui et al., 2007). However, it is
unclear whether the distance from CBD is associate with deprivation.
Fourth, in terms of housing tenure, studies in Europe reveals that extensiveness of rental
housing market is negative associated with housing deprivation (Borg, 2015). Research in
Canada reveals that the food insecurity situation, as a measure of material deprivation, was
the worst among private rental residents, followed by homeowners with mortgage. The
outright homeowners were with less food security problems (Fafard St-Germain & Tarasuk,
2020). Housing assistant program generally showed positive effect in tackling deprivation
(Evans, 1998)
There is research in examining whether housing shows mediate effect among income
and deprivation. A study using structural equation model to investigate relationship between
income, housing, infrastructure, education, employment, and health. In the research,
infrastructure domain refers to outdoor problems such as noise, pollution and hygiene, and
96
accessibility to social service. Housing domain refers to indoor problems such as lack of
facilities and poor housing conditions. Result showed that income exert significant impact on
infrastructure deprivation and health problems through mediation of housing factors.
However, the impact of housing domain just reveals weak negative impact on infrastructure
and health (Rodero-Cosano, Garcia-Alonso, & Salinas-Perez, 2014). Another research in the
UK showed that the production of housing mediates the clustering of deprived groups. The
research revealed that housing production and housing assistance affect the concentration of
poor population. The housing setting further influence the deprivation level of residents
(Smith, 2012). It was argued that the housing situation was determined by income and
employment situation and the consumption of housing cannot mediate the poverty experience
and deprivation. But this cannot explain the substantial housing impact on other life aspects
such as health, education and employment opportunities (Lee, 1994). This reveals the
relationship between housing and poverty was unclear. The impact of income on housing
situation was well-known, however, the housing impact on poverty situation has not been
sufficiently studied.
c) Housing and social exclusion
In term of social exclusion, various housing factors reveal different level of impact on it.
First, housing location is usually found associate with social exclusions. The location of
housing influences the access to workplace, social service, and school. A convenient
97
transportation can enhance people to build social relationship and to get social support
(Malpass & Murie, 1999). A research in UK showed that limited access or concentration of
poor households in certain location lead households to be more socially excluded. This
related to the allocation of social housing in UK, for example, the decision of social landlord
was also a crucial factor in determining the social exclusion level of residents (Munch, 2012).
Research in Latin America also revealed that living location is one of the crucial factors
causing social exclusion. For example, those households lived in rural or peripheral areas felt
more socially excluded (García-Vélez et al., 2020). However, research in Australia showed
that housing location was not directly associated with social exclusions. There is no evidence
for explaining economic segregation by geographical variation (MacDonald, Galster, &
Dufty-Jones, 2018).
Second, the neighbourhoods and community facilities. The research of Park (2019)
suggested that the accessibility of social resource is differ with the same housing tenure and
the accessibility to neighbourhood resource is significant associated with social exclusion.
However, the relationship is complicated. Generally, the resource accessibility is positive
associated with social integration. But research result showed that area with medium and high
level of socio-health service accessibility has the lowest level of social integration in
community. Another research supported that a friendly community with adequate and
affordable facilities and social service enhance social inclusion of elderly residents (Sun,
98
Phillips, & Wong, 2018). Nevertheless, there is also research showing neighbourhood factors
made no significant effect on social exclusion (Friedrichs, 1997).
Third, housing tenure is usually found associated with social exclusions, but the
relationship is ambivalent. A research shows that those lived in social housing were more at
risk of social exclusion because that housing tenure concentrate poor households and the
residents were more excluded from the society (Marsh & Mullins, 1998). However, it was
argued that public housing concentrated households with low income and socially
disadvantage but public housing was a crucial measures in reducing social exclusion of
tenants (Arthurson, 2004). On the other hand, in Australia, renters are generally found more
suffered from different dimensions of social exclusion than owners. However, the outright
homeowners is found as the most excluded from neighbourhood dimension (Pawson &
Herath, 2017).
Fourth, there are also other housing factors affecting social exclusion, including living
space, housing tenure, housing insecurity, household composition and housing system. For
example, scholars supported that residential space was a crucial factor to determine social
exclusion (Somerville, 1998). Another research supported that housing insecurity led to social
exclusions. Households faced housing insecurity when they were unable to secure affordable
and adequate housing or faced eviction (Soyinka & Siu, 2018). Research in Australia showed
that households contain elderly or children are more likely to suffer from neighbourhood
99
exclusion than those households contain working adults only. On the other hand, households
only contain working adults are more suffered from low sense of belonging in community
(Pawson & Herath, 2017). In a macro perspective, scholars proposed that housing system is
one of the key element influencing spatial stratification and social exclusion (Marsh &
Mullins, 1998). Also, housing policy had been widely used a key tool in combating social
exclusion in European countries. Nevertheless, role of housing system in contributing social
exclusion has not been sufficiently studied (Munch, 2012).
Overall, it is not easy to divide housing as different independent factors, numbers of
research studies the impact of housing with consideration of different housing factors
together. Somerville (1998) argued that social exclusion was associated with housing tenure,
housing production, residential segregation and mobility. He proposed housing as a set of
social relation such as interpersonal network and social activities. Individuals suffered from
social exclusion through housing in different ways. For example, shortage of housing supply,
high level of housing cost and poor housing conditions increase the risk of social exclusion.
He argued that the association between housing tenure was ambivalent. Those not owning
houses may not suffer from social exclusion if they can find good quality houses with
affordable rent.
On the other hand, Arthurson and Jacob (2003) found several key relationships between
social exclusion and housing situation in Australia. For example, a poor housing condition is
100
linked with limited access to social services, a poor tenure security negatively affects
education and employment outcomes and a high housing cost discourages the social
participation of residents. In the US, Galster (2002) examined how the residential context can
restrict the availability of economic opportunities and perpetuate poverty. Dwelling unit
conditions, tenure and tenure stability all influence the personal characteristics and local
social networks of individuals (Friedrichs, Galster, & Musterd, 2003). In UK, Taylor (1998)
suggested that housing and social exclusion may form a cycle of labelling. For example, the
public housing concentrates households with low income and the residents may further lose
their confidence and grow sense of failure. The outsiders also exert bad image on those living
in public housing. This further worsens the social exclusion situation. However, he also
proposed alternative for reversing exclusion. Implementation of local service and reformation
of political cultural may help residents to build confident and new relationship with others
and hence get rid of poor image and social exclusion (Taylor, 1998).
In Hong Kong, some studies have examined the impact of housing type or location on
poverty. For instance, Wong (2011) found that those residents living in the remote areas of
Hong Kong had higher levels of job insecurity and transportation cost, which are both
correlated with a higher risk of social exclusion. Meanwhile, Hui et al. (2007) found that
housing cost, which brings much financial burden to residents, is associated with the travel
time to the Central Business District. Forrest et al. (Forrest, LaGrange, & Yip, 2004) added
101
that the relationship between the concentration of deprivation and spatial location was not
statistically significant. By using census data for 1991 and 2001, Delang and Ho (2010) found
that public housing does not promote poverty concentration at the district level, thereby
highlighting the success of the city planning and public housing policy of Hong Kong in
enhancing social heterogeneity. However, by conducting a finer-level spatial analysis, Guo et
al. (2018) found that public housing concentration has a significantly positive association
with poverty. Previous studies also reveal that a high housing density negatively affects the
levels of deprivation, stress and social relationships (Mitchell, 1971) as well as the quality of
living spaces (Chan, Tang, & Wong, 2002).
d) Housing and subjective poverty
Houses are the place for building family relationship, friendships, and community
linkage. Better housing situation generally enhance personal freedom as well as happiness
(Clapham, 2010). For studying the relationship between housing and subjective well-being,
scholars suggested several objective features of housing concerns such as housing type, living
density and living environment (Clapham, Foye, & Christian, 2018; Herbers & Mulder,
2017). Other housing factors such as living space and housing location are commonly found
associated with subjective poverty or subjective well-being in previous studies.
First, living space and housing environment. Foye (2017) proposed there are two
pathways for impact of living space to subjective well-being. One is the increase of space can
102
promote activities and freedom. The other one is status which determined by relative position
with relevant others. Using UK household panel survey data, the result shows that the size of
living space is weakly positively association with subjective well-being and the result is only
applied for men but not women. Moreover, the magnitude of effect is decreased over time
due to adaptation. Nevertheless, it suggests further research is needed to explore the effect of
housing factors of subjective well-being and housing satisfaction separately (Foye, 2017).
Another research in Korea supports that larger living area contributes positive effect to
subjective well-being. Result reveals happiness of households progressively increase with
living area up to a certain limit and the impact of living space diminish after that threshold
(Rudolf & Potter, 2015). A research in China also shows similar result. Living space, which
also reflect the wealth for homeowners, was found particularly important for happiness
among housing factors (Zhang et al., 2018). Physical living environment, including housing
equipment and facilities, is also found significantly correlated with subjective assessment of
living satisfaction (Amérigo & Aragones, 1997).
Second, housing type or housing tenure. Subjective well-being was found to be
associated with home ownership, which not only provide absolute benefit to the owners but
also relative benefit such as individual status. Research reveals home ownership is a kind of
positional good and social norm. The subjective well-being of a person is affected by
comparison with housing status of relevant others and overall home-ownership rate (Foye,
103
Clapham, & Gabrieli, 2018). Research in China also supported that owning house was
positively associated with subjective well-being but varied across different housing types
such as commodity housing and affordable housing (Wu et al., 2019). Renters are found more
at risks in residential insecurity and uncertainty and further causes negative impact on their
subjective well-being. However, several research also argued that the effect of housing type
causes no significant impact on subjective quality of life (Patterson et al., 2013). For
example. a research, using SEM modelling, tested the relationship between housing tenure,
crowding of house, type of dwelling, home environmental quality, and subjective well-being.
Result demonstrated that housing ownership was not associated with housing related
subjective well-being. On the other hand, subjective well-being was positively associated
with living area per capita while negative associated with living in an apartment (Caffaro et
al., 2019). The relationship between housing type and subjective poverty is unclear.
Third, housing location and community factors. A recent research in Hong Kong
studying quality of life demonstrates that community and neighbourhood attributes are
important domain of quality of life and cause significant impact on residents’ subjective
assessment of well-being. Result also reveals attributes such as availability of social service,
social support from other people, safety and convenience to community facilities were
significantly concerned by the residents as important factors in neighbourhood domain (Low
et al., 2018). On the other hand, a research in Malaysia examined impact of social service and
104
community factors on subjective residential satisfaction. Result shows that subjective
satisfaction is highly associated with social service, public facilities, dwelling features, and
community environment. The study proposed objective features of residential factors,
including social and community dimensions of housing, cause significant impact on the
subjective assessment of well-being of residents (Mohit, Ibrahim, & Rashid, 2010).
Fourth, the impact of housing cost on subjective poverty is complicated. Generally,
subjective well-being is positively associated with less expensive and more affordable
housing (Florida, Mellander, & Rentfrow, 2013). However, scholars argue that housing cost is
sometimes correlated with wealthier and higher quality of life and lead to better subjective
well-being (Rentfrow, Mellander, & Florida, 2009). On the other hand, research also
supported that those living in economically affordable housing had a lower housing
satisfaction than others living in commercial housing. This further caused negative impact of
their subjective well-being (Zhang et al., 2018). Another research in US shows that the
housing prices are not significantly associated with life satisfaction, however, the housing
cost to income ratio is negatively associated with subjective well-being for those spending
more than 35% of their income on housing cost (Lawless & Lucas, 2011).
Overall, similar with the previous reviewed literature, number of research study the
combined effect of housing factors on poverty instead of independent factors. Housing
features including house size, number of bedroom or living rooms, housing types was found
105
positively associated with housing satisfaction, which further cause positive impact
subjective life satisfaction. Moreover, the effect of housing factors on subjective satisfaction
was found more significant for the low income families than the high-income group (Zhang
et al., 2018). Using statistics from EU, Sunega and Lux (2016) argued that the objective
measures of housing problems such as overcrowding and unaffordable housing show obvious
distance from the subjective evaluation of the problems by households. For example, those
household with housing cost burden below 40% may perceive the housing cost as heavy, or
those living in overcrowding houses by objective measurement may not perceive there is
shortage of space. Research in Europe found that the magnitude of housing factors impact on
subjective well-being is various across countries with different level of housing markets
regulation and different general housing quality. For example, in countries with more
regulation in housing market, the negative effect of renting on subjective well-being is
smaller. In countries with generally higher level of housing quality, the impact of living area
is weaker (Herbers & Mulder, 2017). A study in China also reveals housing factors such as
housing quality, home ownership and community facilities all cause crucial impact on living
satisfaction, which reflected by subjective assessments of their living environment by
households. However, the magnitude of impact by housing factors on satisfaction was
different across living locations (Ren, Yuan, & Hu, 2019). On the other hand, housing
intervention evaluation reports that improvement of housing safety and living situation
106
significantly enhances the subjective well-being of homeless people (Patterson et al., 2013).
Some research investigates the relationship between income, housing, and subjective
well-being in one study. A research shows that although there is considerable overlapping of
monetary poverty and subjective poverty, some household characteristics distinguish them.
Subjective poverty is affected by a range of factors other than money. For example, housing
ownership, building structure, indoor housing facilities are found associated with subjective
poverty in addition to income (Posel & Rogan, 2016). Another survey in Norway examined
the relationship between housing problems, income poverty and subjective poverty. It found
that the overlap between these dimensions is not large. This implies they measure different
dimensions of poverty situation. Result shows the low-income group faced higher risk in
housing deprivation and subjective deprivation than the general population (Sandbæk, 2013).
On the other hand, some research uses housing factors as measures of subjective poverty
or subjective well-being. For example, Mitra (2016) used adequacy of housing as one of the
key dimension in measuring subjective well-being in Nepal. The housing factors used
includes overcrowding, sanitation, indoor environment, and home facilities. Other housing
variables such as small living area, leaking roof, damp wall, rot in window frames, are also
commonly used as multidimensional poverty indicators, together with income poverty
(Moisio, 2004). Some scholars suggested that subjective poverty is also a multidimensional
concept, which consists of elements in different life domains including housing and
107
subjective assessment of housing (vanPraag & Ferrer-i-carbonell, 2006). Besides, adequacy
of housing and housing consumption are also used as key elements in assessment of
subjective poverty (Pradhan & Ravallion, 2000).
In Hong Kong, there is not much research focus on subjective poverty, but related
measures such as quality of life and subjective well-being. A recent research reports that
housing environment was an influential factor in affecting quality of life. The effect is
particularly significant for those lived in public housing. The housing environment factors
used in the survey included housing location, size, privacy, ventilation, and noise. These
factors were found as significant environmental factors of quality of life. Result also showed
that people living in private housings were much satisfied with location while public rental
housing residents were most dissatisfied about privacy, size and noise (Gou et al., 2018).
Another recent research in Hong Kong studies the quality of urban life and highlighted the
importance of living domain of housing and neighbourhood in subjective assessment of
quality of life. The housing domain included housing affordability, adequacy of space, indoor
and outdoor environment while the neighbourhood domain included accessibility of
community facilities, local social service, safety, building density and neighbourhood
relationship. Both domains were used to reflect quality of urban life and showed association
with quality of life (Low et al., 2018).
108
2.4 Conclusion: Theoretical gaps and significance of research
There are several important points to conclude in this part: 1) poverty is
multidimensional in nature. There are different approaches in studying poverty, including
monetary, deprivation, social exclusion, capabilities, social capital and subjective. It cannot
be simply reduced into one approach; 2) Housing in nature is related to the well-being of
people. It is core element in satisfying basic need and building social relationship and capital
of people; 3) the existing housing situation is highly influenced by the political economy of
that place, such as globalization, housing market regulation, ideology of government and
control of supply and demand; 4) The housing factors had critical impact on poverty situation
of people, but this was relatively underdiscussed in the literature of poverty, either
conceptualization or explanation; 5) the past housing poverty studies mainly focused on
monetary approach but relatively neglected other dimensions of poverty.
The theoretical review of poverty studies reveals the neglect of housing in
conceptualization and explanation of poverty. Monetary approach, especially income
dimension, is dominating in theoretical discussion. Housing factors has appeared in certain
theories, such as deprivation, social exclusion, and capability, but not the focus or core elements.
There was discussion of housing poverty in literature, however, most of them are based on
monetary approach, focus on the discussion of housing affordability.
To sum up, the theoretical gaps in existing literature are 1) the neglect of housing perspective
109
in conceptualization of poverty, 2) the missing piece of the impact of housing factors in
explaining different non-income dimensions of poverty and 3) the process of how poverty
situation is interacting with housing factors.
110
Chapter 3 Empirical Review of Housing and Poverty in Hong Kong
Literature review of theories and concepts of poverty indicates that housing perspective is
neglected in previous poverty studies. In this chapter, an empirical review of literature and
surveys reveals that housing and poverty situation are correlated and interact with each other.
The empirical review focuses on Hong Kong’s situation based on the conceptual framework
(Section 3.1). Academic literature about housing and poverty in Hong Kong is briefly
reviewed (Section 3.2), followed by the poverty (Section 3.3), housing (Section 3.4) and
housing poverty (Section 3.5) profile of Hong Kong. To provide a background of the housing
poverty situation, the political economic background of Hong Kong is discussed (Section
3.6). The conclusion highlights the empirical gaps and significance of this research (Section
3.7).
111
3.1 Academic literature review
To review the literature about housing and poverty in Hong Kong, the word “housing”,
“poverty” and “Hong Kong” and relevant keywords are used for searching journal articles in
Scopus and google scholar. The review is limited in the recent 20 years since 1997. The result
reveals into two parts, poverty in Hong Kong and housing in Hong Kong.
Poverty in Hong Kong
The search about poverty research in Hong Kong used the word “poverty” AND “Hong
Kong” as keywords for searching in Scopus, from 1997 to 2017. It excluded those articles
from disciplines such as medical, nursing, and environmental science. The selected articles
were mainly from the field of social science.
From the result, most articles focus on researching the belief, behaviour or thought of
different groups of poverty population, for example, parental belief and behaviour of low
income families (Lam, 2011; Leung & Shek, 2013a, 2013b; Shek, 2005), psychological
situation of economic disadvantaged youth (Shek, 2003; Shek & Lin, 2014). The other group
of research focus on the overall poverty analysis for specific social group, more attention was
put on the group of elderly, children and youth, for example, the poverty, exclusion and
deprivation situation of elderly (Chan & Chou, 2017; Cheung & Chou, 2018; Chou, 2017) or
impact of poverty on children (Ho, Li, & Chan, 2015; Wong, Ma, & Chan, 2017). Other
groups such as disabled (Chau, Yu, & Boxall, 2017), ethnics minority (Cheung & Chou,
112
2017; Ku, 2006), women and working poor are included, but with limited research. It is not
easy to categorize these two groups of research, the former focus more on specific
characteristic of certain poverty population, the later one put attention on the overall analysis
of the poverty situation for particular social group.
There are researches focus on overall poverty situation in Hong Kong, most of them
using income or economic approach in analysis (Fong & Wong, 2015; Lau, Pantazis, Gordon,
Lai, & Sutton, 2015; Yip et al., 2016). The other approach put attention on the poverty
alleviation, for example, the critique for government orientation (Wong, 2000, 2015; Yu,
2008), the evaluation of anti-poverty projects (Ho & Chan, 2010), or the role of community
(Fung & Hung, 2014) and social assistance (Qi & Tang, 2015) for poverty reduction. Besides
the income and economic dimensions, there are research used the concepts of deprivation
(Saunders et al., 2014a), quality of life (Lam, Guo, Wong, Yu, & Fung, 2016; Tang, 1998;
Wong, 2005, 2011), social capital (Chou & Chow, 2009; Forrest & Kearns, 2001; Holliday &
Tam, 2001) and social exclusion (Lau et al., 2015; Lee, Wong, & Law, 2007; Saunders,
Wong, & Wong, 2014b) for analysing the multidimensional poverty situation in Hong Kong.
The research related to housing and spatial will be analysed in the later part.
Overall, the academic research for Hong Kong poverty are mainly focus on the income
and economic disadvantage of the population, with some focus on specific social group.
There are certain researches used different approaches for investigation, for example, social
113
capital and social exclusion, however, comparatively few. More attention needs to put into
other dimensions to broaden the understanding of poverty situation in the local context.
Housing in Hong Kong
The review on housing research in Hong Kong used similar strategy in searching
“housing” and “Hong Kong” with Scopus, from 1997 to 2017. Those articles from peer
reviewed journals, with clear relation with Hong Kong housing situation, are selected for
analysis Among those researches, numbers of them were study of housing market, price and
financing, for example, the detection of price bubble (Chan, Lee, & Woo, 2001; Yiu, Yu, &
Jin, 2013) the effect of housing price on fertility (Yi & Zhang, 2010), the real estate market
(Haila, 2000; Smart & Lee, 2003a). Beside this, public housing was one of the focuses of
housing research, the topics included privatization (LaGrange, 1998, 1999; Zheng, Zheng, &
Chen, 2017), perception (Forrest & Yip, 2014) and development of public housing (Forrest
&Yip, 2014; Lau & Murie, 2016; Yip & Lau, 2002).
On the other hand, there are studies concerning the topic of home ownership, such as
housing pathway (Campos et al., 2016), owner corporation (Yip &Forrest, 2002) and
ideology of ownership (Chan, 2000). Housing planning was also an important issue in
housing study, especially the impact of planning and land use (Chiu, 2007; Lai & Wang,
1999; Lau, 2017) and gentrification (LaGrange & Pretorius, 2016; Ley & Teo, 2014) on
housing situation. The other fields of housing studies also include housing policy and
114
governance (Li, 2016; Smart, 2001), housing quality (Busiol, 2016; Chan, So, Tang, & Wong,
2008; Chan, Tang, & Wong, 2002) , neighbourhood and community (Forrest, LaGrange, &
Yip, 2002; Yau, 2012).
It is worth noting the categorization mainly based on the themes or keywords provided
by the journal articles. There may be numbers of overlapping topics, for example, housing
poverty and public housing is highly corelated, however, only one theme is chosen from each
article.
Housing and poverty in Hong Kong
Those articles related to poverty and housing in Hong Kong have been selected from
more detailed analysis. Numbers of research focus was on public housing, for example,
redistributive or anti-poverty effect of public housing (Hu & Chou, 2015; Lui, 2007; Lui &
Suen, 2011) and poverty situation of public housing residents (Delang & Lung, 2010;
Monkkonen, 2011).
Other related topics include affordability of housing (Ho & Chiu, 2002; Hui, 2001; Yip &
Lau, 2002), the neighbourhood of residents living in poverty (Chen et al., 2016; La Grange,
2011; Lum, et al., 2016), access to social service in poor area (Guo, et al., 2017), housing
expectation and pathway of youth ( Campos, et al., 2016; Forrest & Xian, 2018) and right to
housing (Yung & Lee, 2014).
Short conclusion of finding
115
From the analysis of academic literature in the past 20 years, it is found that substantial
research in the field of poverty used economic approach for investigation and mainly focus
on specific social group such as elderly and children. The housing researches are more
diversified, topics include market, finance, planning, ideology, perception, and policy. For
research related to housing and poverty, more attention had put into the public housing,
including its anti-poverty effect and the circumstance of residents.
However, there are only limited researches in housing poverty. The poverty situation of
private housing resident is neglected in academic literature. The issues of affordability, right
to housing, living quality and residential satisfaction, although some scholars attempted to
study, still lacking comprehensive research. Moreover, the concept of housing poverty and
the relationship between poverty and housing is still short of in-depth study in the past and
worthwhile for more in-depth research.
116
3.2 Poverty Situation in Hong Kong
The understanding of poverty in Hong Kong is limited, dominated by income approach,
but neglected other dimensions such as inequality (Atkinson, 1987; Therborn, 2013),
capabilities (Nussbaum, 2011; Sen, 1985, 1999), deprivation (Townsend, 1979c, 1987), social
exclusions (Levitas, 2006b; Levitas et al., 2007; Munck, 2005) and social capital (Farr, 2004;
Field, 2008; Putnam, 1995; Woolcock, 2002). This part aims at reviewing the poverty
statistics, report, and related information with different perspectives in Hong Kong, within
recent 10 years
Income poverty
The government set up the poverty line, as 50% of median monthly household income of
different household size, in 2013. The Commission on Poverty (CoP) released the poverty
situation figure each year afterward. In 2018, the poverty line was set at $4,000, $10,000,
$16,500, 21,000 for 1, 2, 3, 4-person family respectively(Figure 3-1) (HKSAR government,
2019). The overall number of people living under these poverty lines were 1.41 million (0.61
million household), with poverty rate 20.4%. After policy intervention, including recurrent
cash subsidies, the poor population dropped to 1.02 million, with poverty rate 14.9% (Figure
3-2).
117
Figure 3-1 Poverty lines by household size, 2009-2018
Figure 3-2 Poor population and poverty rate after recurrent cash intervention
There was more detailed analysis of poverty for the selected socio-economic group.
Within the poor population, elderly and Comprehensive Social Security Assistance (CSSA)
recipients were the poorest social groups. Even after recurrent cash intervention, the poverty
rate were 45.9% and 48.9% respectively (Figure 3-3) (HKSAR government, 2019)
118
Figure 3-3 Poverty rate and poor population by selected socio-economic group, 2018
On the expenditure side, in 2015, the poor populations spent most of their money on
housing and food with 39.7% and 31.4%, compared with 35.6% and 27.7% of overall
households in Hong Kong. Housing was the largest expenditure item among households of
different income level. For the 1-person poor household, housing spending was $5100 of
$9400, near 54% of total expenditure (Figure 3-4) (HKSAR, 2016).
119
Figure 3-4 Expenditure patterns of poor and all households by household size, 2015
Income Inequality / social mobility
Besides relative poverty measured by the official poverty line, income inequality also
getting more serious in the past decades in Hong Kong. The Gini coefficients, which ranged
from 0 to 1 and used to measure level of inequality, increase from 0.533 in 2006 to 0.539 in
2016. After count the tax and social transfer, the Gini coefficient remain high as 0.473 (CSD,
2017a), compared with other countries worldwide.
In addition, although there was general increase of monthly income in different
occupations, by income decile group analysis, the income distribution was highly uneven.
The 10th (highest) decile group shared 39.6% monthly income of all working population,
more than 26 times of the sharing of the 1st (lowest) decile group, which just shared 1.5% of
total (CSD, 2017a). Oxfam used data from CSD to show that the income of the 10th decile
120
group rose from $82,700 in 2011 to $100,000 in 2015, on the contrary, those of 1st decile
group just increase from $3,100 to $3,500 in same period (Oxfam, 2016).
On the other hand, recent studies for social mobilities showed that the youngers, even
with high education level, were much difficult to mobilize upward compared with those in
past years. For example, the starting salaries of graduates with different education levels in
2006/07 cohort were higher than those in 2011/12 cohort, and there were also around 20% of
graduates showed “no mobility” or “downward mobility” in 2007/08 to 2013/14, among
different education level (FS Office, 2016).
In another report conducted by Legislative Council, it showed the social mobility
weakened in recent years and concluded there was “limited opportunities for people moving
higher on the social ladder” (LegCo, 2015, p.9). There was expansion of post-secondary
education places, however, many degrees were self-finance and sub-degree and higher
educational level did not lead to better employment. Research result also showed there was
significant intergenerational factor social mobility, for example, children with father with a
degree or was professional, had high probability to have them too (LegCo, 2015; Vere, 2010).
On the other hand, for the children living in poor families, they were much harder to mobilize
upwards.
Deprivation
There were just few studies focus on analysis of social deprivation in Hong Kong in
121
recent years. In 2011, Hong Kong Council of Social Service (HKCSS) and scholars
conducted a research for social deprivation, interviewed 1037 Hong Kong residents aged 18
or above. In the study, there were first 37 basic living condition items which constructed by
scholars, focus group and from literature, then 35 items of 37 were agreed by more than half
of informants and marked as “items for basic living condition”. Social deprivation was
defined as lacking 4 items or above out of those 35 basic living conditions, such as “having
breakfast everyday”, “able to have dental check-up periodically” or “can have hot shower in
cold winter”. Result found that there was 18.4% of people living in deprivation in 2011
(HKCSS, 2011; Saunders et al., 2014a).
In, the same year, HKCSS also interviewed 754 cases of CSSA recipients, elderly or
families with disabled person. With the same definition of deprivation mentioned above, in
this research, the rate of deprivation (lacking 4 or more basic living items) of CSSA
recipients, elderly or families with disabled person was 56.6%, 33.1% and 20% respectively.
For those CSSA recipients living in private rental housing, the deprivation was even higher as
75.1%, compared with 52.5% for those living in public housing. Research concluded there
was insufficient supports, such as financial subsidies, medication and social network, for
these three groups of people (HKCSS, 2012). In 2014, HKCSS set up another list of
deprivation with 14 items covered areas of family facilities, food, clothes, medication, and
social life. The informants lacking 2 or more items were categorized as under social
122
deprivation. With this definition, it was estimated there was 14.5% of Hong Kong people
living under deprivation. People who were unemployed (26.2%), living in private rental
housing (19.4%), single elderly (37.1%) had a higher chance to be deprived (HKCSS, 2014).
Social Exclusion
The concept of social exclusion was not commonly used in Hong Kong, nevertheless, in
the past years, there were several researches of social exclusion or used the perspective. Lau
et al. (2015) reviewed the researches of poverty and social exclusion in the past 30 years and
suggested a new framework, including quality of life, resources and participation, for study.
In the focus groups of the newly framed research, it was found that the participant felt
difficult to understand the concept of social exclusion, yet, eight key issues were identified by
the interviewed associated with social exclusion, such as “social support”, “labour market
participation” and “discrimination” (Lau, Gordon, et al., 2015).
One comprehensive quantitative research was done by HKCSS mentioned above, 16
items in the survey was chosen with agreement of participants as indicators of social
exclusion. Examples of indicators included “To be treated with respect by other people”,
“Can take transport for visiting relatives and friends” and “Has a mobile phone”. There was
24.1% of people experienced 4 or more items of exclusions and 6.6% experienced 8 or more.
For social group analysis, those aged 65 or above (3.85), with members with chronic disease
or disability (4.18) and unemployed (3.23) were with highest social exclusion index among
123
other groups (Saunders et al., 2014b). On the other hand, research showed that the social
exclusion perceived in society was critically affect morale of people in Hong Kong. The poor,
elderly, with lower education or unemployed were much suffered from social exclusion, with
effect of high life stress (Cheung, 2013).
Social Capital
Holliday and Tam (2001) attempted measuring the overall social capital
comprehensively in Hong Kong society, using data from 1970s to 1990s. They
operationalized social capital into 5 parts, including number of membership in voluntary
associations, participation in voluntary work, donations, informal sociability, and social trust
(Holliday & Tam, 2001). Result found general increase of social capital in the past decades,
for example, rising number of member in trade union and youth organization, the non-taxable
charitable organizations increased from 1163 in 1982 to 3226 in 2000, growing donations
raised by charity, percentage of people believing citizens could be trusted moved from 18%
in 1993 to 40% in 1998. However, the sociality in neighbourhood and community level was
limited, in 1990s, for example, watching TV the most popular leisure activity, but few
respondents viewed visiting people as leisure (Holliday &Tam, 2001).
Researches in Hong Kong supported human capital and social capital were crucial for
reduction of poverty and inequality (Lau, 2010; Lee et al., 2007; Mok, 2015). Chou & Chow
(2009) interviewed 449 respondents from random sample of new arrivals from Mainland
124
China. Social capital was measured by indicators including size of social network, network
quality, support from friends and family, and collective efficacy. Result showed that the social
capital of the employed respondents was much higher than those of not employed, especially
in size of social network. It highlighted the strength of weak ties and the role of human and
social capital was crucial for economic integration of new immigrants. Another study by
Chou (2013), by analysis of census data, showed the poverty rate of children was higher in
immigrant families than local families. For example, the poverty rate of “First-generation
children in immigrant single-parent families” was 57.9%, much higher than 9.3% “Children
in two-parent local families”. Analysis demonstrated child poverty was highly associated with
social capital in families, parental characteristics, and family structure.
On the other hand, there were two main projects in Hong Kong using the concept of
social capital launched by the government, named Community Investment and Inclusion
Fund (CIIF) and Child Development Fund (CDF). For CIIF, the fund was set up in 2002,
$432 million was funded for 323 projects with building of more than 2,000 mutual help
network and 650,000 participants (HKSAR, 2017a). The project was evaluated in recent year.
Result found CIIF projects were effective in building bonding social capital most
significantly, compared with bridging and linking social capital. They also showed major
impact on psychological gain (mean score = 75.2), socially inclusive experience (mean score
= 73.0) and rejoicing experience (mean score = 72.9) to project participants (Chan, Cheung,
125
Ho, & Tija, 2012). Another evaluation report focus on effectiveness of projects in Tin Shui
Wai measured the social capital stock of residents, including “civic engagement”, “trust and
enhancement of the neighbourhoods relationship”, “norms of reciprocity”, “acceptance and
understanding” and “sense of belonging and sense of influence” , and showed CIIF helped in
enhancing all types of social capital (Ting, 2012). Another research also showed CIIF and
community development projects, in deprived community (Tin Shui Wai North), were
effective in building social capital, especially for bonding social capital. However, it needs
more sensitivities and effort in development bridging and linking social capital in community
projects (Fung & Hung, 2014).
Capabilities
In Hong Kong, there was no research specifically investigate poverty using the exact
concept of “capacities”. However, there were several surveys can be related to this, for
example, about social development and quality of life.
The HKCSS launched the project of Hong Kong Social Development Index (SDI) since
1999, aimed at keeping track of the progress of social development and assessing the social
and economic needs of society. The Index consisted of 1) a weighted Social Development
Index at the societal level, 2) 14 Sub-indexes for different domains of development, and 3) 5
Sub-indexes for specific social groups. (HKCSS, 2017). The SDI of 2016 was increased to
205, based on the data on the year of 2014, compared with 196 in 2012, 191 in 2010, 173 in
126
2008 and 171 in 2006. Despite the sub-indexes in most development areas, including health,
education, environmental quality, and public safety, remained positive in 2014, the indexes in
the domain of “housing” and “family solidarity” sharply dropped and scored at -238 and -148
respectively. For social group, the index of children development remained negative in the
past years, mainly due to the poor child poverty situation (HKCSS, 2017).
For the study of quality of life, Wong (2005) measured quality of life as levels of
poverty, income security and expenditure for the poor household in 1990s. It found that there
was significant increase of expenditure of CSSA recipients, the high housing cost caused the
poor families to squeeze the spending on food and daily expense and concluded the quality of
life of poor household were deteriorating in 1990s in sense of living standard and social
exclusion. Another research done by Wong (2011), measured the quality of life of elderly,
women and youth living in remote areas in New Territories, showed the quality of life were
impacted by lacking neighbourhoods in local community, high transportation cost and limited
choice of jobs.
The coverage of “capabilities” can be broad and include many aspects of well-being,
such as health and emotion (Nussbaum, 2000). With this understanding, there were other
researches studying related areas, for example, health inequality (Chung & Wong, 2015),
health related quality of life (Lam et al., 2017), psychological well-being (Shek, 2003),
personal well-being and family quality of life (Shek & Lin, 2014), worthwhile for further
127
study
Short conclusion
The above researches and studies showed the poverty is a multidimensional phenomenon
and the poverty situation in Hong Kong was serious in the past years. Understanding is the
first step for poverty alleviation. It needs more active interventions and all-around anti-
poverty policies to tackle this long-lasting problem.
It should be reminded that the above information was mainly from academic research
and from government official data, however, not covered those surveys or researches
conducted by NGOs or social media, which may be more targeted or with different angles of
analysis. Further study is needed to have a more comprehensive understanding of poverty
situation in Hong Kong.
128
3.3 Housing situation in Hong Kong
Demography of overall population
According to the data of by-Census 2016, there are more than 7.1 million people living in
around 2.5 million of quarters in Hong Kong in 2016. Most of the residents distributed in 4
types of housing: 1) private housing with ownership (33%), 2) private rental housing (20%),
3) subsidized public ownership housing (15%) and 4) public rental housing (30%). Other
types of housing include temporary housing (1%) and non-domestic housing (1%) such as
elderly home (Table 3-1 and Figure 3-5).
Table 3-1 Occupied Quarters, Owner Occupiers and Population in Domestic Households by
Year and Type of Housing
Population in Domestic
Occupied Quarters Owner Occupiers
Households
Number of Unit of Quarters Number Number of Persons
Year Type of Housing
Public rental housing 691488 - 2096126
Subsidised home ownership
362910 351346 1204324
housing
2006 Private permanent housing 1143236 - 3278708
Non-domestic housing 10723 - 10034
Temporary housing 17717 - 47717
Total 2226074 1174132 6636909
379002 367339 1194296
housing
Private permanent housing 1246957 3560047
2011 - ownership 857984
- rental 388973
Non-domestic housing 14425 218 15350
Temporary housing 18580 8054 47388
Total 2381125 1233595 6891659
384849 372453 1144774
housing
Private permanent housing 1336873 3789474
2016 - ownership 835499
- rental 501374
Non-domestic housing 19234 1076 25831
Temporary housing 21186 7975 53931
Total 2526026 1217003 7114136
Note. Adapted from CSD (2017b).
129
Figure 3-5 No. of domestic household by housing type, 2016.
Note. Adapted from CSD (2017b).
Lau & Murie (2016) reviewed the trend of domestic household by tenure and found that
the population in public sector gradually increase from 31.1% in 1971 to 46.4% in 2011. This
mainly contributed by the supply of subsidized home ownership program since 1980s,
increase from 0.6% in 1981 to 15.9% in 2011. However, the portion of public rental housing
dropped from 36.5% in 1991 to 30.4 in 2011, the privatization of public housing in 2000s was
the main factor (Table 3-2).
For the private sector, there was increasing trend of home ownership, sharply rose from
18.1% in 1971 to 33.9% in 1991 and remain steadily in the later years. The percentage of
household living in a private rental housing went in opposite direction, dropped from 42.2%
in 1971 to 18.2% in 1991 and keep stable afterward. The factors behind the growing trend of
home ownership needs more in-depth investigation.
130
Table 3-2 Domestic households in Hong Kong by tenure (%)
Note. Adapted from Lau and Murie (2016).
Housing price and rent
Both housing price and rent increased sharply in recent years. According to the rating and
valuation department, the rent index of domestic private housing dropped from 134.5 at the
peak in 1997 to 73.6 in 2003 and kept rising to 194.5 in 2019, with a slight downward
adjustment from 2008 to 2009 (Figure 3-6).
Figure 3-6 Rental indices for Hong Kong property market
131
Note. Adapted from RVD (2019).
The price of private went with similar trend, but more vigorous. The price index of
domestic private housing dropped from 163.1 at the previous peak in 1997 to 61.6 in 2003
and kept rising to 382.8 in 2019. Except the period of Asian financial crisis started in 1997,
there was an obvious upward trend for the housing price. The housing price increased about 6
times from the lowest to the highest point. It worth noting that the price of other types of
properties, such of offices and retail units, moved up even more than domestic housing. The
price flatted factories increased near 10 times since 2003 (Figure 3-7).
Figure 3-7 Price indices for Hong Kong property market
Significant housing policy change (after 1997)
To study housing policy change in Hong Kong, Li (2016) has done a comprehensive
review of Hong Kong housing policy in recent decades, with analysis of the policy objective,
132
target, merit and demerit of each policy. The important changes included implementation of
public housing, home ownership scheme, sandwich class housing scheme, home starter loan
scheme, and after 1997, tenant purchase scheme, 85,000 plans, “spicy measures” and reverse
mortgage program.
Other than those mention by Li (2016), there were also other important policy change,
such as the nine measures to stabilize the property market introduced by Secretary of Housing
Mr. Suen (commonly known as "the Nine Measures of Michael Suen") in 2002, long-term
housing strategy (LTHS) proposed in 2014, land development plan for New Territory North
East (NTNE), cancellation of rental control in 1998 and tenancy right control in 2004,
introduction of youth hostels in 2010s, housing-related poverty relief measures, etc.
Moreover, the housing policy mentioned in Policy Address of each year and the housing
policy suggestions in platforms of Chief Executive are also worth study.
Housing supply
There was decreasing trend of supply in the past 15 to 20 years. For private housing, the
completion of housing was more than 30,000 units in 1999 and 2002, but it dropped to less
than 10,000 units in 2009. There was a slight increase after 2009, around 10,000 to 20,000
per year, however, far less than those in 10 years before. Generally, by living area, large
portion of the private housing was 40-69.9-meter square units, the second large portion of
supply was <40-meter square units (Figure 3-8). For public housing, the supply reached the
133
top around 30,000 units in the year of 2001-2002 but drop to less than 15,000 units in 2015-
16. On average, there was also a decreasing trend of supply, from 2001-2006, the average
supply per year was 21,553 units, but dropped to 13,980 units per year in the next 5 years,
further decreased to 12,512 units in 2011-2016 (Figure 3-9).
Figure 3-8 Private Domestics - completion by class
Figure 3-9 Completions of public rental housing flats
Note. Adapted from LegCo (2016)
134
3.4 Housing poverty situation in Hong Kong
Housing features of and housing transfer to income poverty households
In 2018, within the 1.41 million poor population, there was 707,200 people (50.2%)
living in PRH, 512,200 (36.6%) living with homeownerships and 148,300 (10.5%) living in
private rental housing. After intervention, within the 1,024,300 people in poverty, 38.3% of
them living in PRH, 48.2% are owner occupiers and 9.2% living in private rental housing
(Figure 3-10, Table 3-3 and Table 3-4).
Figure 3-10 Housing characteristics of poor households
Note. Adapted from HKSAR government (2019).
Table 3-3 Poverty household, population and rate by tenure
135
Table 3-4 Tenure distribution with by selected poor household Group, 2018.
Housing group PRH (%) Owner- Private Others (%)

occupy (%) rental (%)
Overall (1,024,300) 38.3 48.2 9.2 4.4
CSSA (149,500) 74.6 10.8 13.0 1.7
Elderly (all member aged 28.6 60.9 3.7 6.9
65 and above, 240,600)
Single parent (70,100) 66.2 13.8 18.9 1.1
New arrival (69,800) 42.2 12.3 43.7 1.8
With children (399,700) 50.6 27.1 19.7 2.6
Youth (all member aged 9.3 14.8 46.9 29.0
18-29, 6,200)
Unemployed (all member 43.8 38.2 13.2 4.8
unemployed, 46,100)
Economically inactive (all 32.5 54.5 7.4 5.5
member EI, 502,500)
Working poor (at least one 48.4 37.4 11.9 2.3
employed, 475,600)
Non-CSSA working poor 46.9 38.9 11.9 2.3
(454, 600)
For those homeowners living under income poverty, after policy intervention, 88.5% of
household were without mortgages. Within this group, 84.3% of them were economically
inactive and 58.2% of them are elderly.
For PRH households, there was 33.9% of them living in poverty before intervention.
The poverty rate dropped to 20.8% after recurrent cash welfare transfer (Figure 3-11).
136
Figure 3-11 Poverty rate and poor population by housing type, 2018
Housing cost of residents
The trends of housing price and rent were generally increasing in the past years as
mentioned above, however, the burden of housing cost is various to different type of
residents. For example, for the home buyers, the overall median mortgage/loan payment to
income ratio was 18.4 in 2016, which slight decreased comparing with 19.6 in 2011 (Table
3-5). The increase of percentage of mortgage/loan payment was not vigorous as housing
137
price. It may due to the decrease of interest rate. And the ratio decrease may due to the
increase of income of home buyers, who are usually upper class. But this needs more
investigation.
Table 3-5 Mortgage Payment and Loan Repayment
Population in Domestic Households, Median Monthly Domestic Household Mortgage Payment and Loan Repayment,
Median Mortgage Payment and Loan Repayment to Income Ratio and Domestic Households by Year and Household
Population in Median Monthly Domestic Median Mortgage Payment
Domestic Household Mortgage Payment and Loan Repayment to
Domestic
Households
Households and Loan Repayment(1) Income Ratio(1),(2)
Number of Persons Number HK$ Percentage
Year Household Size
1 367653 367653 7000 37.2
2 1071692 535846 8000 27.2
3 - 517108 - -
2006 4 - 504895 - -
5 - 213896 - -
6 and over - 87148 - -
Overall 6636909 2226546 7800 26.6
1 404088 404088 6240 26.6
2 1195394 597697 6900 19.5
3 1725948 575316 6420 19.9
2011 4 2007380 501845 7100 18.4
5 1062635 212527 9060 17.5
6 and over 496214 77323 10000 17.6
Overall 6891659 2368796 7000 19.6
1 459015 459015 8500 26.1
2 1331680 665840 9000 19.3
3 1834467 611489 8300 18.6
2016 4 1959332 489833 9500 16.8
5 1007885 201577 12100 15.9
6 and over 521757 81980 15000 16.1
Overall 7114136 2509734 9500 18.4
On the other hand, for the renters, there are mainly two types, public housing, and
private permanent housing. In 2016, the median rent to income ratio of public housing renters
was 9.3%, the monthly rent was $1060 for a 1-person family and $1750 for a 3-person
family. However, the rent to income ratio was 29.3% for renters living in private housings.
The median rent was $8,000 for a 1-person family and $10250 for a 4-person family (Table
3-6). The housing cost of home buyers, public housing renters and private housing renters are
138
very different.
Table 3-6 Median Monthly Domestic Household Rent and Median Rent to Income Ratio, by
Type of Housing and Household Size
Population in Domestic Households, Median Monthly Domestic Household Rent

and Median Rent to Income Ratio by Year, Type of Housing and Household Size
Median Monthly
Population in Median Rent to
Domestic Household (1),(2)
Domestic Households Income Ratio
Rent(1)
Number of Persons HK$ Percentage
Type of Household
Year
Housing Size
1 153899 1060 14.2
2 398530 1300 10.9
3 564951 1750 8.1
Public rental
4 589336 1870 7.2
housing
5 259335 2250 6.9
6 and over 134075 2380 6.7
Overall 2100126 1500 9.3
1 247701 8000 36.1
2 714332 8750 29.5
Private 3 917097 8500 28.6
2016 permanent 4 1011476 10250 26
housing 5 594245 15000 23.3
6 and over 304623 20000 25.4
Overall 3789474 9200 29.3
1 57415 4600 35.3
2 218818 4580 26.2
3 352419 4500 25.9
Other types of
4 358520 4500 18.4
housing
5 154305 5950 15.2
6 and over 83059 4570 11.7
Overall 1224536 4570 27.5
Housing benefit analysis
CSD attempted to estimate the effectiveness of public rental housing (PRH) on poverty
alleviation. The estimated welfare transfer of PRH to the poor household was $4,100 per
month in 2018. The total estimated transfer of PRH was 38.4 billion, which 35.7% of it was
enjoyed by the poor household, contributed in reducing 97,000 households from poverty or
3.6% point of poverty rate.
It is worth noting that the calculation of welfare transfer can be problematic. As the
139
calculation is based on the market rent, even if there are no change in the PRH provision and
rent, the welfare transfer may increase if the market rent rises. The report also denoted that
the estimated method used was “prudent and conservative” (HKSAR government, 2019).
There are different housing related benefits provided to qualified families currently. For
recurrent cash benefit, the main policy was rental subsidy of CSSA. It provides subsidy with
upper limit for CSSA recipients, for example, in Jan 2018, the upper limit for a singleton was
$1835, and for a 3-person family was $4825.
There are also non-recurrent cash benefit, such as rent subsidies for low-income public
housing tenants, one-off subsidy for elderly living in private housing and living subsidy for
low-income households not living in public housing and not receiving CSSA from
Community Care Fund (CCF). Other housing related in-kind benefits includes home
environment improvement scheme for the elderly, building maintenance grant scheme for
elderly owners.
Waiting list and time for public housing
Another crucial statistic for revealing housing poverty situation is the waiting time and
number of cases in waiting list of public housing. Since all of them needed to pass the means-
test for income and asset, the trend of waiting list may illustrate the housing demand of low-
income families. The number of cases in the waiting list kept in high level in the past years,
from around 110,100 in Jun 2008 to 282,200 in Jun 2015, and slightly dropped to 256,100 in
140
Jun 2019. The waiting time for general applicants also increased from 1.8 year in 2008 to 5.4
years in 2019 (Figure 3-12 and Figure 3-13). Most of the public housing applicants were low-
income families, and more than half of them were living in private housing. The longer
waiting time implied they need to live in the private housing market with high rent for longer
period.
Figure 3-12 Applications for public rental housing
Note. Adopted from LegCo (2019).
Figure 3-13 Average waiting time of public housing
Note. Adopted from LegCo (2019).
141
Demography of household living in sub-divided flats / inadequate housing
In Hong Kong, there were substantial number of poor people living in housing with small
living area, low quality, and high rent. It was sometimes called inadequate housing, but more
people used the term “subdivided units (SDU)” to describe certain kind of housing. The
government also used the naming in recent years and started conducting survey for
investigate the background of SDU and the resident living inside. There are four official
report, including the Policy 21 report in 2013, CSD thematic report in 2014 and 2015 and by-
census report in 2016 (CSD, 2016, 2017a; Policy 21, 2013).
In the report, SDU is defined as “are formed by splitting a unit of quarters into two or
more “internally connected” and “externally accessible” units commonly
for rental purposes” (CSD, 2016, p.13). The number of households living in SDU increased
from 66,900 (171,300 people) in 2013 to 92,700 (209,700 person) in 2016. In 2016, the
median living area per person was 5.3 square meter, the median rent and rent-to-income ratio
were $4,500 and 31.8% respectively. The report in 2013 showed that they faced numerous
housing problem such as water seepage (54.2%), concrete cracking (51.5%), messy electrical
wiring (49.1%) and exposed reinforcement (47.2%). Moreover, near half of them need to
move to other housing at least one time in three years. Generally, half of the residents had
applied and waiting for public housing. The results of these reports are summarized in the
table below (Table 3-7).
142
Table 3-7 Summary of official reports of SDU and residents.
Policy 21 Report CSD Thematic CSD Thematic By-Census

Report no. 57 Report no. 60 2016 Report
Period of Jan to Apr, 2013 Jun to Nov, 2014 May to Sept, 2015 Jun to Aug,2016
conduction
Coverage of 616,700 quarters 637 300 quarters 664 300 quarters N.A.
units
Interview Visited 1 860 Visited 2252 randomly N.A.
buildings, finished buildings，finished selected
5900 short 5075 interviews 2 134 buildings
questionnaires and Interviewed about 1
542 long 000
questionnaires households living in
SDUs
Number of 18800 (averagely 24600 (averagely 25200 (averagely 27100
quarters 3.6 units per 3.5 units per 3.5 units per (averagely 3.4
with SDUs quarters) quarters) quarters) units per
quarters)
Number of 66900 86400 88800 92700
SDUs
No. of 66900 household； 85500 household； 87600 household； 91800
household 171300 people 195500 people 199900 people household；
and people 209700 people
Living area Below 70 feet 2: Below 7 m ︰15.3% Below 7 m ︰13.4% Below 7 m2︰
2 2
10.8% 7 to less than 7 to less than 12%

70 to139 feet2 13m2︰58.9% 13m2︰65.2% 7 to less than
:55.4% Median︰9.5 m2 Median︰10.3 m2 13m2︰66.9%
Median︰10 m2
Area of Mean︰67.6 feet2 Mean︰5.7m2 Mean︰5.8 m2 Median︰5.3
SDUs per (6.2m2) Median︰4.0 m2 Median︰4.5 m2 m2
capita
Monthly Mean︰$3790 Median︰$3800 Median︰$4200 Median︰$4500
Rent (average rent per
feet2: $29.1)
Rent-to- Median︰29.2% Median︰30.8% Median︰32.3% Median︰
income (1-person: 35.3) 31.8%
ratio
143
Household Employment Median︰$11800 Median︰$12500 Median︰
income income: $13500
$10,000-$14,999:
29.5%
$15,000: 38.1%
New household members N.A. N.A. 25.6% of
arrivals were new arrivals: households had
35.4% at least one NA
member
Moved At least 1 time: At least 1 time: At least 1 time: At least 1 time
house in the 47.4% 48.8% 48.3% internally
past 3 years (3 or more: 1.2%) (3 or more: 8.1%) (3 or more: 9.4%) migrated:
17.2% (in past 5
years)
Applied for Yes: 49.6% Yes: 48.9% Yes: 46.8% N.A.
PRH
Note. Adapted from CSD (CSD, 2015, 2016, 2017a).
144
3.5 Political economy background of Hong Kong
Social situation and policy setting are deeply affected by a city’s political economy
background. To investigate the factors behind the current poverty situation, understanding the
structural force and its impact is crucial. The political economy background of Hong Kong,
such as open economy, city position in globalisation, relation with China, corporatist features
and position of welfare regime, has profound impact on the housing and welfare policies as
well as the current social situation.
Minimal, developmental, and open economy
Hong Kong has several state features. Ma (2007a, 2007b) performed a detailed study of
Hong Kong in the colonial and postcolonial period and summarised the contrasting images of
the colonial government as a minimalist and an interventionist state. The former image was
supported by the financial philosophy of the Hong Kong government and the discourse and
implementation of the economic strategy, ‘positive noninterventionism’ and ‘laissez-faire’
(Chan, 1998). The latter was supported by focusing on the interventions made by the state;
for example, the provision of public housing (Castells, Goh, & Kwok, 1990), claiming that
the state had continuously intervened the society by different policies. The colonial
government was also analysed as executive dominant in policy setting and utilised
administrative absorption to work with the opposite parties (King, 1975).
On the contrary, some scholars suggested that the Hong Kong government actively
145
intervened in the country’s economic development. Holliday (2000) described Hong Kong as
facilitative state, which is a sub-division of productivist welfare regime, with minimal effort
in promoting social right, limited stratification effect and prioritised market. Kwon (2005)
suggested that the developmental welfare state of Hong Kong has remained and shown strong
continuity before and after Asian financial crisis, and economic development is still the top
priority of the state. In addition, Hong Kong is an open economy for a long time. For
example, the Heritage Foundation ranked Hong Kong as the freest economy in 2017 for 23
consecutive years, and the Financial Secretary welcomed the ranking (HKSAR, 2017b).
City position of Hong Kong in global economy
Evidently, Hong Kong is an international city under great influence of globalisation,
including the aspects of finance, information technology and capital flow (Yip & LaGrange,
2006). The housing aspect is inevitably affected. The world city hypothesis suggested that
under the current global capital flow, certain cities become critical nodes in global economic.
Those world cities may not need strong support from local manufactory economy, but they
can open externally for global finance to maintain their economic growth (Friedmann, 1986,
2001; Sassen, 1998, 2011). Hong Kong is a classic world city with features of regional node.
Although world cities show similarities in economic characteristics, their difference in terms
of city history, policy and cultural background requires detailed investigation.
China factor and corporatist features
146
The China factor is influential in postcolonial Hong Kong. The state of Hong Kong is
not ruled or governed by a chief executive or appointed secretaries; it is under the increasing
influence of the Chinese government in the economic, political, social, and cultural arenas.
The economic businesses between China and Hong Kong Special Administrative Region are
increasing; for example, the implementation of the Closer Economic Partnership
Arrangement and ‘Belt and Road Initiative’ facilitate the trading between Mainland and Hong
Kong. The Individual Visit Scheme, which was viewed to have saved the Hong Kong
economy after the outbreak of SARS that started on July 2003, dramatically increased the
number of tourists visiting Hong Kong.
On the contrary, Ma (2007b, 2016) criticised that the previous analysis was extremely
state-centred, but the factor of corporatist was usually underestimated. Studies showed that
the business and professional groups are dominant in number and power in the advisory and
working committees of the Hong Kong government (Cheung & Wong, 2004). No political
organisation is available to centralise or to cooperate with the corporatists and professional,
and the business coalition is fragmented or incohesive (Lau, 1999). With detailed study of the
behaviour of functional constituency legislators in LegCo, Ma (2016) found that the policy
outputs, under the system of functional constituency serving interest of corporatist, are
sectoral oriented. The sectoral demands and intervention from professionals and corporatists
are influential in policy setting. The earlier description of Hong Kong state as autonomous or
147
noninterventionism is proposed to be no longer accurate, that is, Hong Kong is in a
corporatist state or under sectoral interventionism (Ma, 2016).
Brief review of housing development before 1997
This study chose the time of transfer of sovereignty in 1997 as a watershed and focused
on studying the housing development after 1997. However, ignoring the historical
background was unwise because this concept greatly impacts the current situation. On this
basis, the housing development before 1997 was briefly reviewed.
Chan (1999) proposed a stage framework for Hong Kong housing development including
the liberal, intervention, reforming and privatisation stages. The liberal stage was in the pre-
1954 period. No provision of public housing existed, and the government orientation of
housing was laissez-faire without active intervention. In the late 1940s, numerous Chinese
citizens migrated to Hong Kong due to internal political struggle in Mainland China. Several
settlement areas, such as lion rock, were found in that period. However, the housing
conditions were generally very poor. Study showed that the living density was extremely
high, for example, approximately 1.08 square meter per person in certain poor areas
(Hopkins, 1971). The housing society, as an independent and non-profit organisation, was
established in 1948 and started to provide housing for the poor in 1952.
The intervention stage was started in 1954 to early 1970s. Since the Shek Kip Mei
housing fire in 1953 which affected than 53,000 people, the government started to provide
148
housing in resettlement areas. The former Hong Kong Housing Authority (HKHA), primarily
a semi-independent organisation, was set up in 1954. It aimed at providing low-cost housing
for low- and middle-income families. Massive housing intervention was observed in this
stage, and the population in public housing estates was approximately one million in 1965
(HKHA, 2017). The reforming stage in the early 1970s to 1980s followed. The critical event
in this period was the 10-year housing plan that started in 1972, as initiated by the Governor
MacLehose. The plan aimed at providing public housing for 1.8 million people and
rebuilding the oldest settlement areas. Different housing organisations were coordinated, and
the Housing Authority responsible for the implementation of public housing planning and
management was reformed in 1973. On the contrary, the government started at implementing
the Home Ownership Scheme (HOS) policy in 1976, providing public subsidised housings
for low- and middle-income class.
The privatisation state was suggested in the mid-1980s. The long-term housing strategy
(LTHS) in 1987 stated an evident turn in housing policy, privatisation of public housing and
promotion of home ownership. The HKHA implemented the well-off tenant policy in 1987,
and some public housing residents were pushed to the private market. Moreover, since the
HOS was welcomed by the public, the HKHA planned to build more HOS instead of public
rental housing (Chan, 1999). On the contrary, LTHS in 1998 targeted that 70% of Hong Kong
families can own their houses in 10 years. In addition, the HOS Secondary Market Scheme
149
was set up in 1997 to promote house trading. Furthermore, the government started the
Tenants Purchase Scheme (TPS) to sell public rental housing to tenants in 1998 (HKHA,
2017). These policies showed a significant turn to housing privatisation.
‘Explanations’ of public housing provision
Although a global trend of declining public housing was observed due to the limiting
role of state, privatisation and regeneration, Hong Kong was exempted because its proportion
of public housing remains large in the past years (Lau and Murie, 2016). Public housing
served as a key solution for tackling poverty and housing problems and important for
investigating the ‘reasons’ behind public housing provision. Forrest and Yip (Forrest &Yip,
2014) summarised the four main arguments in explaining the substantial number of public
housing in Hong Kong, that is, benevolent state, political economy, reproduction of labour
and social stability. They argued that no single argument can provide comprehensive
explanations for the increase in the number of public housing at all time for a particular
period and social situation. They suggested that an integrated approach may provide a
thorough understanding of the expanding public housing sector.
The ‘benevolent state argument’ suggested that the government aims at providing
welfare for the needy and improving the basic living of citizens by building public housing
(Hopkins, 1971; Pryor, 1983). However, the reallocation of residents to public housing was
unsatisfactory, and the living space was still limited. In addition, the newly built public
150
housing was not mainly serving the applicants in the waiting list, at least before the late
1990s. The government’s kindness for all citizens living in poverty was questioned. The state
did not view housing as a social right or at similar provision level with education, medication
and public assistance, which target larger scale of poor population (Forrest & Yip, 2014).
From the ‘political economy perspective’, the development of public housing was
argued to be closely related to the economic development with political consideration. For
example, the massive replacement of poor squatters aimed at land development for industrial
sectors (Drakakis-Smith, 1979) or clearance of land to gain abundant revenues for selling
land to private developers (Cuthbert, 1991). This approach also echoed the view of
Regulationists. Building of public housing does not only meet the housing demands of
citizens, but it also aims at transferring public resources to private companies to employ
citizens and promoting economic growth and capital accumulation (Smart & Lee, 2003b).
This type of political economic explanation has limitations. If the state is totally focused on
industrial or real estate development and the maximisation of revenues from land, then
explaining the high proportion of public housing in Hong Kong is difficult. Evidently, a
leaning for the benefit of private developers and even a state–developer collusion was
observed. However, resilience from the public and other consideration of the government,
which needs other theories and further investigation, was also found.
The ‘reproduction of labour’ argument suggested by Castell (1990) proposed that the
151
construction of public housing was a means to provide social wage to the residents and hence
offer low-cost labour for industrial development. In this view, housing was perceived as
collective consumption, and the state provision of public housing indicates real income
subsidised by the government to the residents for lowering the reproduction cost of labour.
This approach not only supports the low-income families but also relieves the tension
between workers and the industrial sector. Nevertheless, Forrest and Yip (Forrest &Yip,
2014) argued that the social wage argument was ambivalent. The factories decided salaries
depending on their rent level and tenure type of public or private. The reproduction thesis
attempted to provide a framework to explain the state intervention for industrial development.
However, this argument does not work for explaining the public housing intervention after
1990s because factories had no large demand of low-cost labour. Furthermore, for the social
wage argument, distinguishing the nature of public housing as a subsidy to the workers or as
a welfare protection for the nonworking members in a family is difficult.
The thesis of ‘political consideration’ for social stability may be the most suitable for
explaining the massive government intervention in 1970s, especially the 10-year housing
policy in the governing period of Governor Murray MacLehose. This policy suggested that
public housing was a tool settle social crisis after riots in 1966 and 1967. The government
used public housing as a means for obtaining legitimacy and maintaining social stability
(Forrest and Yip, 2014). Most government decisions should involve political thinking. Thus,
152
the ‘political consideration’ argument can explain diverse types of state interventions.
However, a comprehensive analysis is required for validation. Therefore, to understand the
housing intervention in a situation with a highly complicated relationship among government,
market and civil society, such as after 1997, a deeper investigation of housing policies of
different periods of governance by different chief executives and the relationship and struggle
among the public, the legislative councilors, the market and the government is required.
153
3.6 Conclusion: Empirical study gaps and significance of research
Several important points conclude this chapter. Firstly, for academic literature, poverty
studies mainly focus on economic perspective, whereas some focus on special groups, such
as elderly and children. However, the research on housing poverty situation in Hong Kong
was limited. Secondly, no significant change was observed in income poverty rate, before and
after intervention, in the past years, though antipoverty relief and welfare expenditure
increased. Thirdly, the housing problems continuously became serious in the past years,
especially for the low-income group, including increasing housing price and rent and waiting
time for public housing. Fourthly, the political economy analysis provided a macro
background for understanding the existing housing and poverty situation and policy change.
The impacts of globalisation, welfare ideology of government and corporatists groups are
significant. Nevertheless, the power of civil society affected the policy setting and changed
the current situation.
This chapter shows several empirical research gaps including, 1) the absence of ‘housing
poverty line’ in Hong Kong. Although a calculation of housing transfer and the poverty
population after policy intervention is available in government reports, the housing poverty
situation, including ratio and residual housing poverty, is not measured, and compared among
years in previous studies. 2) The relationship of housing and non-income poverty is
insufficiently studied. Studies on the impact of housing factors on deprivation, social
154
exclusion and subjective poverty in Hong Kong are limited. 3) Previous studies on poverty or
housing poverty are usually in a statics view, but not in dynamics or historical aspects. The
dynamics within housing poverty, housing pathway and subjective experience of residents are
unavailable in Hong Kong literature. 4) Literature that explains the persistence of public
housing system is available, but it lacks comprehensive discussion on the existence of
housing poverty situation in Hong Kong.
155
Chapter 4 Philosophical Foundation, Research Method and Design
In social research, the selection of research methods and design is largely dependent on
the philosophical foundation or perspective of the researchers (Corbetta, 2003) which must be
deliberated before choosing the suitable approach. Firstly, the philosophical foundation of
social research is briefly reviewed. After comparing the social research paradigms, the
position of the researcher is revealed (Section 4.1). Secondly, the features of the quantitative
and qualitative approaches are discussed, and the rationale for using mixed methods and
secondary data analysis (SDA) is provided (Sections 4.2). Finally, the quantitative and
qualitative designs of this study and the corresponding research questions and hypotheses are
presented (Section 4.3).
156
4.1 Philosophical foundation of social research
4.1.1 Comparison of social research paradigms
Several basic philosophical questions must be addressed to investigate the social
phenomenon and human behaviour. What is reality? Does social reality exist? Can it be
understood, or is it knowledgeable? How can we know it? These questions are usually
categorised into three streams of philosophical tradition, namely, ontology, epistemology and
methodology (Corbetta, 2003).
Ontology is the study of being or the basic features of reality. It involves the constitution
of nature and reality (Crotty, 1998). By asking questions such as ‘What is the reality?’ or
‘What is the nature and form of society?’, researchers formulated the perception of how
things or realities are (Scotland, 2012). Epistemology is the study of knowledge, such as the
nature of knowledge; the ways knowledge is created and acquired; the scope of knowledge
and the relationship amongst knowledge, truth, belief, and justification. In social science, this
concept questions the knowability of social reality (i.e. whether the reality can be understood
by the observer) and focuses on the relationship between the researcher and the subject
(Corbetta, 2003; Denzin & Lincoln, 2000). Methodology is the study of methods, and it is
concerned with the strategy of actions or methods for studying the reality and the planning
behind the particular methods (Crotty, 1998). The study on methodology focuses on how data
are collected or analysed to fulfil research purpose and how the researchers study the reality
157
that they believed can be analysed (Denzin & Lincoln, 2000a; Scotland, 2012).
The aforementioned streams of philosophical dimensions are interrelated; identifying
clear boundaries amongst the three streams is difficult. Various systems of thought or
paradigms can be used to answer the questions in different philosophical streams or
traditions. Scholars developed various ‘paradigm comparison table’ to compare the
characteristics of such paradigms (Bryman, 1988; Corbetta, 2003; Crotty, 1998; Denzin &
Lincoln, 2000; Lincoln & Guba, 1985; Tashakkori & Teddlie, 1998). Kuo (2011) reviewed
the comparison tables and summarised four mainstream paradigms, namely, positivism, post-
positivism, constructionism/interpretivism and critical perspective, and the researcher then
defined four commonly used dimensions for analysis, including ontology, epistemology,
methodology and method. The comparison table of the ideas of various scholars and related
ideas is presented in Table 4-1.
Table 4-1 Comparison table of research paradigms, with reference of ideas of scholars
Positivism Post-positivism Interpretivism/ critical theory

Constructionism
Ontology Naïve realism; Critical realism; Relativism; the Historical
reality is “real” reality is “real” reality is socially realism; reality is
and knowable but cannot be constructed by constructed by
fully known; different social, economic,
progress by individuals or and cultural
falsification and community process
probabilistic
manner
Epistemology Knowledge is Modified dualism Subjective and Subjective and
objective; objectivity; interactionism; interactionism;
158
explanation with knowledge is result can be Researchers and
generalization; probabilistic; “created” the objects being
dualism- Multiple use of researched are
objectivity theories interdependent;
results are
“filtered”
Methodology Manipulative; Falsification; Interpretation; dialectical;
observation; manipulative; dialectical; induction; mostly
quantitative Mostly deduction Observer- qualitative
and quantitative, observed
with interaction;
supplementary of Induction;
qualitative Qualitative
Method Experiment; Modified Discourse Action research
survey; experiment; analysis;
Analysis by mixed method; narrative;
“variables” Analysis by autobiography;
“variables” case study
Note. Adapted from other frameworks (Bryman, 1988; Corbetta, 2003; Denzin & Lincoln,
2000a; Tuli, 2011).
4.1.2 Pragmaticism and critical realism
The debate on the philosophy of social research is usually caused by the selection
between quantitative or qualitative method or amongst post-positivism, constructionism, and
critical approach. These concepts have certain foundations as well as limitations and
weaknesses. Pragmaticism and critical realism are two of the major approaches for
determining the standpoints among the contentions or out of the debate. The former
emphasises the usefulness of research and actor experience, whereas the latter suggests
multiple dimensions of reality domain and the distinguishing approach of the acquired
159
knowledge. Both approaches can integrate quantitative and qualitative methods, but with
different philosophical foundations. After introducing both ideas, the position of the author of
a social research philosophy will be discussed.
Pragmaticism. One response to the paradigm debate is the rise of pragmaticism. Morgan
(2007) suggested that social science shifted to qualitative studies since the late 1970s and
scholars asked for an alternative to positivism afterwards (Lincoln & Guba, 1985). However,
he argued that the search for metaphysical paradigms is problematic and may produce
incommensurable knowledge. The pragmatic approach can deal with these problems
(Morgan, 2007)
The idea of pragmaticism is not new in the field of social science or academic discussion
(Feilzer, 2010; Howe, 1988). John Dewey promoted a philosophy that focuses on experience
instead of abstract concepts, and he investigated the sources of the beliefs and meanings of
actions for human beings and the linkage amongst them (Morgan, 2014). Dewey further
argued that the different research paradigms are similar in terms of the methods for searching
for the truth and reality, regardless if they are objective, subjective, or mixed. Hence, the
division of paradigms are not important, and attention must be directed on the relationship
between the studies and human experiences (Feilzer, 2010). Another pragmatist, Richard
Rorty, highlighted the practical dimension of philosophy and rejected the acquired knowledge
from the view of an outsider. Knowledge building should be based on experience and on
160
human history (Rorty, 1991, 1999). In summary, no fixed and unchanged ground exists for
any type of knowledge. In social research, pragmaticism focuses on what practically works.
The research objectives should be specific and meaningful, and the applied methods should
be appropriate for solving the problems or achieving the goals (Morgan, 2014).
However, pragmaticism does not provide a solid foundation in ontology and epistemology
and remains sceptical to the existence of external reality and knowledge acquisition. By
contrast, critical realism lies between positivism and constructionism but provides concrete
arguments and foundation for ontological and epistemological debates.
Critical realism. The idea or the movement of critical realism was initiated in 1970s as
a response to the challenges in natural science as well as an attempt to find a compromise
position between positivism and constructionism. Roy Bhaskar, a British science philosopher,
was one of the founders of critical realism. His book ‘The possibility of naturalism’ was
crucial in discussing the foundations of social science (Bhaskar, 1979). His idea of critical
realism can be divided into two parts, namely, ‘transcendental realism’ and ‘critical
naturalism’. The former challenged the traditional positivist approach of scientific practices
in natural science and argued that the causal relationship cannot be reduced and observed
through experiments. The latter suggested that transcendental realism can be applied in
human society, but the human world is more complicated than the physical world; thus, other
strategies are necessary (Chuang & Wei, 1998). Other scholars that support critical realism
161
includes Andrew Collier, Margaret Archer, Ted Benton and Andrew Sayer—all of whom were
influential in the field of social theory and philosophy in social science (Sayer, 2000).
Critical realism suggests the existence of an external reality that is independent of what
humans know or can know but can be understood by human (Baert, 2005; Sayer, 2010).
Reality has three domains, namely, empirical (empirical domain), actual (actual domain) and
real (real domain). The empirical domain is the subset of the actual domain, which is the
subset of the real domain. The empirical domain refers to the observation and sensation
experienced by people; empirical facts can be seen or experienced. The actual domain refers
to different events that can be predicted but cannot be experienced, which also includes
empirical reality. The actual domain are events that happen when the power behind the
objects are activated, which is the core concept of critical realism. This domain refers to the
mechanism behind the actual and the empirical events. The real event is the realm and the
power and structure of objects (Sayer, 2000); such events cannot be directly observed but can
affect the other layers. Some examples of real events include the tendencies behind causation,
the structure of language or the instinct of human beings (Chuang & Wei, 1998).
Objects can be classified into two types, namely, transitive and intransitive. Transitive
objects include facts that can be acquired and understood by human. Theories and science
discourses belong to the transitive dimension. The intransitive objects (e.g., real objects) are
the hidden mechanisms or structures behind human experiences (Baert, 2005). Sayer (2000,
162
2010) used Earth theory to illustrate the difference. Theories on the shape of Earth (i.e.
whether the earth is flat or round) can change and are therefore in transitive dimension.
However, the reality, which is in the intransitive dimension, does not change. This example
distinguishes critical realism from empirical one; the latter suggests searching the reality
using empirical experience. Bhaskar proposed four criteria for realism, namely, objectivity,
fallibility, trans-phenomenality, and counter-phenomenality. These criteria indicate that the
existence of reality is independent from knowability, but people can still obtain knowledge
from reality and can even determine the structure behind the appearance of reality.
Furthermore, knowledge can be falsified and amended (Collier, 1994). These ideas suggest
that people cannot reach the truth directly but can approach it through empirical observations
(Baert, 2005).
Critical realism rejects the naïve idea of causality or regularity (Figure 4-1). This
philosophical approach suggests that simple causation normally happens in closed systems
(e.g. controlled experience) but seldom occurs in an open one (e.g. society). However, the
majority of social phenomena occur in open systems. The experimental methods used in
natural sciences for closed systems cannot be simply copied in social sciences. The causal
processes vary in different contexts, especially in open systems. Similar causal powers may
produce different outcomes depending on the context, which is not fixed in an open system
(Figure 4-2) (Sayer, 2000).
163
Figure 4-1 Positivist or ‘successionist’ view of causation
Note. Adapted from Sayer (2000, p.14)
Figure 4-2 Critical realist view of causation
Note. Adapted from Sayer (2000, p.15)
Apart from searching causation, critical realism also involves the interpretive and
hermeneutic dimension in social studies and proposes that social phenomena are intrinsically
meaningful (Sayer, 2000). However, the meaning cannot be measured but must be understood
through interpretations.
4.1.3 Position of the researcher
The brief review on the philosophical discussions in social science reveals that certain
ideas must be rejected, and some must be further discussed. Instead of simply taking a certain
paradigm as a research foundation, the belief and preference of the researcher should be
elaborated. The key positions and reflections of the researcher are summarised as follows.
164
Firstly, the rejection of positivism does not necessarily mean abandonment of post-
positivism. Numerous valid critiques constitutes traditional positivism, especially those from
Thomas Kuhn (1962) and Karl Popper (1959), which imply that the objective truth proposed
by naïve realism was rejected. However, this implication does not mean that the empirical
spirit and foundation should be totally disregarded. The absolute certainty of social research
may be lost, but the probabilistic explanation and prediction can be still analysed and used as
they can reveal the possibility of studying different social situations and human behaviours
(Corbetta, 2003).
Secondly, alternatives, such as critical realism and pragmatism, are present between
positivism and constructionism. The debate between these two paradigms is ongoing.
However, the extreme types of both perspectives have limitations and defects. For example,
the relativism proposed by extreme constructionism is problematic because the description,
explanation and understatement become extremely subjective and relative. Nothing is true,
and the grounds for conducting the research are lacking. The attempts for finding a space
between these two paradigms in the past decades are increasing. Studies do not need to select
one and reject the other. The term ‘post-positivism’ can be considered as a position between
positivism and constructionism.
Thirdly, compared with pragmaticism, critical realism has a stronger theoretical ground
and fewer limitations. Although pragmaticism tried to sidestep the paradigm debate and focus
165
on the usefulness of the research, the corresponding philosophical foundation was weak.
Pragmaticism did not provide sufficient theoretical arguments to answer these ontological and
epistemological questions for social research, and the focus on human experience cannot
determine the existence of an external reality that is independent of human being or the
nature of knowledge. On the contrary, critical realism proposed a strong basis for the
philosophical arguments in ontology and epistemology.
Fourthly, critical realism provides relatively comprehensive responses to the criticisms
from other ideologies or school of thoughts, such as the response to the challenges of
postmodernism. Sayer (2000) argued that postmodernism tend to be anti-realism, but critical
realism does not need to fully reject its ideas. The ideas of the former revealed the diversity
and complexity of the world and provided a space for the plurality of approaches for viewing
humans and society. Critical realism accepts the diversified views but rejects the naïve
relativistic conclusion. In critical realism, real domains that are independent of human
observation and other layers of reality exist. This condition leads to the generation of the
assumptions of a certain type of social constructionism, which suggests that reality is socially
constructed through language, knowledge, or power. The constructed reality occurs in the
empirical or actual level but not in the real level because the latter is independent of what
humans think or act. This phenomenon contradicts the ideas of naïve relativism and thus
can destroy the meanings of social research (Collier, 1994; Sayer, 2010).
166
Lastly, the directions of social research are not limited within a certain paradigm. The
framework provided by Martin Hollis (1994) is useful for understanding social research
techniques. This framework has two dimensions, namely, (1) holism and individualism and
(2) explanation and understanding. The framework can be broadened by adding dimensions.
The dichotomy of holism and individualism disregards the roles of media, such as institutions
or organisations, when studying a complicated society. Moreover, instead of focusing on
certain corner in a statics view, intensive attention must be devoted on the interaction
between the two. The corresponding framework is displayed in Table 4-2.
Table 4-2 amended framework for social study
Explanation Understanding Change

Holism Systems “Games”
Medium (institutions,
organizations, etc.)
Individualism Agents Actors
Although this study did not purpose at articular the philosophy and theory of critical
realism in detail, the author adopts critical realism as the basic ontological and
epistemological foundations where an external and objective reality exists, and the
knowledge is fallible but acquirable. The research directions and designs of this thesis are
informed by the philosophy of critical realism. Objective and subjective approaches can be
used to acquire knowledge. The objectives of this study are to objectively describe and
explain social phenomenon and to understand the subjective views of human. Moreover, this
study aims to change the social injustice and inequality. The methodology used is a mixed
167
method with abductive approach. Mixed methods, including quantitative and qualitative, are
utilised to answer the research questions and present a comprehensive explanation of the
issues and subjects.
168
4.2 Research method
4.2.1 Quantitative and qualitative research
Nature of quantitative studies. The quantitative approach is generally based on a
positivist or post-positivist position (Corbetta, 2003; Tuli, 2011). By accumulating observable
and empirical facts, this approach evaluates theories using a deductive approach and searches
for the causality and causal relationship between the concepts and the variables, especially
dependent and independent ones. Advanced statistics techniques can be used to identify such
relationship (Bryman, 1988). Quantitative approaches aim to explore generalisation and
replication; such techniques investigate the profiles or demography of the samples, usually
through surveys, and then expand the descriptive or explanative results to the entire
population (Bryman, 1988).
Logical flow of quantitative studies. Bryman (1988) designed a logical flow for
quantitative research processes. The main phase started with the establishment of theories,
followed by the generation of the hypotheses deduced from theories, observations and then
data collection. The results are obtained after the analysis and data interpretation. The
findings further enrich theories through the induction process. The variables used in the
quantitative study are based on the conceptualisation and operationalisation process. The
concepts are operationalised as measurable variables, such as different indicators or indices
(Hollis, 1994). Although quantitative approaches are supposed to start with theories, grand
169
theories are excessively broad and abstract and offer limited guidelines for further empirical
research (Bryman, 1988). Several research methods are commonly applied in a quantitative
approach, such as social survey, experimental design, analysis of collected data and structural
observation. Statistical techniques are usually required in quantitative research (Bryman,
1988; Creswell, 2014; Tuli, 2011).
Strengths and limitation of quantitative studies. Quantitative studies are
advantageous in testing theories and hypotheses. With adequate sample size in data, these
approaches can generalise the finding to a related population. The data collection process is
structured and easy to follow and the results are relatively independent from the researchers
and obtains high credibility (Corbetta, 2003; Johnson & Onwuegbuzie, 2004). However,
several limitations are still present. For instance, the concepts and variables may be abstract
and difficult to apply in daily practices. Moreover, because of the objective approach, the
subjective positions of the participants are not considered and the investigation is often not
sufficiently thorough or diversified (Corbetta, 2003; Johnson & Onwuegbuzie, 2004).
Nature of qualitative studies. According to several scholars, quantitative studies have
no single definition or unique way of implementation (Neuman, 2014; Ritchie & Lewis,
2003; Rubin & Babbie, 2017). Several attempts were taken to identify the common
characteristics of such studies. Denzin and Lincoln (2005) proposed a generic definition
stating that ‘Qualitative research is a situated activity that locates the observer in the world. It
170
consists of a set of interpretive, material practices that makes the world visible. These
practices transform the world’ (Denzin & Lincoln, 2005, p.3). Other scholars concluded
certain common grounds and basic features of qualitative studies (Bryman, 1988; Creswell,
2014; Ritchie & Lewis, 2003). Research targets are usually placed in a natural setting of their
own context instead of in a laboratory. The meanings and ideas of the participants about
certain issues are the key data or frame of reference in the research. The data collected are
assumed to be rich in content. The researchers serve as the main instruments for the research
instead of questionnaires. Thus, they should maintain reflexivity and awareness of their roles
and position in the study and should realize the influence of their personal reflection,
background, and cultural and social assumptions on data interpretation.
Qualitative methods. Although qualitative studies are flexible, many scholars proposed
different approaches different foci, units of analysis, purposes and data collection methods
(Booth, Colom & Williams, 2008; Creswell, 2007; Ritchie & Lewis, 2003). Creswell (2007)
proposed five qualitative approaches, namely, narrative, phenomenology, grounded theory,
ethnography and case study. Multiple sources, including interviews, observations, and
documents, are used to obtain a comprehensive understanding of the cases. Qualitative
approaches are not limited to these five sources. Ritchie and Lewis (2003) suggested similar
divisions that focus on theory and philosophical foundations and added other categories, such
as symbolic interactionism, constructivism and critical theory. The commonly used methods
171
in qualitative studies include participatory observation, in-depth interviews, focus group
discussions, narrative study, document analysis and audio–visual analysis (Creswell, 2014;
Ritchie & Lewis, 2003).
Sampling in qualitative studies. Sampling is among the major issues in qualitative
research. The discussion for such activity includes the appropriate type of sampling, number
of required cases, logic or stages of sampling and types of sampling methods (Coyne, 1997;
Flick, 2009; Miles & Huberman, 1994; Morse, 1991; Patton, 1990; Strauss & Corbin, 1990).
The first issue is determination if the sampling should be random or purposeful.
Marshall (1996) argued that random sampling is not suitable in qualitative studies due to the
following reasons: (1) the general population is usually unknown; (2) the value and attitude in
the research process are not normally distributed; and (3) the random sample cannot provide
the target information. He then proposed three non-mutually exclusive sampling strategies,
namely, convenience, judgement, and theoretical sampling. Convenience sampling is based
on the accessibility and conservation of time and money. The judgement sampling targets the
samples who can properly answer the research questions. The last sampling strategy is driven
by theories, such as the grounded theory approach.
The second issue is the logic or criteria of sampling. Patton (1990) suggested 15
sampling strategies, including extreme or deviant case, maximum variation, homogeneous,
snowball or chain, theory-based, politically important and convenience sampling. Flick
172
(2007) proposed several crucial considerations for the selection of cases. Firstly, the typical
cases that represent the average or majority of population can be selected. Secondly, the
extreme or deviant cases should be selected to reflect the extremities of the situations (e.g.
poorest family). Lastly, the variation of cases should be maximised to reveal the differences
in the study area. Marshall (1996) proposed several sampling methods with specific purposes.
For example, the maximum variation sampling involves a broad subject range, and deviant
sampling aims to analyse outliers. Critical case sampling is concerned with specific
experiences, whereas key informant sampling is designed to investigate expertise and
confirming. Disconfirming sampling identifies the that cases agree or disagree with a certain
issue.
Another issue is the number of required cases. Padgett (2017) stated that the background
of informants is significantly related to the research topic, and the number of sampling is not
a critical factor. However, Sandelowski (1995) argued that the number cannot be extremely
small or extremely large and should depend on the purposes and target output of the study. A
certain number may be enough for homogeneous grouping but not for illustrating the
variation of complicated situations. Nevertheless, the variations in the sampling process, such
as demographic variations in the characteristics of people, phenomenal variations in the
features of the studied phenomenon and theoretical variations in the constructs of theories,
must be maximised (Sandelowski, 1995).
173
Merits and weaknesses of qualitative studies. Qualitative studies provide
comprehensive understanding of human behaviour and social phenomena because of the
involved in-depth analysis of cases. Moreover, these studies reveal the multiple subjective
views of different subjects and the contextual background of cases, which cannot easily
inferred using statistical figures (Hesse-Biber, 2010). However, qualitative studies cannot
satisfactorily provide generalised and objective results. Hence, providing predictions is
difficult and the credibility for policymaking or administration is low. In addition, data
processing is time consuming, and the results are subjective and easily affected by the view
of the researchers (Johnson & Onwuegbuzie, 2004).
Comparison of quantitative and qualitative approaches. Quantitative and quality
approaches are different in several aspects, including research foundation, use of theories,
research processes and research design. Scholars suggested different frameworks for
comparing quantitative and qualitative studies (Bryman, 1988; Corbetta, 2003; Crotty, 1998;
Fok & Hu, 2012; Neuman, 2014; Rubin & Babbie, 2017). Bryman (1988) emphasised that
quantitative researchers are distant from subject, whereas qualitative ones are often close. The
relationship between theories and research is respectively confirmative and emergent for
quantitative and qualitative studies. Corbetta (2003) compared the two approaches during the
stages of research planning, data collection, data analysis and result production. The
quantitative approach is more structured and logical than the qualitative one during the
174
planning stage. In addition, the relationship of the researcher and the subject is distant, and
the data collection process is standardised and manipulative. The relationships amongst the
variables are the object of the data analysis, and the result usually aims to produce
correlations and generalisations. For the qualitative study, the planning is relatively open and
interactive with a naturalistic approach. The data collection process is unstructured and non-
standardised. The analysis aims to understand the research subjects, and the results usually
produce specific themes and classification.
4.2.2 Mixed method research (MMR)
The trend for using mixed methods have been increasing in the past decades (Creswell &
Plano Clark, 2007; Hesse-Biber, 2010; Johnson, Onwuegbuzie & Turner, 2007; Teddlie &
Tashakkori, 2003).
Definition of MMR. Johnson et al. (2007) enumerated several important themes for
MMR, such as quantitative and qualitative, paradigm mixing, execution time and research
purpose and orientation. They defined MMR as ‘the type of research in which a researcher or
team of researchers combines elements of qualitative and quantitative research approaches
for the broad purposes of breadth and depth of understanding and corroboration’ (Johnson et
al., 2007, p.123). MMR can be analysed through a continuum (Figure 4-3)from pure
qualitative to pure quantitative, in which the centre area denotes equal status. In essence,
MMR lies between two extremes. If the MMR is leaning on the qualitative approach, then it
175
is labelled as ‘QUAL+quan’ or qualitative dominant research. Otherwise, MMR is denoted as
‘QUAN+qual’ or quantitative dominant research.
Figure 4-3 Continuum of mix methods research
Note. Adapted from Johnson et al. (2007, p.124).
Reasons for using MMR. Scholars cited the multiple functions and advantages of using
mixed methods instead of monomethod (Sung & Pan, 2010; Teddlie & Tashakkori, 2003).
For example, quantitative and qualitative approaches shared similar research goals, such
acquiring additional knowledge on the world and human behaviour (Sale, Lohfeld, & Brazil,
2002), and are therefore useful for studying complex social phenomenon with different
perspectives when applied together (Reichardt & Rallis, 1994). MMR has a high capacity to
deal with complex research questions (Hesse-Biber, 2010) and enhances the understanding of
reality in empirical and subjective processes (Johnson & Onwuegbuzie, 2004). Moreover,
MMR is compatible with different paradigms (Shannon-Baker, 2016). Greene, Caracelli and
Graham (1989) proposed the following important reasons for using MMR: 1) triangulation or
176
seeking the convergence and correspondence of various methods to increase the validity of
constructs; 2) complementarity or enhancing the result of one method using other method to
improve the interpretability and meanings; 3) development or using one method to develop
the other method (e.g. quantitative data can provide guidance for setting qualitative interview
questions); 4) initiation or determining the contradiction of frameworks and providing new
perspective to enhance the depth of the research by analysing different methods; and 5)
expansion or using different components from various methods to extend the scope of
investigation (Hesse-Biber, 2010).
Criticism of MMR and responses. Despite the advantages of using MMR, criticisms
and challenges still exist. Sale et al. (2002) highlighted the tension between qualitative and
quantitative paradigms. The ontological and epistemological assumptions of both paradigms
are different. However, they agreed that the two paradigms are not incompatible for several
reasons. For instance, both paradigms aim to understand the world and share the views of
value- and theory-laden enquiries. In addition, knowledge is fallible. MMR is useful in
studying complicated social phenomena. Although the qualitative and quantitative methods
are incommensurate, multiple methods can still be combined in a single research (Sale et al.,
2002). Morgan (2007) suggested the MMR is compatible with both approaches in several
ways. For example, MMR utilizes an abductive approach, which combines induction and
deduction to connection theories and data. MMR also implies intersubjectivity, which
177
considers the subjectivity and objectivity in the research process. However, rather than
proposing the use of MMR in the beginning stage, the research must start with the research
questions, hypotheses, and theories before deciding if MMR is appropriate for application.
Using MMR in this study. This study comprises several research questions that are
developed from the literature, including ‘How are the interactions between housing situations
and poverty?’, ‘What is the “housing poverty” situation?’, ‘Why does housing poverty exist,
persist or worsen?’ and ‘What is the experience of people living in poverty and housing
poverty situations?’. Using either quantitative or qualitative approach is not sufficient to
answer these questions. MMR is appropriate for addressing complicated research questions
(Hesse-Biber, 2010). For example, the overall picture housing poverty should be examined
using a survey (i.e. quantitative method). The experience of the underprivileged, on the other
hand, can be effectively analysed using qualitative approach, which is appropriate for
exploring the subjectivity of the informants. Moreover, the result of the quantitative research
can provide guidelines for the qualitative interviews (Greene et al., 1989). After conducting
the quantitative analysis, the questions can be answered through interviews; such interviews
can provide comprehensively explain the research results. Given that the quantitative part of
result is based on secondary data with limited variables, interviews can serve as
supplementary materials by asking questions that are not covered in the secondary dataset.
4.2.3 Secondary data analysis
178
Nature of SDA. Secondary data analysis is not a new concept (Glaser, 1963; Hakim,
1982). An increasing trend of using SDA in studies has been observed in the past decades
(Donnellan, Trzesniewski & Lucas, 2011; Hofferth, 2005; Johnston, 2014; Vartanian, 2011).
With the advancement of technology, large amounts of data are collected and organised and
datasets are becoming available and accessible (Johnston, 2014). Glaser (1963) defined
secondary analysis as ‘the study of specific problems through the analysis of existing data,
which were originally collected for other purposes’ (Glaser, 1963, p.11). Glass (1976)
proposed a definition with a similar meaning, stating that secondary analysis ‘is the re-
analysis of data for the purpose of answering the original research question with better
statistical techniques, or answering new questions with old data’ (Glass, 1976, p.3). Johnston
(2014) argued that ‘SDA is the analysis of data that were collected by someone else for
another primary purpose’ (Johnston, 2014, p.619). The secondary datasets are distinct types
of primary datasets that are originally collected to answer certain questions (Johnston, 2014).
Advantages of SDA. SDA has several advantages, including money- and time-saving
features, convenience, high research quality, enhanced representation, and suitability for
policy change. If the data needed to answer the research questions is available, then primary
data collection is not necessary. Using the existing datasets can save time (Donnellan et al.,
2011). When the data required to answer certain research questions are not feasible to collect,
using secondary data can conserve time resources (Vartanian, 2011). The data collection
179
process is not only time consuming, but is also costly (Hofferth, 2005). Secondary datasets
are usually free or have low cost (Vartanian, 2011). Moreover, these datasets obtains larger
sample size or longer period of longitudinal data than primary datasets, and can therefore
provide a highly representative and generalised result for the targeted population (Hofferth,
2005). A large sample size reduces the chances for violating the assumptions of certain
statistical models (Vartanian, 2011).
Limitations. One of the major limitations of using secondary datasets is the limited
information. Given that secondary data are not designed to answer the primary research
questions, researchers may have insufficient information to address such questions when the
secondary datasets are used (Donnellan et al., 2011; Hofferth, 2005). The frame or the
variable set in the secondary datasets cannot be controlled by the researchers. Moreover, the
definitions or the operationalisation of certain concepts may also differ from the primary
ideas of the researchers (Vartanian, 2011). If the researchers did not participate in the data
collection process, then they would have a limited understanding of many factors, such as the
face-to-face responses of the informants, difficulties in the survey or the reasons for missing
data (Johnston, 2014). In conclusion, the advantages of using secondary data outweigh the
demerits and limitations. The disadvantages can be reduced by the high awareness of the
researchers. Moreover, the use of other primary data can mitigate the demerit of using SDA.
Using SDA in this study. Given that the research questions are concerned with the
180
overall housing poverty situation in Hong Kong, SDA with large sample size and random
sampling are the most suitable data to answer the questions. In this study, the sample data of
the population census and another survey studying the poverty situation using random
sampling are used to examine the general housing situation in Hong Kong. The details will be
discussed in the subsequent Section. The researcher conducted a face-to-face interview to
supplement the quantitative data. The primary data collection will illustrate the limitations of
using SDA in answering the research questions.
181
4.3 Research design
4.3.1 QUAN: Corresponding research questions and hypotheses
The quantitative research design and analysis is based on a revised conceptual
framework studying the relationship amongst housing and poverty situation (Figure 4-4).
Figure 4-4 Revised conceptual framework of relationship amongst housing and poverty
situation for quantitative study
The quantitative part aims to answer the following research questions:
■ RQ2.1: To what extent do the housing factors affect the different dimensions of poverty?
■ RQ2.2: Which type of non-income poverty is the most affected by the housing factors?
■ RQ 2.3: Are housing factors the mediators of income and non-income poverty?
182
■ RQ3. Why does housing poverty exist, persist, or change in Hong Kong?
The corresponding hypotheses are:
■ Hypothesis 1: The housing poverty situation in Hong Kong is worsening in the past 20
years.
■ Hypothesis 2.1: Poor housing situations are associated with high level of poverty.
■ Hypothesis 2.2: The overall impact of the housing factors is different for each type of
non-income poverty situation.
■ Hypothesis 2.3:
Three types of housing factor, three types of non-income poverty and nine hypotheses
are derived.
■ H1 (SEM): The impact of income on deprivation is mediated by the housing cost.
■ H2 (SEM): The impact of income on deprivation is mediated by the living area.
■ H3 (SEM): The impact of income on deprivation is mediated by the indoor housing
problem.
■ H4 (SEM): The impact of income on social exclusion is mediated by the housing cost.
■ H5 (SEM): The impact of income on social exclusion is mediated by the living area.
■ H6 (SEM): The impact of income on social exclusion is mediated by the indoor housing
problem.
■ H7 (SEM): The impact of income on subjective poverty is mediated by the housing cost.
183
■ H8 (SEM): The impact of income on subjective poverty is mediated by the living area.
■ H9 (SEM): The impact of income on subjective poverty is mediated by indoor housing
problem.
Structural equation modelling (SEM) is adopted to test whether the impact of income on non-
income poverty is partially mediated by the housing factors.
4.3.2 QUAN: Data collection strategy
Source and background of the secondary datasets. The ‘Strategic Public Policy
Research (SPPR)’ and Census’ datasets were used to answer research questions.
The former was collected from the SPPR’s ‘Trends and Implications of Poverty and
Social Disadvantages in Hong Kong: A Multi-disciplinary and Longitudinal Study’ to
investigate the trend of poverty, deprivation, and health outcomes in Hong Kong by Professor
Hung Wong and his team from 2014 to 2016. The surveys were collected from a random
sample of Hong Kong households. The samples were selected from 25,000 addresses from
the Census and Statistics Department using a two-stage stratified method, which were drawn
by living quarters and living districts. Face-to-face interviews were conducted with the
sample households in 2014–2015.
Five Census datasets were used in this study, which are the 5% sample data of the census
survey in Hong Kong, including the by-census survey in 1996, 2006 and 2016 and census
survey in 2001 and 2011. The data were collected by the Census and Statistic Department.
184
Each Census dataset covered the demographic and socioeconomic characteristics of the Hong
Kong population. About one-seventh of all quarters in Hong Kong were sampled in the
datasets of 1996 and 2001 and about one-tenth were sampled in the 2006, 2011 and 2016
datasets.
Appropriateness of the datasets. The selected datasets are designed to answer some of
the research questions. The Census datasets are the largest datasets in Hong Kong that
involves random sampling; such datasets are appropriate for investigating the general
situation of Hong Kong residents. Several variables in the dataset are related to the research
theme. For example, the ‘income’ and ‘household size’ are related for poverty measurement,
whereas the ‘type of quarter’, ‘tenure of accommodation’, ‘rent’, ‘mortgage and loan
payment’ and ‘living district’ provide information for measuring housing affordability and
other housing situations. On the one hand, the Census datasets provide a broad and static
background of the housing and poverty situations in Hong Kong. On the other hand, the
SPPR dataset provides information for analysing multidimensional poverty and housing
situations, which involves the variables ‘income’, ‘deprivation’, ‘subjective poverty’ and
‘social exclusion’. To assess the housing situation, the variables ‘living area’, ‘indoor housing
problem’ and ‘rent’ are included.
The datasets were obtained by the researcher through the help of his supervisor, which
provides a convenient and feasible way to answer the research questions. Moreover, because
185
the Census datasets were obtained through an official survey, the research results are
representative and significant for policy implication. The concept operationalisation and
index construction are discussed in the next chapter.
4.3.3 QUAN: Data analysis strategy
The quantitative data analysis consists of three parts.
Firstly, the Census datasets were used to examine the income and housing poverty
situation from 1996–2016. The result of the overall population was examined before the
analysis by housing types. To achieve an in-depth investigation, the by-census data of 2016
were used to analyse the housing poverty situation through the demographic, economic and
housing characteristics of the households. This process aims to answer RQ1 and RQ3.
Secondly, the SPPR dataset was used to study the impact of housing factors on the non-
income poverty situation. Three logistic regression models are utilised. Logistic regression
analysis aims at finding out the risk factors of the dependent variables by measuring the
probability of a certain class. Deprivation, social exclusion, and subjective poverty are set as
the dependent variables (DVs) in Models 1, 2 and 3, respectively. The DVs was set as binary
for modelling. The demographic variables of the respondents are controlled in each model.
Multiple linear regression is conducted to measure the explanatory power of the housing
factors on the DVs. This step aims to address RQ2.1 and RQ2.2.
Finally, the SPPR dataset was used to determine whether the housing factors act as a
186
mediator between income and non-income poverty. SEM analysis was performed using the
AMOS software. SEM helps analysing independent variables, dependent variables and error
terms within a theoretical framework. It allows operationalization of the constructs, which
cannot directly be observed, with different observed variables. Moreover, SEM takes
measurement error into account in the analysis. It can also address the mediating and
moderating effect by estimation of model with variables with interconnections (Hoyle, 2012).
Income and housing factors were set as the main exogenous and mediating variables,
respectively, in the models. Deprivation, social exclusion, and subjective poverty were set as
the endogenous variable in the three SEM model. After the overall population was modelled,
the models were tested using various housing types. This part aims to answering RQ2.3.
4.3.4 QUAL: Corresponding research questions and hypotheses
The quantitative research design investigates the relationship of housing and poverty in an
objective and static perspective but cannot provide the description and explanation of the
dynamics and the process. In addition, the subjective understanding of the residents was
overlooked.
Individual in-depth interviews were conducted to obtain deep understanding of the
relationship between housing and poverty and the subjective view of the people living in
poverty or housing poverty situations. The respondents who showed a close relationship with
the research themes were invited for in-depth interviews. Note that this part only serves as a
187
supplement in this study.
■ RQ2.4 What are the impacts of housing factors on poverty and vice versa?
■ RQ4. What is the experience of people living in poverty and housing poverty situations?
■ RQ4.1 How do the housing factors affect the poverty situation and daily lives of the
residents?
■ RQ4.2 How do they understand the relationship of housing and poverty situation?
■ RQ4.3 How do they react, deal with, or attempt to change the existing situation?
No concrete hypothesis was established for the qualitative part.
4.3.5 QUAL: Data collection strategy
Data collection. Scholars (Morgan, 1996; Neuman, 2014; Padgett, 2017; Rubin &
Babbie, 2017; Seidman, 2013) proposed numerous preparation ideas for the data collection
from individual interviews, such as interview guides, case selection and recruitment and
researcher participation.
For the in-depth individual interviews, semi-structured face-to-face interviews were
conducted by the researcher. The interviews were conducted in the field or in the centres of
non-governmental organisations (NGOs) depending on the preference of the informants. The
interviews focused on the living history and experience of residents, such as the change of
living place, problems that they encountered and the impact of housing factors on their lives.
Interview guides were developed on the basis of the research questions and literature review.
188
The researcher was trained to acquire interview skills and familiarised with the topic before
conducting the interview. The respondents were recruited through the assistance of the social
workers in the NGOs. All interviews were conducted by the researcher. The people living in
different housing types, especially the poor families residing in inadequate and public
housing, are the research targets.
Case selection. On the basis of the above sampling strategies and the preliminary
quantitative data analysis, 20 cases were selected for the in-depth interviews depend on the
following considerations: residents under different housing situations (maximum variation
sampling), residents with different family backgrounds (maximum variation sampling),
households with severe housing poverty situation or problems (critical case and deviant
sampling) and public housing residents with different periods of living (maximum variation
or confirming and disconfirming sampling).
4.3.6 QUAL: Data analysis strategy and research quality
Data analysis. The data collected from the interviews were transcribed into words. The
researcher read through all the data before coding using the NVivo software. On the basis of
the theories, reviewed literature and research questions, the researcher identified meaningful
units by repeatedly analysing the data. The meaningful units were organised into themes,
which are abstract and conceptual in nature. Conceptual frameworks were formed. The
researcher read and compared the critical themes and relevant literature to formulate a
189
comprehensive conceptual map (Creswell, 2014; Padgett, 2017).
Ethical issue. Ethical considerations are crucial in conducting a social research. The
interviewer was aware of the ethical concerns, including informed consent, voluntary
participation, life quality and safety of informants and confidentiality (Padgett, 2017; Ritchie,
Lewis, Nicholls, & Ormston, 2014). The researcher applied for the approval of the Survey
and Behavioural Research Ethnics Committee of Chinese University of Hong Kong before
proceeding with this study.
Informed consent and voluntary participation. The researcher obtained the consent of
all informants before conducting the interviews. The interviewer introduced the background,
research purposes, usage of data collected and time schedule to the informants. The
researcher devised a consent form and invited the informants to sign the form before joining
the study (Appendix D). The participation was voluntary, and the informants were free to
withdraw any time.
Confidentiality. All collected data, including fieldnotes, recordings and interview
transcripts, were used for research purposes only. The raw data with the identities of the
informants will not be disclosed to other person or organisations. In the data analysis process
and report construction, the data with the identities were excluded. The informants were also
ensured that their information and the collected data will be kept confidential.
Research quality. The quality of this study was enhanced through several ways. Firstly,
190
triangulation and member checking (Denzin, 1971; Padgett, 2017). The researcher invited
other Ph.D. students to give their comments on the research process and data analysis. This
step can confirm if the coding and themes are understandable to further improve the
credibility (Lincoln & Guba, 1985; Padgett, 2017; Thomas & Magilvy, 2011). Secondly, data
management and review. The collected data and information, including the fieldnotes,
recording and documents, were kept and managed appropriately and with caution (Padgett,
2017). The data transcripts were reviewed several times to improve the credibility (Thomas &
Magilvy, 2011). Thirdly, interview quality enhancement. The researcher established
relationships with the informants to reduce the respondent’s bias by introducing himself
through the social workers before the interviews. Fourthly, personal experience. The
experience of the researcher in working for housing and poverty campaigns was beneficial in
selecting the suitable case, conducting interviews, interacting with informants, and enriching
the data analysis. The researcher remained critical and self-reflective during the research
process with respect to the impact of his personal values, belief and research ability and the
relationship between him and the informants (Kornbluh, 2015). Lastly, accurate reporting is
also essential. This report maintained transparency and presented a clear state of the research
processes, including data collection and sample selection, and remained open for criticisms
and questions from the readers.
191
Chapter 5 Quantitative Analysis: Part 1
5.1 Introduction of quantitative analysis
The result and discussion are divided into three parts. The first two parts are quantitative
analysis (Chapter 5 and Chapter 6), and the third part is qualitative analysis (Qual).
The first part of quantitative analysis (Chapter 5) focuses on the overall housing poverty
situation in Hong Kong. The housing poverty situation from 1996 to 2016 are examined
using the Hong Kong Census dataset, with focus on the circumstance in 2016 (Section 5.2).
The analysis of Hong Kong census data is followed by a discussion of the application of the
housing poverty concept in Hong Kong. The political economic explanation behind the
housing and income poverty situation is discussed as well (Section 5.3).
The second part of quantitative analysis (Chapter 6) focuses on the impact of housing
factors on different dimensions of poverty. The impacts of housing factors on deprivation,
social exclusion and subjective poverty are explored using the SPPR dataset and multiple
regression models (Section 6.1). The mediating impacts of housing factors between income
and non-income poverty are tested using SEM modelling (Section 6.2). The result is firstly
presented, followed by a discussion about the importance of the housing factor in poverty
studies and implication on Hong Kong’s situation (Section 6.3).
192
5.2 Housing poverty situation in Hong Kong (Census)
5.2.1 Data and sample
Population Census
The dataset used in this part was the 5% sample of the Hong Kong Population Census of
1996, 2001, 2006, 2011 and 2016. The data of Population Census were collected by the
Census and Statistics Department of Hong Kong every five years, by-census in 1996, 2006
and 2016 and census in 2001 and 2011. The 5% sample datasets were used because they
comprised the largest proportion of the sample available with random sampling for Hong
Kong population. Census covered the demographic and socio-economic characteristics of
Hong Kong population. About 10% of all quarters in Hong Kong were sampled, and all
households therein were included. This census is appropriate for investigating the living
situation of Hong Kong residents.
Generally, the five datasets contained data, including demographic background (e.g.
gender, age and educational level), economic characteristics (e.g. employment status and
income), household information (e.g. household size and family composition) and housing
attribute (e.g. housing type and housing cost). These data help in measuring the poverty and
housing situation of Hong Kong’s population. However, the data available in five census
datasets had little difference. The latter contained much substantial information. For example,
information about the living area was only available in the 2016 dataset.
193
The five datasets were used to measure the change in the overall poverty and housing
situation in the past 20 years. The 2016 dataset with updated and comprehensive information
was used for detailed analysis. The data items and sample size of household and persons in
each census dataset and total population in corresponding year were listed in Table 5-1.
Table 5-1 Background information of census data set from 1996 to 2016
Data No. of No. of No. of No. of Total

items household domestic persons persons in population
available households domestic in Hong
households Kong
(person)
1996 56 93,051 92,689 309,879 306,611 6,217,556
2001 73 108,140 102,776 332,632 323,757 6,708,389
2006 74 117,670 111,381 342,527 332,511 6,864,346
2011 80 126,171 121,723 364,127 354,151 7,071,576
2016 92 131,434 125,448 366,619 355,419 7,336,585
Corresponding research questions and hypothesis
The ‘Census’ datasets provide a broad and statistical background of housing and poverty
situation in Hong Kong. The following datasets are used for answering the following
questions:
■ RQ3. Why does housing poverty exist, persist, or change in Hong Kong?
The corresponding hypothesis of RQ1 is as follows:
■ Hypothesis 1: The housing poverty situation in Hong Kong is worsening in the past 20
years.
194
No concrete hypothesis is presented for RQ3, and it is answered in the discussion part.
5.2.2 Operationalization of concepts and measurements
Several poverty thresholds were constructed to measure and compare the housing and
poverty situation in Hong Kong in the past years. In this part, the main types of poverty
measurement are income and housing poverty. The definitions of each poverty threshold are
illustrated below.
Income poverty
Income poverty (50%): This poverty line threshold follows the definition of the official
poverty line of the Hong Kong government. The median income of different households with
varying sizes was calculated using census datasets of corresponding year. This threshold was
set at 50% of the household median income.
Equivalised household income (EQI) (50%): EQI was calculated by dividing the
household income by the square root of the household size. Households with income <50%
of the median EQI were counted as income poverty.
Housing poverty
The term ‘housing cost’ was firstly defined. ‘Housing cost’ of each household was
calculated as the sum of ‘monthly household rent’ and ‘monthly household mortgage
payment and loan repayment’.
Ratio housing poverty
195
Ratio housing poverty (25%): The concept of ratio housing poverty was widely used
amongst countries. This concept counts the ratio of housing cost to household income.
Households with ratio >25% were counted as housing poverty.
Housing poverty (30/40) indicator: This measurement of threshold was widely used in
Australia (Stone et al., 2011). It aims to identify the housing affordability for a relatively low-
income group and assumes that those households with high income that pay >30% of their
income on housing were given their own choice. By contrast, households with a household
income level in the bottom 40% of the place and pay >30% of their income on housing were
counted as facing housing affordability stress or living in housing poverty.
Residual housing poverty
The concept of residual housing poverty measures if the amount of after housing income
is enough to afford basic living. Those households with AH − B < 0 are counted as residual
housing poverty cases, where AH is after housing income, and B is basic living standard. The
basic living standard (B) has different definitions. The standards used in existing policy or in
previous research were used for calculation.
Residual housing poverty (public housing): The public rental housing (PRH) was
supposed to protect the basic living of residents. The income limitations for applying for
public housing were used to calculate the basic living standard. The amount of income
limitations consisted of the following components: housing cost, non-housing cost and
196
contingency. The non-housing cost was counted as the average non-housing expenditure of
the lower half expenditure group obtained from the Household Expenditure Survey. This
threshold was used for calculating the amount for basic living (Appendix B).
5.2.3 Statistical analysis strategy
The analysis firstly focused on the profile of by-census 2016 data, including the
demographic, economic and housing characteristics. Only domestic households and
individuals were selected because this study focused on the housing related situation. The
number of people living in income and housing poverty was then counted on the basis of the
above-mentioned definition. The results in different years were compared. Further detailed
comparison was conducted by splitting the result by housing types.
5.2.4 Result and Findings
a) Demographic background (2016)
In the sample of by-census 2016, males and females accounted for 45.9% and 54.1% of
the sample, respectively. With regard to age, children (0–17 years old), young adult (18–40
years old), adult (40–59 years old) and elderly (60 years old or above) comprised 14.1%,
31.0%, 33.1% and 21.8% of the sample, respectively. In terms of marital status, 52.2% of the
respondents were married, and the remaining 47.8% were never married, widowed, divorced,
or separated. With regard to education, 27.9% of the respondents obtained primary as the
highest educational attainment, whereas the remaining 44.6% and 27.5% finished secondary
197
and tertiary or above, respectively. In terms of occupation, 52.2% of the respondents have
fill-time work, whereas the remaining 45.1% were not working or economically inactive
(Table 5-2).
Table 5-2 Demographic information of respondents (Bi-Census 2016)
(N=355419) %
Sex Male 45.9
Female 54.1
Age 0-17 14.1
18-39 31.0
40-59 33.1
>=60 21.8
Marital status Married 52.2
Never married/Widowed/Divorced/Separated 47.8
Educational primary 27.9
Attainment secondary 44.6
tertiary or above 27.5
Occupation Full-time work 52.2
Part-time work 2.8
not working/economic inactive 45.1
In terms of households, the median values of the monthly and equivalised household
income were $25,000 (SD = 33637) and $15,019 (SD = 19858), respectively. The median of
income of people from main employment was $15,000 (SD = 22327).
b) Housing characteristics (2016)
The population mainly lived in six types of housing (Appendix A), including PRH
(29.6%), public subsidised housing with mortgage (3.4%), public subsidised housing without
mortgage (12.2%), private rental housing (15.3%), privately owned housing with mortgage
(15.0%) and privately owned housing without mortgage (19.9%). The other housing (4.5%)
198
included those staff quarters, collective living quarters and quarters provided by employers.
In terms of living location, 17.0% of the respondents lived in Hong Kong Island, 30.5% were
living in Kowloon, and 47.1% and 5.4% of the residents lived in new towns and other areas
of New Territories, respectively.
In terms of living area, residents mainly lived in housings with four categories of living
area, including <20 m2 (5.0%), 20 m2 to <40 m2 (36.2%), 40 m2 to <70 m2 (46.6%) and ≧70
m2 (12.2%). With regard to housing cost, the median of the monthly household rent was
$2153 (SD = 8959.18) and that of monthly household mortgage payment and loan repayment
was $8,123 (SD = 10102.42) (Table 5-3). In terms of family composition, 31.2% of the
respondents were living without elderly aged 60 years old or above and children aged under
15 years old. About 35.4% of the respondents lived with elderly but not with children, 25.4%
of them lived with children but without elderly, and 8.0% of the residents lived with elderly
and children.
Table 5-3 Housing cost of respondents (by household)
Mean Median SD
Monthly household rent (N=60838) $5494.67 $2153.00 8959.18

Monthly household mortgage payment $10365.93 $8123.00 10102.42
and loan repayment (N=20935)
c) Poverty situation (2016)
In terms of income poverty, this study used the relative income poverty (50%)
threshold and found that 19.0% of the population lived under poverty. It also used the
199
threshold of EQI poverty (50%) and found that 17.5% of the households lived under income
poverty. Three indicators were developed for measuring housing poverty. The housing
poverty rates of respondents were 15.9% for ratio housing poverty (25%), 4.9% for housing
poverty (30/40) indicator and 22.5% for residual housing poverty (PH) (Table 5-4).
Table 5-4 Income and housing poverty situation (by person) (2016)
%
Income poverty (50%) 19.0
EQI poverty (50%) 17.5
Ratio housing poverty (25%) 15.9
housing poverty (30/40) 4.9
residual housing poverty (PH) 22.5
The income and housing poverty rates were analysed by dividing the respondents into
groups by housing type, household size, living area of the living place, living location and
family composition.
By housing type
In terms of income poverty, the households living in PRH (26.0%), public subsidised
(without mortgage) (18.4%) and privately owned (without mortgage) (20.8%) had relatively
high EQI poverty rates. By contrast, the households living in privately owned (with
mortgage) housing had lower EQI poverty rate (just 3.1%) compared with the overall 17.5%.
With regard to the situation of ratio housing poverty, the public housing households had
low poverty rate (5.1%), whereas the private ones had a very high rate (61.2%). Respondents
living in privately owned houses (with mortgage) were also faced with a prominent level of
200
housing poverty (27.7%).
This study counted the bottom 40% of income group and found that the overall housing
poverty (30/40) was 4.9% and that of PRH was 3.1%. This finding indicates that a substantial
portion of low-income families were housed in existing public housing. However, the
housing poverty (30/40) indicator remained high, with 21.1% for the private rental housing
group (Table 5-5 and Figure 5-1).
The residual housing poverty rates were higher for those that needed paying mortgage
than those without. Although the respondents owned houses, the mortgage payment might
make them live under poverty.
Table 5-5 Income and housing poverty rate (%), by housing type (2016)
public public public private private private other

rental subsidiz subsidiz rental own own
ed (with ed (NO (with (NO
mort) mort) mort) mort)
Income poverty 26.9 10.0 20.1 15.1 4.3 22.0 20.3
(50%)
EQI poverty 26.0 7.3 18.4 12.2 3.1 20.8 19.1
(50%)
Ratio housing 5.1 19.4 0.0 61.2 27.7 0.0 5.0
poverty (25%)
housing poverty 3.1 5.4 0.0 21.1 3.1 0.0 1.2
(30/40)
residual housing 30.5 13.7 17.4 31.7 8.3 19.7 19.3
poverty (PH)
N 105159 12110 43524 54411 53359 70804 16052
201
Figure 5-1 Income and housing poverty rate (%), by housing type (2016)
By household size
All the income and housing poverty indicators of singletons were the highest amongst all
the groups. The EQI poverty rate was extraordinarily high (43.7%). The residual poverty
(PH) rate was high, thereby showing that 43.4% of the singleton respondents cannot afford
basic living after housing cost.
The households with large family size generally had low income and housing poverty
rates. For example, the ratio housing poverty (25%) rate and housing poverty (30/40) of a
four-person family were 14.0% and 3.7%, respectively (Table 5-6).
202
Table 5-6 Income and housing poverty rate (%), by household size (2016)
Household size 1 2 3 4 5 6 or
above
Income poverty (50%) 26.3 22.2 17.4 17.5 18.6 16.8
EQI poverty (50%) 43.7 30.7 15.8 11.3 8.5 7.6
Ratio housing poverty 25.7 17.6 15.1 14.0 14.9 15.6
(25%)
housing poverty (30/40) 13.0 7.4 4.9 3.7 2.1 1.1
residual housing 43.4 33.3 21.4 17.3 15.2 14.2
poverty (PH)
By living area
The income poverty rate generally inversely increased with the living area of households.
Households in small areas had high income poverty rate. For example, the EQI poverty
(50%) rate of those households living in <20 m2 place was 39.7%, whilst it was 8.4% for
those living in ≧70 m2 place.
The ratio housing poverty rates of households living in <20 and >70 m2 were high (43.5%
and 27.2%, respectively). By contrast, the housing poverty (30/40) indicators of those
households living in >70 m2 was just 1.7%. This finding showed that a certain amount of
households had high house cost but also high income.
In terms of residual housing poverty, households living in small living areas were faced
with quite severe housing poverty. The residual housing poverty (PH) rate was 54.5% for the
group living in a small area (Table 5-7).
203
Table 5-7 Income and housing poverty rate (%), by living area (2016)
<20m2 20 m2- 40 m2- >=70 m2

<40 m2 <70 m2
Income poverty (50%) 31.5 15.9 15.9 9.7
EQI poverty (50%) 39.7 13.0 13.0 8.4
Ratio housing poverty (25%) 43.5 14.1 14.1 27.2
housing poverty (30/40) 31.3 3.2 3.2 1.7
residual housing poverty (PH) 54.5 17.3 17.3 13.2
By location
Households in Hong Kong Island had a low EQI poverty rate (14.5%), whereas those in
Kowloon had a high one (18.7%). The after-housing income poverty rate in Kowloon was
also high (23.1%).
The ratio housing poverty (25%) in the new towns of New Territory (NT) was 12.6%,
which was much lower than that of Hong Kong Island (21.8%). This situation might be due
to the high proportion of public housing in the new towns (Table 5-8).
The residual housing poverty rates of those respondents living in Kowloon were slightly
higher than those living in other areas, but the difference was not obvious.
Table 5-8 Income and housing poverty rate (%), by location (2016)
Hong Kowloon NT (new NT (other)

Kong town)
Island
Income poverty (50%) 16.2 20.2 19.2 20.1
EQI poverty (50%) 14.5 18.7 17.8 17.9
Ratio housing poverty (25%) 21.8 17.1 12.6 20.0
housing poverty (30/40) 4.4 6.6 3.8 6.3
residual housing poverty (PH) 20.0 24.8 21.9 23.4
By family composition
204
Households with elderly generally had high income poverty rates. For example, the EQI
poverty (50%) rates of those households with elderly but no children were 26.1%, which was
more than double of those household without both (10.3%).
The ratio housing poverty (25%) and residual housing poverty (PH) rates of household
with children were 24.2% and 24.7%, respectively, which were the highest amongst
groups(Table 5-9).
Table 5-9 Income and housing poverty rate (%), by family composition (2016)
without both with elderly, with With both

elderly and no children children, no elderly and
children elderly children
Income poverty (50%) 9.7 23.3 21.6 28.4
EQI poverty (50%) 10.3 26.1 14.9 16.1
Ratio housing poverty 16.7 10.3 24.2 11.7
(25%)
housing poverty (30/40) 4.8 4.1 6.7 2.7
residual housing poverty 14.1 28.2 24.7 23.7
(PH)
By education
In terms of education, people with low educational attainment generally had high
poverty rate. The EQI poverty and residual housing poverty (PH) rates were comparatively
high for the group with primary education or below. By contrast, the EQI poverty rate of
those people with degree of above was 6.9%. This finding clearly showed the income
variation amongst people with different educational backgrounds.
However, the ratio housing poverty (25%) rate of those households with a degree or
205
above was 18.6%, which was the highest amongst the groups. Nevertheless, the residual
housing poverty rate was 9.8% (Table 5-10). Thus, these individuals could spend a large
amount of money in housing without sacrificing their basic living standards.
Table 5-10 Income and housing poverty rate (%), by education (2016)
Primary or Secondary Post- Degree or

below secondary/Sub- above
degree
Income poverty (50%) 28.4 19.5 11.3 7.4
EQI poverty (50%) 27.2 17.4 9.9 6.9
(25%)
housing poverty (30/40) 6.2 5.4 3.1 2.3
(PH)
By age group
The analysis by age group showed that the children and elderly greatly suffered from
income poverty. EQI poverty rate was 34.6% for individuals aged 60 years old or above.
Such a rate was nearly triple that of the groups aged 18–39 and 40–59 years old.
The ratio housing poverty rate of the elderly group was 11.4% which is the lowest
amongst the groups. This result may be due to the large portion of elderly living in public or
privately owned housing without mortgage. By contrast, the ratio and residual housing
poverty rates of the children group were 21.9% and 28.1%, respectively (Table 5-11). Thus,
families with children were probably under high housing cost burden.
206
Table 5-11 Income and housing poverty rate (%), by age group (2016)
0-17 18-39 40-59 60+

Income poverty (50%) 26.1 13.2 14.3 29.9
EQI poverty (50%) 18.0 10.5 12.7 34.6
(25%)
housing poverty (30/40) 6.6 4.6 4.2 5.2
(PH)
By place of birth
The place of birth of the cases were divided into three groups, namely, Hong Kong,
Mainland China and others. The income poverty rate of households born in China was higher
than those of other groups. The EQI poverty rate of China group was 26.2%, which was
nearly double than that of the group born in Hong Kong (13.7%). However, the ratio housing
poverty (25%) rate of the China group was 14.4%, thereby making them the lowest group
(Table 5-12).
In the crosstab analysis, the group of Mainland China exhibited a large population
with low education, consisted of older individuals and lived in PRH. This situation might
explain the high level of income poverty and low level of ratio housing poverty of the China
group.
207
Table 5-12 Income and housing poverty rate (%), by place of birth (2016)
HK Mainland other
China
Income poverty (50%) 15.5 27.0 16.6
EQI poverty (50%) 13.7 26.2 15.4
Ratio housing poverty (25%) 15.1 14.4 25.8
housing poverty (30/40) 3.9 6.9 4.9
residual housing poverty (PH) 18.1 31.9 21.8
By duration of residence
In terms of duration of residence, seven years was used as a threshold for analysis.
Those households that resided for <7 years were identified as new immigrants. No obvious
difference of EQI poverty (50%) rate was observed between both groups. However, the ratio
housing poverty rate of new immigrants (29.9%) was much higher than those households that
resided for ≧7 years. In the crosstab analysis with housing type, a larger proportion of
immigrants lived in private rental housing (35.8%) compared with the other groups (12.6%)
(Table 5-13). This finding may explain the reason that new immigrants are faced with a
severe ratio housing poverty situation.
Table 5-13 Income and housing poverty rate (%), by duration of residence (2016)
<7 years 7 years or more

Income poverty (50%) 21.6 18.7
EQI poverty (50%) 16.8 17.6
Ratio housing poverty (25%) 29.9 14.1
housing poverty (30/40) 9.7 4.2
residual housing poverty (PH) 27.1 21.9
d) Poverty situation from 1996 to 2016
In this session, the income and housing poverty situations from 1996 to 2016 were
208
compared. The statistics of the overall population amongst different years was firstly
analysed. The profiles were then analysed by division of housing types. Finally, the mortgage
payment was analysed to investigate the falling trend of the housing poverty rate in certain
groups.
Overall population
In the past 20 years, the income poverty rate generally increased from 1996 to 2006 and
slightly dropped from 2006 to 2016. The poverty rate in 2016 was nearly the same 20 years
ago. The overall ratio housing poverty (25%) dropped from 23.7% in 2001 to 15.9% in 2016.
The housing poverty (30/40) indicator remained relatively low (from 6.1% in 2001 to
4.9% in 2016). This phenomenon may be due to the proportion of low-income families living
in PRH, thereby providing low rent option for those with low income. The residual housing
poverty (PH) remained high for many years, but it dropped from 21.4% in 2001 to 18.2 in
2011 but increased to 22.5 in 2016 (Table 5-14 and Figure 5-2). Thus, a certain portion of
residents remained below the basic living standard.
Table 5-14 poverty situation from 1996 to 2016 (overall population)
Poverty rate % 1996 2001 2006 2011 2016

income poverty (50%) 18.1 19.7 20.8 20.2 19.0
EQI poverty (50%) 18.0 17.7 19.7 18.9 17.5
ratio housing poverty (25%) NA 23.7 23.0 15.4 15.9
housing poverty (30/40) NA 6.1 7.0 4.8 4.9
residual housing poverty (PH) NA 21.4 21.9 18.2 22.5
209
Figure 5-2 poverty situation from 1996 to 2016 (overall population)
PRH
The EQI poverty rate of PH residents boosted from 24.1% in 1996 to 32.8 in 2011 but
dropped to 26.0% in 2016. This rate is lower compared with the overall income rate. This
finding showed the substantial income improvement of the low-income families in public
housing.
The ratio housing poverty (25%) and housing poverty (30/40) remained low and even
dropped in the recent 10 years. Hence, the housing cost burden of public housing residents
declined. However, the relatively high level of residual housing poverty (PH) rate (30.5% in
2016) showed that part of the PH residents that still cannot afford basic living (Table 5-15
and Figure 5-3).
210
Table 5-15 poverty situation from 1996 to 2016 (public rental housing)
Poverty rate % 1996 2001 2006 2011 2016

income poverty (50%) 23.9 30.2 34.8 35.2 26.9
EQI poverty (50%) 24.1 26.7 32.7 32.8 26.0
ratio housing poverty (25%) 4.3 7.7 15.8 9.2 5.1
housing poverty (30/40) 2.5 4.6 10.1 5.8 3.1
residual housing poverty (PH) 27.5 29.8 34.2 30.1 30.5
% of total population 39.8 32.9 32.0 30.4 29.6
Figure 5-3 poverty situation from 1996 to 2016 (public rental housing)
Public subsidised housing with mortgage
The income poverty rate in this group was relatively low compared with the overall
population. The EQI poverty (50%) rate was 7.3 in 2016. Hence, the income level of
residents living in public subsidised housing was not very low. By contrast, a substantial drop
of ratio housing poverty rate can be observed from 39.2% in 2001 to 19.4% in 2016 (Table
5-16). Further analysis of mortgage payment was needed.
211
Table 5-16 poverty situation from 1996 to 2016 (public subsidized housing with mortgage)
Poverty rate % 1996 2001 2006 2011 2016

income poverty (50%) NA 8.7 10.3 8.8 10.0
EQI poverty (50%) NA 7.4 9.4 7.7 7.3
% of total population 10.7 8.6 5.8 3.4
Public subsidised housing without mortgage
The number of people living in this housing type without mortgage increased (104.8%
from 2001 to 2016). The residents in this type of housing did not have any housing cost.
However, these residents did have a relatively high EQI poverty rate, and it increased from
17.1% in 2001 to 18.4% in 2016 (Table 5-17). The poverty rates in this group did not greatly
change in the past 20 years compared with those of other housing groups.
Table 5-17 poverty situation from 1996 to 2016 (public subsidized housing without
mortgage)
Poverty rate % 1996 2001 2006 2011 2016

EQI poverty (50%) NA 17.1 18.1 16.7 18.4
Private rental housing
The percentage of residents living in private rental housing increased from 10.1% of the
212
overall population in 1996 to 12.0% in 2011 and further increased to 15.3% in 2016.
The poverty rate of most dimensions declined from 1996 to 2011 but increased from 2011
to 2016. For example, the income poverty (50%) rate increased from 11.9% in 2011 to 15.1 in
2016. Thus, many low-income families were forced to live in private rental housing.
The ratio housing rate maintained a high level in the 20 years and was extremely high in
2016 (61.2%). The residual housing poverty (PH) rate was the highest amongst various
housing types (31.7% in 2016) (Table 5-18 and Figure 5-4). Hence, the housing poverty
measurements correspond to the housing situation (with increasing rent) which cannot be
measured by using the income poverty approach.
Table 5-18 poverty situation from 1996 to 2016 (private rental housing)
Poverty rate % 1996 2001 2006 2011 2016

income poverty (50%) 18.8 19.5 14.0 11.9 15.1
EQI poverty (50%) 18.9 17.6 14.0 11.2 12.2
ratio housing poverty (25%) 52.7 58.1 49.6 50.6 61.2
housing poverty (30/40) 20.4 25.0 19.9 15.6 21.1
residual housing poverty (PH) 29.8 31.0 23.4 17.9 31.7
% of total population 10.1 11.2 10.5 12.0 15.3
213
Figure 5-4 poverty situation from 1996 to 2016 (private rental housing)
Privately owned housing with mortgage
Income poverty rate was the lowest amongst different housing groups. The EQI poverty
(50%) rate was 3.1 in 2016. The residual housing poverty rates were also low. Thus, the
living standard of residents was relatively good.
The ratio housing poverty (25%) rate was very high in 2001 (60.5%). However, such rate
declined by 27.7% in 2016 (Table 5-19), thereby showing that the housing cost burden of
families paying mortgage in this group significantly dropped.
Table 5-19 poverty situation from 1996 to 2016 (private owned housing with mortgage)
Poverty rate % 1996 2001 2006 2011 2016

EQI poverty (50%) NA 4.4 5.2 3.7 3.1
214
Privately owned housing without mortgage
This group comprised well-off residents who are living in owned housing with housing
cost. However, these people had relatively high-income EDI poverty rate (20.8 in 2016),
which was even higher than that of the overall population (Table 5-20).
The residual housing poverty (PH) rate was also relatively high (19.7% in 2016). Hence,
a substantial proportion of resident cannot sustain basic living even if they owned their own
houses.
Table 5-20 poverty situation from 1996 to 2016 (private owned housing without mortgage)
Poverty rate % 1996 2001 2006 2011 2016

EQI poverty (50%) NA 19.9 20.8 20.5 20.8
Mortgage payment analysis
In the analysis of the poverty rate of the overall population, the housing poverty rate
generally dropped in the recent 10 years despite of the increase in rent and housing price in
that period. The number of households living in owned housing was studied with their
mortgage payment to investigate this case. The analysis was based on samples from 2001 to
2016 because no datum for mortgage payment in 1996 was available.
The total percentage of sample persons that needed to pay mortgage dropped from 52.0%
in 2001 to 36.4% in 2016. The percentage in the public subsidised housing group dropped.
215
About 34,521 out of the 55,775 cases needed to pay mortgage in 2001 but only 12,110 out of
the 55,634 cases in 2016. The percentage of cases with mortgage in the public subsidised
housing group dropped from 61.9% in 2001 to 21.8% in 2016. The percentage in the
privately owned housing group dropped from 47.2% in 2001 to 43.0% in 2016 (Table 5-21
and Figure 5-5).
Table 5-21 Mortgage payment analysis (2001 to 2016)
2001 2006 2011 2016

public subsidized, with mortgage 61.9% 48.6% 34.1% 21.8%
/total (%)
private own, with mortgage /total 47.2% 47.8% 44.1% 43.0%
(%)
Owned, with mortgage /total (%) 52.0% 48.1% 41.0% 36.4%
Figure 5-5 Mortgage payment analysis (2001 to 2016)
216
5.3 Discussion for census data analysis
5.3.1 How is the income poverty situation in the past 20 years and why?
Poverty has been a long-lasting problem in Hong Kong. The government addressed the
issue and initiated various anti-poverty policies in the past years (Goodstadt, 2013). However,
the income poverty rate of the overall population remained almost unchanged in the past 20
years. Although a minimum wage legislation was passed in 2011 and the wage of labour was
generally increased (HKSAR government, 2017), the EQI poverty rate of the overall
population just slightly dropped from 18.9% in 2011 to 17.5% in 2016 compared with 18.0%
in 1996. The income inequality remained serious and caused the high level of income poverty
rate in the past decades.
The income poverty rate was analysed by housing type, and the result showed that the
income poverty rate (50%) of those households living in PRH decreased from 35.2% in 2011
to 26.9% in 2016, whereas those living in private rental housing increased from 11.9% to
15.1% in the same period. The percentage of population living in private rental housing
increased from 12.0% in 2011 to 15.3% in 2016 which was the highest proportion in the past
20 years. Several factors possibly influenced such phenomenon. One possible reason is that
the legislation of minimum wage showed a significant effect on the low-income groups, and a
certain portion of households were living in PRH. By contrast, the percentage of people
living in private rental housing increased from 12.0% in 2011 to 15.3% in 2016. This
217
phenomenon is possible because low-income families were forced to live in private rental
housing as the waiting time of public housing was prolonged, thus increasing income poverty
rate.
The income poverty analysis showed that the Hong Kong government failed to alleviate
income poverty by regulating the labour market. This finding is consistent with the local
research about the working poor in Hong Kong (Cheung & Chou, 2016; Wong, 2015). The
legislation of minimum wage enhanced the income protection of the working poor; however,
it did not show a significant effect on the reduction of income disparity, which is reflected by
the income poverty rate. Aging is another factor contributing to the high level of income
poverty rate. Lee and Chou (2016) explained that the decrease of the number of earners in
families with elderly was the main cause of increase in elderly poverty rate. The result of this
study is consistent with the above-mentioned findings. Although the EQI measurement
considered the number of members in the household, the poverty rate remains high in the
elderly groups. Overall, this analysis was limited because it did not consider the effect of
social welfare for income enhancement. However, the report of the Commission of Poverty in
recent years showed that the income poverty rate after government intervention remained
stably high (HKSAR government, 2019).
5.3.2 What is news for using housing poverty measurements?
The poverty measurement in Hong Kong was dominated by the income approach, which
218
was used as an official poverty measurement by the Commission of Poverty. However, the
existing measurement cannot comprehensively describe the poverty situation. This
measurement only showed part of the whole picture and particularly overlooked the housing
situation, which was crucial in a global city with high housing price and rent and relatively
high proportion of public housing (Wong & Chan, 2019).
Firstly, the existing official poverty line uses the 50% of the median household income as
the threshold. However, this measurement underestimated the poverty situation faced by the
singletons and two-person families, especially the working population. The official poverty
lines of one and two persons were in a low level because the proportions of single elderly and
elderly couples without income are high (for example, $4,000 HKD and $9,000, respectively,
in 2016). This situation indicates that those people working with minimum wage or receiving
CSSA were most likely not counted as poverty. On this basis, the EQI measurement adjusted
the effect of household size. The EQI poverty line (50%) threshold was approximately $7,500
HKD in the 2016 by-census data. This mechanism was more reasonable than that of the
official poverty in counting the poverty situation particularly to the singletons and two-person
families. The literature review indicated that the existing income poverty measurement
cannot capture the essential expense of households and hence failed to describe the full
picture of the poverty situation (Alcock, 1997; Townsend, 1979b), especially for those with
high housing cost (Stephens & vanSteen, 2011). Three housing poverty lines were
219
constructed to fill this gap.
The ratio housing poverty line was a convenient indicator to describe the housing cost
burden to the households (Cai & Lu, 2015). This indicator showed a distinct picture from the
income poverty measurement. For example, the analysis of the by-census 2016 data indicated
that the income poverty rate of private rental housing residents was just 15.1%, which is even
lower than that of the overall population (19.0%). However, this rate does not include the
high housing cost that residents pay. The ratio housing poverty rate was 61.2%, thereby
showing that a substantial proportion of private rental housing residents faced housing
affordability problem. The other example was the residents living in privately owned housing
with mortgage. The income poverty rate of such group was just 4.3%, but the ratio housing
poverty rate was 27.7%. The housing cost burden of the high-income group may be
overlooked when the income poverty analysis is applied.
The housing poverty (30/40) indicator was applied to supplement the limitation of the
ratio housing poverty measurement. These relatively high criteria may result in lower poverty
rate compared with those of the other measurements because this indicator only counted the
lowest 40% income group (Daniel, Baker, & Lester, 2018). Nevertheless, the indicator
captured ‘the housing poor of the income poor’ (Cai & Lu, 2015). The housing poverty rate
was just 4.9% in the overall population when this indicator was used, but it was 21.1% in the
group of private rental housing residents. This rate excludes the high-income group who
220
‘choose’ to spend high housing cost for optimal housing quality. The analysis of poverty by
living area indicated that the housing poverty (30/40) rate for the group of <20 m2 was
31.3%, which is much higher than those of other groups. This finding indicated that the low-
income families lived in a small flat and even spent a considerable proportion of their income
on rent. The housing poverty (30/40) indicators increased from 15.6% in 2011 to 21.1% in
2016. Such finding was consistent with the increasing trend of rent in the housing market.
This indicator was more accurate in reflecting the housing burden of the grassroots than using
the income poverty line.
The focus of residual housing poverty is to measure whether the households can afford
basic living after paying the housing cost (Kutty, 2005). Housing cost is an essential expense
to many households. This indicator highlights the number of people that suffered after paying
the housing cost. The residual housing poverty rate of private rental housing residents is
31.7% in 2016, which is the highest amongst different housing residents. This rate captures
the high housing cost burden faced by the residents. The use of the income poverty rate
(15.1%) cannot depict the number of residents that were unable to maintain a basic living
standard after paying the housing cost. In contrast with the ratio housing line and housing
(30/40) indicator, the residual housing poverty line also accounts the poverty situation of
those residents without housing cost. For example, the residual housing poverty rate of
residents living in privately owned housing without mortgage payment was 19.7% in 2016.
221
This notion indicates that a substantial number of households cannot sustain a basic living
standard even without any housing cost. Nevertheless, which level the threshold of basic
living should be set is still debatable. Further studies are needed to measure the basic living
standard in Hong Kong.
Overall, the three measurements of housing poverty not only supplement the weaknesses
of income poverty measurement but also capture the housing cost burden faced by
households. The use of three housing poverty lines reflects different dimensions of housing-
induced poverty (Hancock, 1993; Kutty, 2005; Thalmann, 1999). The three housing poverty
measurements have various strengths and weaknesses. For example, the ratio housing poverty
measurement is easy to understand and to compare internationally but it may count those
living in high quality housing as poor. The residual housing poverty approach is strong at
identifying the poor who cannot afford basic living, but the measurement of basic living
standard is controversial. More details for comparing these three housing poverty
measurements will be discuss in the conclusion session. This result agrees that the use of
composite methods in measuring housing poverty can help in demonstrating a comprehensive
view of society and enhance policy design and implementation (Li, 2014; Thalmann, 2003).
The housing poverty measurements were important especially in the Hong Kong context with
high housing price and rent.
222
5.3.3 Whose ‘housing problems’? Whose ‘housing profit’?
‘The housing problem is serious in Hong Kong.’ This statement is generally agreed by the
Hong Kong government officials, media, and the public. Nevertheless, this statement was
general and simple to describe the housing situation in Hong Kong. Further questioning with
regard to who benefits or suffers from the housing situation, how serious it was and how it
changed over years should be conducted. This research broadens the understanding of
housing problem analysis in previous studies that focus only on certain types of housing in
Hong Kong (Delang & Lung, 2010; Monkkonen, 2011).
The analysis of the overall population showed that the ratio poverty rate declined from
23.0% in 2006 to 15.4 in 2011 and slightly increased to 15.9 in 2016. The housing poverty
(30/40) indicator also demonstrated a similar trend in that period. This circumstance seems
showing a contradicting picture with the boost of housing price and rent in the recent 10
years. The analysis of census data by housing type may answer this puzzle.
The following are the factors that contribute to the drop of ratio housing poverty rate.
Firstly, the ratio poverty rate of PRH residents reduced from 15.8% in 2006 to 9.2% in 2011
and further dropped to 5.1% in 2016. This decline may be caused by the increase of income
of residents with stable situation and low rent level in public housing. The PRH residents
occupied approximately 30% of population in Hong Kong; this situation was one of the
critical elements for the decrease of ratio housing poverty of the overall population. Secondly,
223
the proportion of residents living in owned houses with mortgage payment gradually dropped
from 52.0% in 2001 to 36.4% in 2016. The proportion of people living in public subsidised
housing even declined from 48.6% to 21.8% in the same period. The residents in these types
of housing did not have to pay rent or mortgage, and the ratio housing poverty rate was zero.
This phenomenon was another crucial factor for the overall drop of ratio housing poverty. By
contrast, the ratio housing poverty rate of private rental residents increased from 49.6% in
2006 to 61.2% in 2016, which was extremely high compared with those of the other groups.
Although the number of people living in private rental housing increased, it consisted just
15.3% of the total population in 2016 and did not show a significant effect on the overall ratio
housing poverty rate.
When discussing the housing affordability issue in Hong Kong, the problem should be
regarded as diversified. Residents living in distinct types of housing demonstrate varying
situations. Low-income families living in private rental housing greatly suffered at least in the
past 10 years. However, those people who owned houses and finished their mortgage
benefited from the increase of housing price and became wealthy, especially in booming
private housing market (Yip, Forrest, & LaGrange, 2007). This situation distinguished the
people who faced ‘housing problem’ or made ‘housing profit’. The increase of housing price
and rent was not an issue to people living in PRH because the tenancy and rent were stable
there.
224
Other groups of people besides private rental housing residents may have suffered from
certain factors in the past years. Young adults who want to build new families and were
forced to live out of their original families probably experienced hardships due to the increase
of housing price and rent. For example, some residents were unable to stay with their original
families and needed to move out due to different reasons, such as getting married or divorced,
migrated from Mainland China, and having bad relationship with family members. The
housing price was high for these families to buy, and they were forced to live in the private
rental housing with high rent and small living area. In the recent years, even the middle class
and young working couples found it difficult to buy houses (Yip, 2013). Most couples tended
to find new flats after getting married, and they either chose to rent or buy. If they bought a
house, a large amount of down payment was needed, and they need to pay the monthly
mortgage in the next 25 or 30 years. If the couples chose to rent, they will suffer from the
increasing rent
5.3.4 Who are the groups that greatly suffer from housing poverty? Is it different from
income poverty analysis?
When housing poverty line was used in the analysis, it highlighted several groups
suffering from poverty, which is not indicated by the income poverty line alone.
In terms of household size, those people who were singleton and living in two-person
families generally suffered from housing poverty. In 2016, the residual housing poverty rates
225
of singletons and two-person families were 43.3% and 33.3%, respectively, which are much
higher than the income poverty rates (26.3% and 22.2%). Different housing poverty rates
showed a decreasing trend with the increase of household members. This situation occurs
because the households with few members exhibited difficultly in sharing the housing cost
with others regardless of the type of housing, they were living in.
With regard to family composition, those families with children exhibited a higher ratio
housing poverty rate compared with those without children. For instance, in 2016, the ratio
housing poverty rate of families with children and no elderly was 24.2%, which is much
higher than those with elderly and no children (10.3%). This finding showed the difference
with the measurement with income poverty line alone. The families with elderly generally
exhibited higher income poverty rate compared with those with children. The housing cost
spent by families with elderly and children may be overlooked when income poverty line
alone is used.
In terms of living area, the income and housing poverty rates of those living in tiny flats
(<20 m2) were relatively high. Nevertheless, the housing poverty measurement further
showed the poverty situation faced by the underprivileged. For example, the residual housing
poverty rate of those people living in tiny flats (<20 m2) was 54.5%, which is much higher
than those living in larger ones (20–<40 m2) (17.3%). The difference was large when the
income poverty measurement was used. Overall, the use of housing poverty measurement not
226
only highlighted the suffering of families by housing type but also indicated the poverty
situation faced by different groups, such as singleton, families with children and those living
in small living areas.
Previous studies in Hong Kong attempted to highlight the vulnerable groups that
suffered from housing problems. These groups include youngsters (Castro Campos et al.,
2016; Yip & Forrest, 2014), elderly (Hui, Wong, Chung, & Lau, 2014), private rental housing
residents (LaGrange & Pretorius, 2002) and homeowners with mortgage (Tam, Hui, &
Zheng, 2010). However, a comprehensive analysis of the housing poverty situation using
census data is unavailable. This study reveals the characteristics of those disadvantaged
groups suffering from housing poverty by using a large representative dataset in Hong Kong
5.3.5 Success and failure of public housing policy
The public housing policy exhibits a distinctive feature in the Hong Kong housing
system. The proportion of population living in public rental and public subsidised housing
was relatively high compared with those of other global cities worldwide (Chiu, 2010; Lau &
Murie, 2016). The PRH provided low-income families with a stable environment with low
rent. The low level in ratio housing poverty rate and housing poverty (30/40) indicators
showed that most PRH residents were able to afford their houses. For example, in 2016, the
ratio housing poverty rate was 5.1%, and the housing poverty (30/40) rate was 3.1%. The
ratio poverty rate of people living in public subsidised housing with mortgage also decreased
227
from 2001 to 2016, thereby showing the decline of housing burden faced by residents. The
trend in the past 20 years showed the success of public housing policy in helping certain
families eliminate housing poverty.
However, the proportion of population enjoying this benefit was declining, especially for
PRH. Although the Hong Kong government highlighted the building of additional PRH in the
Chief Executive’s annual policy address, the proportion of PRH residents was decreasing
(from 39.8% in 1996 to 32.0% in 2006 to 29.6% in 2016). This percentage of public
subsidised housing residents also declined from 17.3% in 2001 to 15.6% in 2016. Despite
building of new public housing units every year, public housing estates are also redeveloped
or removed. Besides the well-known situation of increasing waiting time of PRH, this result
demonstrated the failure of public housing policy in the past two decades and caused less
fraction of people living in the houses with low housing cost. This echoes the residualisation
and decline of public housing situation in Hong Kong, as suggested by Lau and Murie
(2016). However, some people questioned whether the resilience of public housing can be
sustained because the Hong Kong government actively promotes home ownership schemes
and have been selling PRH in recent years.
5.3.6 Explanation of housing poverty: political economy of housing in Hong Kong
The housing poverty situation cannot be simply explained by a traditional economic
demand and supply thesis. The political economy background of housing poverty is
228
complicated. The effect of globalisation on the housing market, the role of state and
corporates and other crucial factors, such as the role of housing authority, Heung Yee Kuk
and the culture of housing, also contributed in influencing the current housing poverty
situation in the past 20 years.
Effect of globalisation on housing market and residents
In the age of globalization, no countries or cities can neglect the interactions with other
regions in the world. Under advancement of technology, speeding up of transport,
development of internet, the world was tended to shrink (Munck, 2005), globalization nearly
affects everything in the society, including information and knowledge, income and
employment, prices of good and service (Milanovic, 2016; Ravallion, 2003; Wade, 2004).
Under rapid globalization, the financialisation of housing may create unstable society because
the capital flow cannot be efficiently controlled (Boyer, 2000a; Fernandez & Aalbers, 2016;
Smart & Lee, 2003b). For example, Hong Kong is not the only city that suffered from
housing price and housing deprivation. Other cities, such as London, New York, Vancouver,
and Melbourne, also face the same problem. The high housing price is a global problem,
especially for the cities with elevated level of capital flow. The sharp increase of rent and
housing price highly affected the livelihood of Hong Kong residents. The global inflow of
capital to the land and housing market was one of the main factors contributing to the rise in
housing price and rent (Yiu, Yu, & Jin, 2013).
229
The housing market and policies in Hong Kong were highly affected by the global capital
flow and global financial crises. For example, in 1997, the Hong Kong’s first CE Mr. Tung
proposed his vision of housing, such as ‘85,000 housing plan’ and 70% of the households will
own their flats. Later, Hong Kong faced Asian financial crisis, such as the increasing demand
from the tenant owners and property sectors to the government to ‘save the housing market’
in the late 1990s, especially with increasing property owners with ‘negative equity’. The
government finally introduced ‘the Nine Measures of Michael Suen’, including the banning
of the selling of land and HOS flats. This situation created shortage of land and private and
public housing in the later years. Public housing, such as housing with good availability,
affordability, living quality and accessibility, is important for people living in poverty
compared with those in the private market (Chiu, 2010). Public housing is also crucial for
anti-poverty (Hu & Chou, 2015; Hui, 1999, 2001) and to stabilise the society (Hu & Chou,
2015; Hui, 1999). The sharp decrease of public subsidised housings deeply affected the
livelihoods of the grassroots in Hong Kong.
Role of state and corporates in the change of housing policy
Housing problems remain a top concern of Hong Kong residents and top policy agenda
of the government (Forrest, 2003; Goodstadt, 2013). The claim that the government totally
ignored the housing issues, especially after 1997, is unreasonable. Active interventions were
initiated by the government, but whether they were for the interest of the citizens or
230
corporates was questioned. The effect of the corporate sector cannot be overlooked.
The power of business and property sectors was crucial in affecting housing policy
changes. The government/developer nexus was formed and strengthened after 1997, and the
term of ‘government–business collusion’ was widely used by Hong Kong media (Smart &
Lam, 2009; Smart & Lee, 2003a). The issues of ‘Hung Hom Peninsular public to private
housing scam’ (Smart & Lam, 2009) and the corruption case of the former Chief Secretary
for Administration Rafael HUI Si-Yan in CE Donald Tsang’s period were typical examples to
show the complicated interest between the government and the business sectors. Under the
governance of CE Donald Tsang, who had been criticised for forming the ‘government–
business collusion’, the supply of housing and land in those years was low, and the
government had not increased the supply in later years. The real estate sectors also had great
power in financing. In 1997, nearly half of the financial lending was accounted by the real
estate (Fung & Forrest, 2002). Moreover, nearly half of the housing properties were built by
six developers (Tse & Webb, 2000). However, Smart and Lam (2009) commented that the
civil society has initiated changes, especially in urban and housing planning. The Hong Kong
government learnt from the policy process and responded to the demands of the civic society.
The constraint from the civic society is still a crucial factor for policy learning and policy
change.
The third CE CY Leung highlighted that housing was the ‘top priority’ amongst the
231
government agenda, which includes policies for increasing the supply of land and public
housing, such as LTHS, planning of reclamation and re-implementation of HOS. The
government also introduced ‘spicy measures’ with different new types of stamp duty,
including Buyer’s Stamp Duty, Special Stamp Duty and Double Stamp Duty, to decrease the
demand of investors. However, the housing development plan also faced different pressure
from the society, including the environmental protection organisation, the residential
organisation and district councils. The ‘spicy measures’ was also opposed by the property
sectors.
Determination of the role of state and civil society in the changing of housing policy is
difficult. The dynamics are highly complicated and affected by different social, economic,
and political factors, including global economy change, ideology of the Chief Executive,
power of business sectors and housing demand from the society. The development of housing
policy is a continuous struggle amongst state, market, and society.
Other crucial factors for housing development
Besides the above-mentioned analysis of political economy, other crucial factors
influence the housing development in Hong Kong and are worthy of further study.
The first factor is the role of Housing Authority (HA). The HA is a statutory organisation
established in 1973 and is responsible for the development and management of the public
housing program in Hong Kong. Although the chairman and vice-chairman of HA are
232
government officers, many non-official members are from different parties, including
academies, representative of NGOs and legislative councillors. The vision of HA is clear, that
is, to provide affordable rental housing to low-income families and subsided home ownership
to middle-income families (HKHA, 2017). The role of HA is crucial for resisting the
neoliberal view of housing and showed resilience in providing public housing in the past
decades (Lau & Murie, 2016).
The second factor is the culture of housing. Two important housing cultures in Hong
Kong affect not only the housing policy but also the housing market. The first culture is the
belief of home ownerships. Substantial families prefer owning a house rather than renting.
The Chinese traditional culture of home ownership for retirement and sense of belonging may
be one of the reasons. Another reason is the increasing trend of housing price in the past
decades, making people believe that renting in not beneficial in the long term. The second
culture is speculative, that is, people treat housing as an investment tool rather than living. In
the 1990s before the Asian financial crisis, the market regulation was loose and promoted
speculations of housing (Chan, 2000).
The third factor is the housing need of the young generation. The youth housing
problems became increasingly serious in the past years. The lack of housing supply and
moving up of housing price and rent are part of the reason. The younger generation has been
asking for optimal living condition and home ownership. Some scholars argued that the
233
vigorous social movements in the past years are related to the housing need and mobility
problems of the youth (Forrest & Xian, 2018). The government tried to react to the youth
voice, for example, the implementation of youth hostel. However, the problems are yet to be
solved.
The fourth factor is the development in the New Territories grand land development
projects are conducted in the recent years. For example, the NTNE project and Hung Shui
Kiu development project in NTNW aimed to provide additional land for housing
development. However, such projects involve highly complicated political dynamics,
including the welfare of residents, the interest of landlords and the business in the brown-
field area. Some scholars argued that the government should develop other areas, such as
Golf Course in Fanling or wasted brown-field areas, instead of NTNE. Other scholars argued
that the development projects are urgent to satisfy housing needs. However, the benefit to
landlord or businessmen should be controlled and limited.
The fifth factor is the dynamics with Heung Yee Kuk, rural area, and NT indigenous,
which is related to the fourth point. The small houses and village houses in the rural area
occupied a large portion of the land in Hong Kong compared with public and private housing.
The leaders of NT indigenous claim that their rights of small houses were protected by the
basic law, but such rights were challenged by a court case in 2014. By contrast, the
development of land in the rural area by the government faced challenges from the rural
234
force.
These structural elements cause critical effects on the setting of housing policy,
regulation of housing market and distribution of land and housing and hence influence the
housing poverty situation. Specifically, social agencies in different parties, such as residents,
legislative councillors, media, and opinion leaders, may also reveal their influence on
changing the housing poverty situation.
235
Chapter 6 Quantitative Analysis: Part 2
6.1 The impact of housing factors on non-income poverty
The Strategic Public Policy Research (‘SPPR’) dataset was collected by Professor Wong
Hung and his team in their research project titled ‘Trends and Implications of Poverty and
Social Disadvantages in Hong Kong: A Multidisciplinary and Longitudinal Study’ which
aimed to investigate the trend of poverty, deprivation and health outcomes. The surveys were
administered on a random sample of Hong Kong households. The samples are selected from
25,000 addresses from the Census and Statistics Department by using a two-stage stratified
method, by living quarters and by living district. Face-to-face interviews were conducted by
trained interviewers from June 2014 to August 2015. Members of the household aging 18
years old or above were interviewed. This sample was designed to represent all the adults
aged 18 years old or above in Hong Kong. From the 4,947 addresses obtained, 3,791 valid
cases were available. Amongst these cases, 2,282 adults were successfully interviewed with a
response rate of 60.2%. A total of 301 respondents were randomly selected to determine if
they consider the selected items as necessity. Other respondents were invited to answer the
whole questionnaire. Finally, 1,978 valid questionnaires were obtained for analysis in this
study. This survey provided a representative sample of Hong Kong adults with a large variety
of variables related to housing and poverty. The use of this secondary dataset helps in
236
answering the research questions.
6.1.2 Corresponding research questions and hypothesis
This part of the study attempts to use the ‘SPPR’ data to answer the following research
questions:
◼ RQ2.1: To what extent do housing factors affect the different dimensions of poverty?
◼ RQ2.2: Which type of non-income poverty is the most affected by housing factors?
The corresponding hypotheses are:
◼ Hypothesis 2.1: Poor housing situations are associated with high level of poverty.
◼ Hypothesis 2.2: The overall impact of housing factors is different on each type of non-
income poverty situation.
6.1.3 Operationalisation of concepts and measurements
Demographic and socioeconomic status. The demographic background information of
the respondents in this study included sex, age, educational level, and marital and
occupational statuses. Age was divided into three levels, namely, ‘young adult (18–40 years
old)’, ‘adult (41–59 years old)’ and ‘elder (60 years old or above)’. The three levels of
education are ‘primary’, ‘secondary’ and ‘tertiary or above’. Occupation was divided into
‘full-time work’, ‘part-time work’ and ‘not working/economically inactive’. Marital status
included ‘married/cohabit’ and ‘single/separated/divorced/widowed’.
The equivalised household income measurement was used for a good comparison of the
237
economic situation of families with different numbers of household members. This measure
was calculated by dividing household income by the square root of the number of people
living in the same household. Data on demographic and socioeconomic status were used as
control variables in the regression analysis.
Non-income poverty measurement. Mainly three indicators, including deprivation,
social exclusion, and subjective poverty, were developed for measuring the non-income
poverty situation of respondents. The threshold of deprived, socially excluded and
subjectively poor was set with reference of the rate of relative income poor.
Deprivation. A total of 301 respondents were randomly selected to confirm the items for
measuring deprivation, while the remaining 1978 respondents were asked if they had the
proposed 21 necessity items, such as ‘diet and clothing’, ‘medical care’ and ‘household
facilities’, in their daily lives. The respondents who lack a certain number of items were
counted as ‘living in deprivation’ (for example, lacking two or more items out of 21 in
‘SPPR’). The items were summed up as a Deprivation Index (DI), scored from 0 to 21, and
were with 0.846 of Cronbach’s α, thereby showing high reliability for measurement. Cases
with DI >=2 were classified as deprived in this study.
Social exclusion. In terms of social exclusion, the respondents were asked nine
questions about their social networks and sources of support. The responses were used to
construct the Social Exclusion Index (SEI). Specifically, these respondents were asked about
238
how frequent they meet and communicate with their friends, how much support they receive
from their families and how often they feel respected and understood. Respondents who
answered, ‘once a month’ or less for the question ‘How often do you meet your friends or
family?’ were given an SEI score of one. The SEI score for each respondent ranged from 0 to
9, whilst its Cronbach’s α was 0.730, thereby implying its good measurement reliability.
Respondents who score four or more were counted as socially excluded. Cases with SEI>=4
were classified as socially excluded in this study.
Subjective poverty. The questions asking the subjective perception of the standard of
living were used to construct the Subjective Poverty Index (SPI). The participants were first
asked a monthly income amount which ‘necessary to keep a household, like yours, out of
poverty’. They then answered the question ‘How far above or below that level would you say
your household is?’ Those who answered ‘a little below’ or ‘a lot below’ score 1 or 2,
respectively. Other questions included ‘Do you think you are poor now?’ and ‘How would
you rate your standard of living?’ These questions contributed eight scores of SPI in total.
The Cronbach’s α of SPI was 0.654, thereby indicating moderate reliability for measuring
subjective poverty. Respondents who scored three or more were counted as subjectively poor.
Cases with SPI>=3 were classified as deprived in this study.
Housing factors. Various dimensions of housing factors were used for analysis. These
factors include housing type, living location, housing cost, housing affordability, living area
239
and situation inside a living place and community.
Housing type. In this study, the four main types of housing were ‘public rental housing’,
‘public subsidised housing’, ‘private rental housing’ and ‘privately owned housing’. Data
about mortgage payment were limited. Hence, no further division of housing types with
mortgage was performed.
Housing location. The living location of residents was divided into 18 districts on the
basis of the district council division of Hong Kong. The housing locations were grouped by
their distance to the Central Business District (CBD) to assess their differences and analyse
their relationship with poverty. The three groups include ‘near CBD’, ‘mid-range from CBD’
and ‘far from CBD’ (Appendix A).
Housing cost. The residents either rented or owned their living place. Housing cost
means the monthly rent paid by the renters and mortgage payment for the owners. This
expense also included electricity and water fees, management fees and government rates. The
housing cost per capita was obtained by dividing the housing cost by the number of family
members. The housing cost per capita was divided into two groups, low or high, taking the
median as a threshold.
Living area. In this study, the respondents were asked ‘How large is your living space?’
with 10 answer options from ‘<20 m2’ to ‘100 m2 or above’. The absolute living area was
estimated at the midpoint of the response. The lower limit was marked as 15 m2, and the
240
upper one was 105 m2. The living area per capita was obtained by dividing the housing cost
by the number of family members. The living area per capita was divided into two groups,
taking 10 m2 as a threshold. The use of per capita in calculation of housing indicators was
referred to the official measurement of Census and Statistical Department of Hong Kong.
Indoor housing problem index (IHPI). The respondents were asked nine questions
about the problems they faced in their living quarters, including lack of privacy, damping
wall and ceiling, poor ventilation and with rats and insects. The answers for these questions
were summed up to construct the IHPI scored from 0 to 9. The Cronbach’s α was 0.743,
thereby implying relative high reliability in measurement. The IHDI was divided into two
groups, low or high. Those who score 2 or more was classified as facing high level of
housing problems.
Community facilities problem index (CFPI). The respondents were asked seven
questions about whether they used the facilities, such as public sport, wet market, banks, and
private clinics, in their living community. If no or inadequate facilities were available or were
not affordable for the respondents, then such situation was counted as 1 score for the
community facilities problem index (CFPI). The score ranged from zero to seven with
Cronbach’s α of 0.743, thereby showing relative high reliability. The CFPI was divided into
two groups, low or high. Those score 3 or more in CFPI was classified as facing high level of
community facilities problems.
241
6.1.4 Statistical analytical strategy
Weighting was first performed on the basis of the distribution of sex, age, and housing
type in the census data of mid-2014 to improve the representativeness of the sample. After
weighting, the cases with missing data were eliminated. In the end, 1962 respondents
remained in the analysis. The descriptive data, including demographic information, economic
situation, poverty situation and housing characteristics, were first presented. Different logistic
regression models were applied with three dependent variables (DV), deprivation, social
exclusion, and subjective poverty. The demographic background information variables were
used as control variables in the model. The variables related to housing were used as main
independent variables (IVs) to measure the explanatory power of housing factors on the DVs.
Multiple linear regression models were further applied with those three DVs, with three
levels of IVs, demographic, income and housing factors. This aims at comparing the level of
impact of housing factors on different types of non-income poverty.
6.1.5 Descriptive Result
Demographic background
The results below were obtained after the weighting process. Males and females
comprised 45% and 55% of the sample, respectively. With regard to the age group, 39.4%
was young adult (18–40 years old), 35.1% was adult (41–59 years old) and 25.5% was elder
(60 years old or above). In terms of education, the percentage of respondents with highest
242
attainment in primary, secondary, and tertiary or above were 23.8%, 54.1% and 22.1%,
respectively. With regard to occupation, 43.6% of the respondents have full-time work,
whereas 48.6% of them were not working or economically inactive (Table 6-1).
Table 6-1 Demographic background of respondents
Demographic background
Weighted % N
Sex (N=1962) Male 45.0 883
Female 55.0 1078
Age (N=1962) 18-40 39.4 773
41-59 35.1 689
>=60 25.5 500
Educational primary 23.8 464
Attainment secondary 54.1 1054
(N=1949) tertiary or above 22.1 432
Occupation Full-time work 43.6 854
(N=1961) Part-time work 7.9 154
not working/economic inactive 48.6 953
Marital status Married/cohabit 64.5 1264
(N=1960) Single/separated/divorces/widowed 35.5 696
Economic characteristics
The indices EQI, DI, SEI and SPI were used to measure the income and non-income
poverty situation of the respondents. The mean value of these indices was HKD15,712 (SD =
10924), 0.688 (SD = 1.824), 1.716 (SD = 1.894) and 1.025 (SD = 1.439), respectively. About
14.6%, 17.4% and 14.9% of the respondents were classified as deprived, socially excluded.
and subjectively poor, respectively (Table 6-2).
243
Table 6-2 Income and non-income poverty indices and poverty rate
Mean SD
Equivalized household income (N=1845) HKD HKD
15,712 10,924
Deprivation Index (N=1960) 0.688 1.824
Social exclusion Index (N=1942) 1.716 1.894
Subjective poverty Index (N=1962) 1.025 1.439
Poverty rate
Income poverty (N=1845) 13.1%
Deprived (>=2 of 21) (N=1960) 14.6%
Socially excluded (>=4 of 9) (N=1942) 17.4%
Subjectively poor (>=3 of 7) (N=1962) 14.9%
Housing characteristics
The cases included respondents living in four types of housing, namely, public rental
housing (31.6%), public subsidised housing (17.1%), private rental housing (14.8%) and
privately owned housing (36.6%). In terms of living location, 28.7% of the respondents lived
near CBD, whereas 24.8% lived mid-range from CBD and 46.5% lived far from CBD (Table
6-3).
The mean values of housing cost and house cost per capita were $3119.92 (SD =
3930.56) and $1165.57 (SD = 1623.29), respectively. The mean value of living area was
48.33 m2 (SD = 22.70). By contrast, the mean value of living area per capita was 16.37 m2
(SD = 9.69). In addition, the mean values of IHPI and CFPI were 1.47 and 1.82, respectively
(Table 6-3).
244
Table 6-3 housing characteristics of respondents
Housing characteristics
Weighted % N
Housing type Public rental housing 31.6 619
(N=1962) Public subsidized housing 17.1 335
Private rental housing 14.8 290
Private owned housing 36.6 717
Housing location Near CBD 28.7 564
(N=1962) Mid-range from CBD 24.8 486
Far from CBD 46.5 912
Mean SD
Housing cost (N=1962) (in HKD) 3119.92 3930.56
Housing cost per capita (N=1962) (in HKD) 1165.57 1623.29
Living area (N=1960) (in m2) 48.33 22.70
Living area per capita (N=1960) (in m2) 16.37 9.69
Indoor housing problem index (N=1962) 1.47 1.86
Community facility problem index (N=1962) 1.82 1.97
6.1.6 Inferential Analysis
In this part, the impacts of housing factors on different non-income poverty situations
were examined by various logistic and multiple linear regression models. Models with similar
IVs were constructed to explore their impact on different DVs, including deprivation, social
exclusion, and subjective poverty.
a) Impact of housing on deprivation
Logistic Regression Model 1: Probability of deprived
Logistic regression analysis is performed to examine the impact of various independent
variables on deprivation. For demographic variables, only sex was not significant with
deprivation, while age, education level, occupation and marital status showed significant
245
association with deprivation. For income level, those who are income poor are near five times
more likely than non-income poor to be deprived (adjusted OR 4.90, 95% CI [3.40 – 7.07]).
In terms of housing characteristics, housing cost and living location did not significantly
associated with deprivation. For other housing factors, those with high level of community
facilities problems (adjusted OR 1.61, 95% CI [1.17 – 2.20]) and indoor housing problems
(adjusted OR 2.04, 95% CI [1.48 – 2.81]) had higher risk to be deprived. On the other hand,
taking living area per head >= 10 m2 as reference group, those lived in living area per head <
10 m2 showed significant association with and had higher risk of being deprived (adjusted OR
2.94, 95% CI [2.09 – 4.13]) (Table 6-4)
Table 6-4 Logistic Regression Model 1: (DV: Deprivation)
Variables Adjusted OR 95% CI
Demographic and SES

Gender
Male 1
Female 1.24[0.90-1.70]
Age
18 to 40 1
41 to 59 1.79*[1.19-2.69]
>=60 1.55[0.95-2.52]
Education
primary 1
secondary 0.83[0.57-1.22]
tertiary or above 0.35**[0.18-0.67]
246
Occupation
Full-time work 1
Part-time work 2.50**[1.47-4.27]
Not working/economic inactive 2.01**[1.35-2.99]
Marital Status
Married/cohabit 1
Single/separated/divorces/widowed 1.67**[1.20-2.31]
Income poverty
NOT Relatively poor 1
Relatively poor 4.90***[3.40-7.07]
Housing Characteristics
Housing cost
Low 1
High 1.28[0.95-1.74]
Location
Near CBD 1
mid-range from CBD 1.10[0.74-1.65]
far from CBD 0.95[0.67-1.35]
Community Facilities
Low level 1
High level 1.61**[1.17-2.20]
Living Density
living area per head >= 10 m2 1
living area per head < 10 m2 2.94***[2.09-4.13]
Indoor housing problem
Low level 1
High level 2.04***[1.48-2.81]
Significant level, Odd ratio (OR) and 95% confidence interval (CI) for Deprivation.
*p < 0.05, **p < 0.01, ***p < 0.001
b) Impact of housing on social exclusion
Logistic Regression Model 2: Probability of socially excluded
variables on socially excluded. For demographic variables, sex and occupation were not
247
significant with socially excluded. Age, education level, and marital status showed significant
association with socially excluded. For income level, income poor were more likely than non-
income poor to be socially excluded (adjusted OR 2.53, 95% CI [1.79 – 3.59]).
In terms of housing characteristics, housing cost, living location and community
facilities problems did not significantly associated with socially excluded. For other housing
factors, those with high level of indoor housing problems had higher risk to be socially
excluded (adjusted OR 1.38, 95% CI [1.04 – 1.84]). In addition, taking living area per head
>= 10 m2 as reference group, those lived in living area per head < 10 m2 showed significant
association with and had higher risk of socially excluded (adjusted OR 1.43, 95% CI [1.05 –
1.94]) (Table 6-5).
Table 6-5 Logistic Regression Model 2: (DV: Social Exclusion)
Demographic and SES

Gender
Male 1
Female 1.13[0.86-1.49]
Age
18 to 40 1
41 to 59 1.49*[1.06-2.11]
>=60 1.13[0.73-1.75]
Education
primary 1
secondary 0.74[0.52-1.04]
tertiary or above 0.36***[0.21-0.61]
248
Occupation
Full-time work 1
Part-time work 1.34[0.83-2.16]
Not working/economic inactive 1.07[0.77-1.48]
Marital Status
Married/cohabit 1
Income poverty
Housing cost
Low 1
High 1.05[0.81-1.37]
Location
Near CBD 1
far from CBD 1.07[0.78-1.45]
Low level 1
High level 1.05[0.79-1.40]
Living Density
living area per head < 10 m2 1.43*[1.05-1.94]
No or moderate 1
Serious 1.38*[1.04-1.84]
Significant level, Odd ratio (OR) and 95% confidence interval (CI) for social exclusion.
*p < 0.05, **p < 0.01, ***p < 0.001
c) Impact of housing on subjective poverty
Logistic Regression Model 3: Probability of subjectively poor
variables on subjectively poor. For demographic variables, age and occupational status were
249
not significant with subjectively poor while sex, education level, and marital status showed
significant association with subjectively poor. For income level, income poor were more than
seven times likely than non-income poor to be subjectively poor (adjusted OR 7.16, 95% CI
[4.92 – 10.41]).
In terms of housing characteristics, living location was not significantly associated with
subjectively poor. For other housing factors, those with low level of housing cost (adjusted
OR 0.74, 95% CI [0.55 – 0.99]), high level of community facilities problems (adjusted OR
1.65, 95% CI [1.22 – 2.24]) and high level of indoor housing problems (adjusted OR 2.49,
95% CI [1.82 – 3.41]) had higher risk to be subjectively poor. On the other hand, taking
living area per head >= 10 m2 as reference group, those lived in living area per head < 10 m2
showed significant association with and had higher risk of subjectively poor (adjusted OR
1.87, 95% CI [1.35– 2.58]) (Table 6-6).
Table 6-6 Logistic Regression Model 3: (DV: Subjective Poverty)
Demographic and SES

Gender
Male 1
Female 0.59**[0.44-0.80]
Age
18 to 40 1
41 to 59 1.33[0.91-1.94]
>=60 0.91[0.57-1.47]
250
Education
primary 1
secondary 0.76[052-1.11]
tertiary or above 0.39**[0.22-0.70]
Occupation
Full-time work 1
Part-time work 1.66[1.00-2.75]
Not working/economic inactive 1.10[0.76-1.59]
Marital Status
Married/cohabit 1
Income poverty
Housing cost
Low 1
High 0.74*[0.55-0.99]
Location
Near CBD 1
far from CBD 1.00[0.71-1.40]
Low level 1
High level 1.65**[1.22-2.24]
Living Density
living area per head < 10 m2 1.87***[1.35-2.58]
Low level 1
High level 2.49***[1.82-3.41]
Significant level, Odd ratio (OR) and 95% confidence interval (CI) for subjective poverty.
*p < 0.05, **p < 0.01, ***p < 0.001
251
Impact on deprivation, social exclusion and subjective poverty
The above analysis showed that in logistic regression models with similar structures,
control variables and housing factors had different levels of influence on non-income poverty.
Obviously, income poverty was significantly associated with all non-income poverty. The
large magnitude of odd ratio in three models indicated that being income poor was a crucial
risk factor for deprivation, social exclusion and subjective poverty. Among housing factors,
living location was not significant associate with three types of non-income poverty. On the
other hand, high level of indoor housing problems and living in denser housing were risk
factors of deprived, socially excluded and subjectively poor. Housing cost and community
facilities problems showed different impact on various types of non-income poverty.
Multiple linear regression models were further applied for comparing the level of
impact of housing factors on the DVs. Different IVs were inputted to the model with logical
consideration. In regression analysis, the categorical variables were computed as dummy
variables. The continuous variables with relatively large value to DV, including income,
housing cost, and living area, were taken as natural logarithm for calculation. Demographic
variables were first inputted in level 1. The EQI, which was an influential factor of DV
theoretically, was then incorporated into level 2 as a control variable. Finally, the housing
factors were inputted into the next levels with consideration of the normal practice of
residents.
252
This analysis compared the R2 of the models (Table 6-7). The total R2 in the model of
subjective poverty (R2 = .301) was the largest. Hence, subjective poverty was better
explained by the IVs than the other factors, whereas social exclusion was less explained.
With regard to deprivation, the DV was most explained by housing factors (R2 = .080)
compared with the demographic background and EQI amongst the three levels of analysis.
The housing factors had the least effect (R2 = .028) amongst the three levels. Nevertheless,
the demographic factors showed a comparatively large effect (R2 = .080) on social exclusion.
In terms of subjective poverty, the effect of EQI was critical (R2 = .171) and demonstrated the
largest explanatory power. By contrast, the effects of demographic and housing factors were
similar.
The significance of standardised beta of IV amongst different regression models was
studied. The result showed the common factors demonstrating the effect on non-income
poverty. Marital status significantly affected three DVs amongst the demographic background
data. In level 2, the effect of EQI was not only significant but also with large magnitude of
standardised beta in all models. With regard to the housing factors, the housing cost per
capita and indoor housing problem were the common significant factors. The living area was
significant on deprivation and subjective poverty but not social exclusion.
253
Table 6-7 Comparison of R2 among different regression models
DV in regression model
Deprivation Social Exclusion Subjective Poverty
2
R of demographic .062 .080 .065
R2 of EQI .067 .064 .171
R2 of housing factors .080 .028 .066
Total R2 209 .172 .301
6.2 The mediating effect of housing factors on non-income poverty
The logistic and multivariate linear regression models used in the previous session studied
the direct relationship between IVs and DVs. The model examined the relationship amongst
income, housing factors and non-income poverty (Figure 6-1).
Figure 6-1 The independent effect of income and housing factors on non-income poverty
However, the relationship amongst variable can be complicated. For example, housing
factors can be the mediators between income and non-income poverty (Figure 6-2), or
interacting factors exist amongst housing variables. The indirect mediating effect of the
variable can be examined by path analysis and structural equation modelling (SEM). SEM
can analyze, dependent variables, independent variables and error terms within a theoretical
framework. Moreover, SEM can address the mediating and moderating effect of variables
with interconnections (Hoyle, 2012). The minimum sample size of 100 to 200 in total is
254
acceptable for SEM analysis (Kline, 2011; Loehlin, 1992). SEM was an appropriate method
to analyse the path relationship among housing factors and non-income poverty using the
SPPR dataset.
Figure 6-2 Housing factors as mediators among income and non-income poverty
The date sample used to answer the research question was the SPPR dataset, which was
a random sample designed to represent all the adults in Hong Kong. The background was
introduced in the previous session.
6.2.2 Corresponding research questions and hypothesis
The literature review asked the question of if or how the housing factors had mediating
effects amongst income and non-income poverty, including deprivation, social exclusion, and
subjective poverty. The research question and hypothesis were as follows:
RQ 2.3:
■ Are housing factors the mediators of income poverty and non-income poverty?
Hypothesis 2.3.
The following nine hypotheses listed in section 4.3.1 will be tested by SEM analysis:
6.2.3 Operationalisation of concepts and measurements
255
Several variables were constructed for measurement to examine the relationship amongst
income, housing factors and non-income poverty situation. Income, deprivation, social
exclusion, and subjective poverty were the main measurements for poverty. In the previous
session, housing cost, living area and indoor housing problems significantly affected non-
income poverty. These housing factors were used in the analysis. In SEM, continuous
variables were more preferable than the categorical or binary ones for measurement. The
SEM analysis allowed the formation of latent variables by different observed variables for
measuring certain concepts. In this study, social exclusion and subjective poverty were
reconstructed instead of summing them up as indexes as used in the last session. The
variables and constructs in SEM models were listed as follows.
Income. Similar to the previous session, the equivalised housing income (EQI) was used
to measure income. This framework was taken as a natural logarithm in the model.
Deprivation. In terms of deprivation, the deprivation index (DI) used in previous session
was kept for analysis. The index consisted of 21 binary items with well-validated
measurement. Hence, another latent variable must be constructed.
Social exclusion. The questions used in the SEI were used to construct a latent variable
for measuring social exclusion to advance the measurement instead of summing up. Those
factors with loading <0.3 were eliminated. After confirmative factorial analysis (CFA), six
observed variables were used for measurement.
256
Subjective poverty. The four questions used in the measurement of subjective poverty
index incorporated into CFA. The four observed variables were with a relative high loading
and kept for measurement.
Housing cost (per capita). The measurement of this variable was similar to that in the
previous session.
Living area (per capita): The measurement of this variable was then similar to that in
the previous session.
IHPI: The measurement of this index was similar to that in previous session.
Housing type: Group analysis was performed in the SEM analysis. The four housing
types, namely, public rental, public subsidised, private rental, and privately owned housing,
were used as groups.
Housing location. The categories used for this variable are similar to those in previous
session.
6.2.4 Statistical analytical strategy
The associations amongst variables were tested via SEM with assistance of software
AMOS. The coefficients of the impact of housing factors on non-income poverty were
compared with those of the impact of income on non-income poverty by using SEM models
to measure the effect of meditation. CFA was performed for the latent variables, social
exclusion, and subjective poverty, to eliminate factors with low weight loading. This task was
257
carried out to ensure that the constructs in the models were well-explained by those observed
variables.
SEM model 1 was first constructed by measuring the impact of deprivation using three
housing factors as mediating variables. SEM model 2, which has a similar structure to SEM
model 1, was then tested by grouping the housing types. The result of the private rental
housing group was significant and thus presented. SEM model 3 was constructed by using
social exclusion as the endogenous variable. SEM model 3 was tested by grouping the
housing types. The result was not shown because the difference was insignificant. Finally, the
endogenous variable was changed to subjective poverty and regarded as SEM model 4. This
variable was also analysed by using private rental housing as SEM model 5.
In each SEM model, the regression weight amongst variables, the direct and indirect effects
on the DVs and the goodness of fit of model were discussed. The chi-square test was
inappropriate to evaluate the model fit for the large sample size (Byrne, 2001). The model for
the goodness of fit indices is considered as a good fit for the RMSEA value < 0.08 (Hu &
Bentler, 1999). The IFI, TLI and CFPI were satisfied if they were higher than 0.90 (Bentler,
1990). The goodness of fit index (GFI) and adjusted GFI (AGFI) were considered acceptable
if the value was > 0.90 (Hooper, Coughlan & Mullen, 2008).
258
6.2.5 Descriptive Result
The descriptive results of the sample, including the demographic background,
economic situation, and housing characteristics, were illustrated in the previous session.
In particular, the descriptive results of the construct were listed in the Table 6-8. Those
factors with loading less than the minimum acceptable loading of 0.3 (Hair, Black, Babin, &
Anderson, 2014; Tabachnick & Fidell, 2019) were eliminated from the analysis. The factor
loading of the observed variables for social exclusion ranged from 0.33 to 0.89. Meanwhile,
the factor loadings for subjective poverty ranged from 0.53 to 0.75. The Cronbach’s α of
social exclusion and subjective poverty were 0.728 and 0.614 respectively, which showed
moderate reliability for measurement.
Table 6-8 Descriptive statistics of the construct in SEM models
Construct Item Mean SD Cronbach's Factor

alpha loading
Social SE1 5.15 1.32 0.728 0.37
Exclusion SE2 4.07 1.79 0.33
SE3 2.28 0.89 0.86
SE4 2.29 0.87 0.89
SE5 2.09 0.82 0.73
SE6 2.72 0.64 0.40
Subjective SP1 3.09 1.07 0.614 0.61
Poverty SP2 1.76 0.43 0.75
SP3 2.82 1.15 0.53
SP4 3.02 0.58 0.54
259
6.2.6 CFA
The correlations amongst the observed variables used in the creation of social
exclusion and subjective poverty were listed in the following table. All variables had
significant correlations (Table 6-9 and Table 6-10).
Table 6-9 Correlation among observed variables for construct of social exclusion
SE1 SE2 SE3 SE4 SE5 SE6

SE1 1.000
SE2 .437 1.000
SE3 .280 .261 1.000
SE4 .299 .267 .784 1.000
SE5 .351 .239 .618 .644 1.000
SE6 .266 .241 .317 .336 .373 1.000
Table 6-10 Correlation among observed variables for construct of subjective poverty.
SP1 SP2 SP3 SP4

SP1 1.000
SP2 0.453 1.000
SP3 0.283 0.418 1.000
SP4 0.365 0.384 0.291 1.000
After modification, the comparative fit index (CFI), RMSEA and Tucker–Lewis fit index
(TLI) for the CFA of social exclusion were .991, .062 and .973. These indicators showed a
relatively good fit between the observed data and the model. All factor loadings were
significant (p < .001) in the model.
In the CFA model of subjective poverty, the construct link to four observed variables and
all the factor loadings were significant (p < .001). Such link also showed a good model of fit,
with CFI = .998, RMSEA = .033 and TLI = .990. With the CFA, the constructs social
260
exclusion and subjective poverty were used for further analysis in the following SEM models.
6.2.7 Path Analysis and structural equation modelling (SEM)
a) SEM model 1 (deprivation and path analysis)
In this model, no measurement model was used because the DVs deprivation was an
observed variable. This model was path analytical, and the regression weight and path effect
were examined (Figure 6-3). About 14% of the variance of deprivation was explained in the
model. All the paths on deprivation were significant. The effect of EQI on deprivation
showed a large effect (r = −0.31, p<0.01). The regression weights on deprivation of indoor
housing problem, living area per capita and housing cost per capita were 0.15, −0.10 and
−0.06, respectively. Interaction effects can be observed amongst housing factors. The
regression weights of the paths from housing cost to living area and from living area to IHPI
were 0.17 and −0.27, respectively (Table 6-11).
261
Figure 6-3 SEM model 1: impact on deprivation
Note. *p<0.05, **p<0.01, ***p<0.001.
Table 6-11 Regression Weights of SEM model 1
Standardize S.E. C.R. P

d Estimate value
Housing cost per <--- EQI .037 .036 1.629 .103
capita
Living area per <--- EQI .054 .018 2.434 .015
capita
Living area per <--- Housing cost .165 .011 7.447 ***
capita per capita
IHPI <--- Living area per -.267 .073 -12.153 ***
capita
IHPI <--- EQI -.012 .060 -.573 .566
IHPI <--- Housing cost -.008 .038 -.384 .701
per capita
Deprivation <--- Housing cost .062 .041 2.936 .003
per capita
Deprivation <--- Living area per -.096 .083 -4.390 ***
262
capita
Deprivation <--- IHPI .148 .025 6.841 ***
Deprivation <--- EQI -.308 .066 -14.744 ***
The standardised effect of variables was studied, and the EQI showed a large effect on
deprivation. Living area (−.136) and IHPI (.148) also showed relatively high effects on the
DVs amongst the housing factors. The housing cost had a relatively indirect effect on IHPI
(−.044) and deprivation (−.024). Living area also had an indirect effect on deprivation
(−.040).
b) SEM model 2 (deprivation, path analysis and private rental housing)
The analysis was split by housing types for further investigation because the housing
characteristics were varied in different housing types. The effect of housing factors
demonstrated more significant results in the group of private rental housing residents amongst
the four housing types (Figure 6-4).
263
Figure 6-4 SEM model 2: impact on deprivation, private rental housing residents
In this model, approximately 21% variance of deprivation was explained, which was
much higher than that of the model of overall population. The regression weights from EQI to
deprivation, housing cost and living area were significant. The weights on the paths from EQI
to housing cost (r = .36, p < 0.001) and living area (r = .20, p < 0.01) were relatively large.
Income also significantly affected housing characteristics of those living in private rental
housing. The paths from housing cost to living area (r = .34, p < 0.001) and from living area
to IHPI (r = −.47, p < 0.001) were also with high regression weighting (Table 6-12).
264

d Estimate value
Housing cost per <--- EQI .362 .093 5.437 ***
capita
capita
capita per capita
capita
IHPI <--- EQI -.114 .259 -1.712 .087
per capita
Deprivation <--- Housing cost .054 .224 .752 .452
per capita
Deprivation <--- Living area per -.131 .322 -1.651 .099
capita
Deprivation <--- IHPI .176 .083 2.352 .019
Deprivation <--- EQI -.317 .303 -4.510 ***
In this model, EQI had a relatively high total effect on deprivation and housing
variables. The total effect of living area (−.213) and IHPI (.176) on deprivation was also large
in magnitude. The indirect effect of EQI to deprivation was significant (−.071), whereas those
of housing cost (−.076) and living area (−.082) were even larger. Overall, this model
demonstrated the mediating effect of housing factors. The effect of income on deprivation
was mediated by housing cost, living area and indoor housing problem.
c) SEM model 3 (social exclusion and SEM)
In this model, social exclusion was a dependent latent variable, and about 12% of the
variance was explained by other exogenous variables. The regression weight of the paths was
265
analysed, and the path from EQI to social exclusion had a large weight (r = .31, p < .05). The
direct effects of housing cost (r = .08, p < .05) and IHPI (r = .14, p < .05) on social exclusion
were also significant. The path from housing cost to living area and that from living area to
IHPI were significant amongst the housing factors, and their regression weights were 0.17
and 0.27, respectively (Figure 6-5 and Table 6-13).
Figure 6-5 SEM Model 3: impact on social exclusion

d Estimate value
capita
capita
capita per capita
266
per capita
IHPI <--- EQI -.012 .060 -.573 .566
capita
Social Exclusion <--- EQI -.307 .022 -10.117 ***
Social Exclusion <--- Housing cost .081 .010 3.448 ***
per capita
Social Exclusion <--- Living area per .027 .020 1.156 .248
capita
Social Exclusion <--- IHPI .136 .006 5.451 ***
This study examined the effect on social exclusions. Amongst the housing factors, IHPI
had the largest effect (.136), and housing cost also obtained significant effect on SE (.078).
The indirect effect from living area on SE was −0.36, which was relatively large compared
with those of other variables. Overall, the effect on social exclusion in this model was mainly
contributed by income. Although certain housing factors directly affected social exclusion,
the mediating effect of housing factors amongst income and social exclusion was
insignificant.
This model demonstrated relatively good model of fit (X/df = 2.670, p <.001;
RMSEA = 0.021, RMR = 0.060, CFI = 0.979, AGFI = 0.964, GFI = 0.983). As there was
relatively large sample size, the checking of chi square and p value can be overlooked.
Besides this, most of the indexes are with good standard for adaption.
In the group analysis of housing type, no significant difference was observed amongst
the groups in terms of factor loadings and directions of effect. Thus, no illustration of the
result was provided.
267
d) SEM model 4 (subjective poverty and SEM)
In this model, all the paths to the observed latent variable in the measurement model
were significant with factor loadings from 0.54 to 0.71. About 43% of the variance of
subjective poverty was explained by the other exogenous variables. The paths of EQI, living
area and IHPI on subjective poverty were significant. EQI significantly affected SP in the
model (r = .59, p < −.05), and the regression weights of living area and IHPI on SP were 0.08
and 0.24, respectively. The paths from housing cost to living area and from living area to
IHPI were significant with regression weights of −0.17 and −0.27 amongst the housing
variables (Figure 6-6 and Table 6-14). An interaction effect can be observed amongst housing
variables.
Figure 6-6 SEM model 4: impact on subjective poverty
268

d Estimate value
capita
capita
capita per capita
capita
IHPI <--- EQI -.012 .060 -.573 .566
per capita
Subjective poverty <--- EQI -.586 .031 -23.222 ***
Subjective poverty <--- Housing cost .033 .016 1.549 .121
per capita
Subjective poverty <--- IHPI .241 .010 10.553 ***
Subjective poverty <--- Living area per -.079 .033 -3.493 ***
capita
The standardised total effect was the largest amongst EQI and subjective poverty
(−.596). Amongst the housing factors, housing cost per capita significantly affected living
area per capita (0.165), which also had relatively high influence on IHPI (−.267). The IHPI
significantly affected SP (.241) besides EQI. In this model, the indirect effect of EQI on
subjective poverty was relatively small (−.010). However, the indirect effects of housing cost
per capita (−.026) and living area per capita (−.064) were relatively large compared with
those of other variables.
The present study tested the goodness of fit. This model demonstrated a relatively good
269
fit (X/df = 3.035, p < .001; RMSEA = 0.023, RMR = 0.032, CFI = 0.974, AGFI = 0.966, GFI
= 0.988). The checking of chi square and p value can be overlooked because of the relatively
large sample size. All the indexes are with good standard for adaption.
e) SEM model 5 (subjective poverty, SEM, and private rental housing)
Similar to the previous model, group analysis by housing type was performed.
Significant results were also found in the group of respondents living in private rental
housing. About 52% of the variance of subjective poverty was explained by the other
variables. The regression weight of EQI (r = −.54, p < .05) and IHPI (r = .36, p < .05) on
subjective poverty was significant with a relatively high factor loading. The high regression
loadings were located in the path from EQI to housing cost (r = .36, p < .05), from housing
cost to living area (r = .34, p < .05) and from living area to IHPI (r = −.47, p < .05), which
were similar to SEM model 4 but with larger magnitude (Figure 6-7 and Table 6-15). This
finding showed the path of the effect from EQI to SPI through housing factors.
270
Figure 6-7 SEM model 5: impact on subjective poverty, private rental housing residents
Standardized S.E. C.R. P

Estimate value
Housing cost per <--- EQI .362 .093 5.437 ***
capita
capita
capita per capita
capita
IHPI <--- EQI -.114 .259 -1.712 .087
per capita
271
Subjective poverty <--- EQI -.543 .090 -7.027 ***
Subjective poverty <--- Housing cost .031 .059 .442 .659
per capita
Subjective poverty <--- IHPI .364 .023 4.758 ***
Subjective poverty <--- Living area per .003 .085 .042 .967
capita
Although the direct effects from housing cost and living area on SP were small, they
significantly and indirectly affected SP with loadings of −0.65 and −0.170, respectively. By
contrast, the indirect effect of EQI on SP was −0.088, which was much higher than that
(0.010) in the overall population analysis. This finding not only demonstrated that the effect
of EQI on SP was mediated by the housing factors, but the effect was particularly significant
in the group of private housing residents.
272
6.3 Findings and Discussion
6.3.1 Impact of housing factors on deprivation
In terms of the influence on deprivation, income level significantly affected all variables
in the regression and SEM models. This result is consistent with previous studies supporting
that household income exerted a strong influence on the deprivation level of families
(Saunders, Wong, & Wong, 2014). This result reveals that the impact of income on
deprivation is more significant than those of other individual demographic and housing
factors. However, the total R2 of housing factors (.080) is larger that of income (.067). Hence,
the total effect of housing factors explains the larger variance of deprivation than that of
income. Previous studies normally focus on particular housing variables to investigate the
effect of housing (Daniel et al., 2018; Warren, 2018). This result enriches the existing
literature in examining the effect of different housing factors in the same model. After the
demographic variables, including sex, age, and educational level, were controlled,
community facilities problem, living area per capita and indoor housing problems showed
significant effects on deprivation amongst all housing variables.
Living area per capita showed the largest odd ratio amongst housing factors, followed by
the indoor housing problems. Households living in small flats with indoor environmental
problems are more likely suffered from deprivation. Such notion echoes the previous studies
that emphasise the crucial effect of housing condition on deprivation (Guio & Maquet, 2007;
273
Napiorkowska-Baryla & Witkowska-Dabrowska, 2018). This finding is particularly
important in Hong Kong. Residents face many issues with small living area and indoor
environmental problems because of the increasing number of subdivided flats and small size
private housing units, thereby leading to deprivation. Though the income level of Hong Kong
working population generally increased in the previous years, the deprivation situation of
residents may not improve because the housing condition is worsening.
The influence of housing cost on deprivation was not significant. This finding shows a
different result with previous research, which emphasises the effect of housing cost on
deprivation (Daniel et al., 2018; Warren, 2018). High housing cost does not directly imply
high level of deprivation. One possible explanation is that households attempted to squeeze
the cost of renting and scarified the living quality. For example, certain households may
limited by their budget and have to rent a smaller flat. Thus, part of the effects of housing
cost on deprivation was reflected by the other housing factors. The SEM analysis probably
explains this. In SEM model 1, the direct effect of housing cost on deprivation is small.
However, the indirect effect of housing cost through other housing factors on deprivation is
relatively large. This situation may explain why housing cost in the regression model does not
reveal a large effect. Nevertheless, the interaction of housing factors needs to be examined in
the SEM and qualitative analyses.
Housing location did not show a significant effect on deprivation. This finding shows a
274
different result with a previous study, which found that deprivation varied with distance from
the city centre (Markkanen & Harrison, 2013). One plausible reason is the small difference of
deprivation situation between various districts in Hong Kong. Another probable reason is that
the survey of this research cannot reach those relatively rich people living in wealthy areas.
This situation may limit the variation of effect of housing location on deprivation.
In terms of housing type, the SEM model analysis demonstrates that the housing factors
generally show a significantly greater effect in the private rental housing model than those of
the general population. For example, the total effect of living area on deprivation is −.213 in
the private rental housing group, which is larger than that in the general population (−.136).
This finding echoes the research of Borg (2015) wherein the extension of rental housing
market caused a worse deprivation situation. The result is also consistent with the previous
studies supporting that private rental tenants greatly suffered from deprivation amongst
different housing tenures (Fafard St-Germain & Tarasuk, 2020).
The SEM analysis confirms that housing factors demonstrate a mediating effect between
income and deprivation. This finding responds to a previous study that questions if housing
can mediate the poverty situation and deprivation (Lee, 1994). Although income level still
shows the largest influence on deprivation in the model, the total effect of all three housing
factors is even larger. This finding reveals a different conclusion with previous research
showing that the effect of housing factors is significant but weak (Rodero-Cosano, Garcia-
275
Alonso, & Salinas-Pérez, 2014). The SEM analysis further illustrates the interaction amongst
housing variables. For example, the housing cost shows a direct effect on living area and
further influence on the indoor housing environment. These housing factors show direct
effects on deprivation. This finding is new because the existing literature normally focuses on
the independent impact of housing factors but ignores the interaction amongst them. Such
result fills this gap in confirmation of the interaction effect amongst housing factors.
Overall, this study broadens the understanding of deprivation in Hong Kong in previous
studies, which focus on the measurement and prevalence of deprivation situation (Cheung,
Chan, & Chou, 2019; Saunders et al., 2014a). This research instead focuses on the causes of
deprivation and the impact of housing factors besides income.
6.3.2 Impact of housing on social exclusion
For social exclusion, number of demographic factors, including age, education, marital
status, showed significant relationship with social exclusion. Among housing factors, only
living density and indoor housing problem demonstrated significant effect. Compared with
deprivation and subjective poverty, the odd ratios of significant housing factors were the
smallest in the model of social exclusion. This implied the impact of housing factors was
relatively weak.
In the linear regression model, the demographic variables in total show a relatively
large effect (R2 = .080). The demographic variables totally exerted greater effect on social
276
exclusion than those of income (R2 = .064) and housing factors (R2 = .028). This notion is
different with the analysis of effect on deprivation. Such notion implies that the influence of
housing variables on social exclusion was the least important. The demographic factors, such
as educational level and age, showed great association with social exclusion. Nevertheless,
income again demonstrates a large effect on social exclusion as an individual factor.
In the logistic regression models, indoor housing problems and living space revealed
significant impact on social exclusion. This finding responds to the study of Somerville
(1998) that residential living space significantly influences the social exclusion of residents.
However, the odd ratio of living space in this study is not large compared with other models.
This study shows that the physical housing condition does not cause a strong impact on social
exclusion. Specifically, the social network or support of residents is not seriously affected by
their living environment. The reasons behind this phenomenon need to be explored via the
qualitative interview.
With regard to the housing location, a number of previous studies support that
geographic factor is influential to social exclusion because the housing location normally
affects the access to social service and workplace (Malpass & Murie, 1999). However, the
impact of housing location on social exclusion is not significant in this study. This result is
different from that of pervious study (García-Vélez et al., 2020; Munch, 2012). One possible
explanation is as follows: Hong Kong is a society with dense population. The geographic
277
distance does not cause a substantial impact on the social interaction of residents. Certain
districts have a relatively large population, and transformation from one district to another is
easy. The housing location is not a significant factor for building a social relationship.
Community facilities problem did not show significant effects on social exclusion.
Theoretically, community resource is highly correlated with social support and relationship
(Park et al., 2019; Sun et al., 2018). However, the situation in Hong Kong may be different
with those of foreign countries. The social interaction of residents in Hong Kong may not
depend on the availability of neighbourhood facilities because the population is relatively
dense. Instead, people may establish their social life through workplaces and schools. For
example, a woman may establish her social network through the school activities of her
children. The result of the qualitative interview will reveal the social network of residents in
detail. Nevertheless, this notion does not imply community facilities problem are not
important, but their effect is not determining.
With regard to housing tenure, this study attempts to split the analysis of social
exclusion by housing type on regression and SEM models. However, no significance is found
amongst different housing tenures. This result is different from those of previous studies. For
example, Marsh and Mullins (1998) argued that social housing concentrates on poor
households, and they are further excluded in the society. However, residents living in public
housing in Hong Kong may not face this situation. Although public housing is designed for
278
low-income families, diverse types of housing exist in the same district or local community.
The mixture of housing tenure in one community may reduce the effect of housing tenure on
social exclusion.
Overall, the impact of some housing factors on social exclusion is significant but its
magnitude is relatively small in Hong Kong. This result counters the findings in previous
studies (Arthurson & Jacobs, 2003; Marsh & Mullins, 1998; Somerville, 1998). The
difference in urban planning and housing allocation between Hong Kong and other countries
is one of the explanations. The other explanation may line with the social interaction and
lifestyle of Hong Kong people. People may establish social support, and social network does
not heavily rely on housing and community means. This notion needs to be further examined
in the qualitative interviews.
6.3.3 Impact of housing on subjective poverty
In terms of the impact on subjective poverty, the multiple linear regression analysis shows
that after the demographic variables were controlled, income showed a significant and the
largest impact amongst the three different DVs of non-income poverty in the model. The
effect of income reveals a relatively substantial influence on subjective poverty compared
with the regression models of deprivation and social exclusion. This finding demonstrates the
importance of income level in explaining the subjective feeling of being poor by residents.
By contrast, the total explanation power of all housing factors on subjective poverty is larger
279
than that on social exclusion. This result shows the role of housing in subjective assessment
of poverty. Various housing factors demonstrate different effect on subjective poverty.
High level of community facilities problems shows the significant impact on subjective
poverty compared with low level. The community dimension of housing is usually neglected
in the discussion of housing and poverty issue in Hong Kong. This finding indicates that the
households are concerned about not only the indoor environment but also the local
communities, and these aspects further affect their self-perception of poverty. The study
reaffirms the result of previous studies that highlight the importance of community factors
and availability of social service in the subjective assessment of well-being (Low et al., 2018;
Mohit et al., 2010). However, housing location is insignificant with subjective poverty in
study. Households that lived far away from CBD are with low income or poor, and this
situation may lead to subjective poverty. Nevertheless, the distance from CBD is not a crucial
factor for people to consider themselves as poor or not. However, further studies should be
conducted in other district divisions and locations (Guo, Chang, Sha, & Yip, 2018).
Living area and indoor housing problem show a significant relationship with subjective
poverty. This finding reveals that the physical environment of housing is crucial for residents
to assess their poverty status. The aforementioned result is coherent with a number of studies
in the past (Amerigo & Aragones, 1997; Rudolf & Potter, 2015; Zhang et al., 2018). This
finding reaffirms the importance of housing conditions for the self-assessment of the
280
subjective well-being. Hong Kong people have become increasingly concerned with the
quality of life. This finding also reveals the importance of housing perspective in promoting
the subjective well-being of residents. However, the quantitative analysis uses a cross-
sectional data and cannot examine whether an adaption effect of subjective poverty exists
(Foye, 2017). The relationship between housing factors and subjective poverty needs to be
examined via qualitative interviews.
Housing cost was associated with subjective poverty. This finding was different with
the research of Lawless and Lucas (2011) who found no significant relationship between
housing price and life satisfaction. The result of the SEM model analysis of subjective
poverty reveals that the effect of housing cost on subjective poverty is mediated through
living area and indoor housing problems. This result enriches the understanding of the effect
of housing cost on poverty situation in existing literature. The use of housing cost as an
independent factor to study the impact of housing is difficult.
This finding is important in Hong Kong because of the limited research about the impact
of housing factors on the subjective assessment of well-being. Although income shows a
large effect on subjective poverty, the influence of other housing factors is substantially
important. Income enhancement is not enough to make people feel less poor. Improvement of
housing situation, especially the physical environment, may significantly help people
eliminate subjective poverty.
281
6.3.4 Mediating and interacting impact of housing factors
The SEM analysis shows that the income influence on non-income poverty is partially
mediated by housing factors. The result confirms several hypotheses and rejects some (Table
6-16).
Table 6-16 Hypothesis and acceptance
Hypothesis Accept?
H1(SEM): the impact of income on deprivation is mediated by housing Rejected
cost
H2(SEM): the impact of income on deprivation is mediated by living Accepted
area
H3(SEM): the impact of income on deprivation is mediated by indoor Accepted
housing problem (partial)
H4(SEM): the impact of income on social exclusion is mediated by Rejected
housing cost
H5(SEM): the impact of income on social exclusion is mediated by Accepted
living area
H6(SEM): the impact of income on social exclusion is mediated by Rejected
indoor housing problem
H7(SEM): the impact of income on subjective poverty is mediated by Rejected
housing cost
H8(SEM): the impact of income on subjective poverty is mediated by Accepted
living area
H9(SEM): the impact of income on subjective poverty is mediated by Accepted
indoor housing problem (partial)
The housing cost for the overall population shows no mediating effect between income
and deprivation, social exclusion, or subjective poverty; thus, H1(SEM), H4(SEM) and
H7(SEM) are rejected. Living area mediates the effect of income on deprivation and social
exclusion and subjective poverty; thus, H2(SEM), H5(SEM) and H8(SEM) are confirmed.
The indoor housing problem does not directly mediate income and other three dimensions of
282
poverty. However, this problem mediates the impact of living area on deprivation and
subjective poverty. No significant mediating effect between income and social exclusion was
found, H6(SEM) is rejected. However, H3(SEM) and H9(SEM) are partially accepted
because the effect on income on deprivation and subjective poverty is mediated through
indoor housing problem by living area. In terms of private rental housing, H1(SEM) and
H7(SEM) are partially accepted. The impact of income shows significant influence on
housing cost and cause further impact on deprivation and subjective poverty through living
area.
The result of SEM models reveals the key role of income on other non-income
dimensions of poverty. The mediating impact of housing depends on different housing
dimensions. Generally, living area is a crucial housing factor in these models for contributing
a mediating effect. This result is an important finding in Hong Kong because the living space
of Hong Kong residents is declining in the past years, especially for the low-income families.
Many households are forced to live in small living areas because of the continued increase of
the housing price and rent, and the private rental housings become less affordable. This result
demonstrated that the small living area not only causes poor indoor problems but further
influence the deprivation, social exclusion, and subjective poverty situation.
Housing type is found as a key moderator in the models. The weight of housing effect is
amplified in the sample of private housing residents. The variance of deprivation and
283
subjective is explained by the housing variables in the private rental housing group. For
example, the effect of living area on deprivation is larger for residents living in private
housing rental than those living in other types of housing. This result is consistent with the
findings in the previous session, which shows the crucial role of housing type in influencing
the poverty situation of residents. Private rental housing residents suffered more from
housing poverty problems. This finding is especially important to Hong Kong because the
percentage of population living in private rental housing is increasing in the past years.
Specifically, the private housing tenants suffered more from housing problems than other
factors considering income level.
Another noteworthy finding from the SEM analysis is the interaction amongst housing
factors. The regression models focus on the independent effect of each housing variables but
do not study the interaction amongst them. The result of the SEM models indicates that
housing cost shows no significant effect on the DVs but reveal an indirect influence through
living area and indoor housing problems. This circumstance reminds researchers not to
consider housing cost as an independent factor in housing or poverty research. Housing cost
is usually correlated with other housing factors, such as housing size (Sunega & Lux, 2016)
and affordability (Zhang et al., 2018). Although the SEM model demonstrates the path
relationship amongst income, housing, and non-income poverty in general, the dynamic and
mechanism in the relationship is yet to be explored. In terms of interaction amongst housing
284
factors, the qualitative interviews can help in examining why and how the residents make a
choice amongst housing factors, and how the housing situation affect their daily lives. The
result will be presented in the next chapter.
6.3.5 Exploration of ‘multidimensional housing poverty’
This research examined the relationship of housing factors and non-income poverty. To
the best of my knowledge, this study is the first to examine the impact of housing factors on
deprivation, social exclusion, and subjective poverty in Hong Kong. This study reveals the
‘housing poverty’ situation in Hong Kong via the illustration of ‘housing induced poverty’
(Kutty, 2005) or ‘poverty caused by housing factors’ (Stephens & vanSteen, 2011). The
nature of poverty (Alkire & Foster, 2011; Chen, Leu, & Wang, 2019) and ‘housing poverty’
are multidimensional. For example, the result in this study shows the situation of ‘housing-
induced deprivation’ or ‘housing-induced subjective poverty’. The concept of
multidimensional housing poverty not only highlights various dimensions of poverty, such as
material, social and subjective, but also emphasises the role of housing as a crucial factor in
affecting multidimensional poverty.
Amongst the housing factors, indoor housing problems show a relatively large and
significant effect on three dimensions of non-income poverty. Living area also reveals a
critical role in mediating the effect of income on non-income poverty. This study highlights
the crucial role of physical housing environment in contributing material, social and
285
subjective poverty. The direct effect of housing cost on all non-income poverty dimensions is
weak compared with those of other housing factors. Housing cost reveals the effect through
living area and indoor housing problem. The influence of housing location and community
facilities show moderate effect amongst housing factors. This result demonstrates that the
physical housing factors are more crucial than monetary and geographic ones due to the
multidimensional housing poverty in this study. Nevertheless, research with additional
comprehensive variables is needed to validate a general conclusion.
Subjective poverty is the most explained DVs in the all models amongst the three
dimensions of non-income poverty. More than 30% of variance of subjective poverty is
explained in the multiple linear regression model. Meanwhile, more than half of the
explanation is contributed by income, which is coherent with the traditional understanding of
poverty theory. By contrast, housing factors generally demonstrate a large effect on
deprivation and least influence on social exclusion. Such notion implies that the concept of
housing poverty is applicable in the material perspective and less in social dimension.
This finding contributes in confirming whether housing is a crucial factor in
multidimensional poverty. Previous study of multidimensional poverty commonly focuses on
measurement and prevalence (Alkire & Foster, 2011; Bourguignon & Chakravarty, 2003;
Laderchi et al., 2003). The cause of multidimensional poverty has been insufficiently studied.
The housing cause of multidimensional poverty is rarely explored. This study fills in this
286
research gap and proposes a framework for further research in this area.
6.4 Short conclusion of the quantitative analysis
The use of census population data from 1996 to 2016 helps in answering RQ1 by
examining the housing poverty situation with various measurements. The overall ratio of the
housing poverty rate did not increase in the period from 2001 to 2016 despite the increase of
the housing price and rent. The increase of homeowners that finished mortgage payment in
the past years is one of the main explanations. However, the housing poverty situation of
private rental housing residents is becoming seriously worse. The housing poverty
measurement with analysis by housing type helps us distinguish the ‘housing problem’ of the
poor and ‘housing profit’ of the rich. The residual housing rate also increased from 2011 to
2016, thereby indicating that many people cannot maintain a basic living standard after
paying a housing cost. The housing poverty measurements are crucial in Hong Kong to
identify those vulnerable groups that suffered from high housing cost. The political economy
analysis helps us explain the changes in the housing poverty situation in the past 20 years.
The analysis of the SPPR dataset examined the impact of housing factors on deprivation,
social exclusion, and subjective poverty. Overall, living area and indoor housing problems
caused significant impacts on non-income poverty amongst the housing factors. However, the
impact of housing cost and living location are weak. By contrast, subjective poverty is more
influenced by different types of housing factors compared with other non-income poverty;
287
however, social exclusion is the least affected dimension of poverty. The SEM analysis
revealed that living area is the common mediator between income and non-income poverty.
Housing type is also the moderator in the models. Private rental housing amplified the impact
of the housing factors.
288
Chapter 7 Qualitative Analysis
This section discusses the quantitative analysis performed in this study. It aims at
answering the research questions on the mutual impacts between housing and poverty and
examining the experience of people living in housing poverty. It also attempts to address the
questions that remain unanswered in the quantitative analysis, such as the influence of
housing factors on the livelihood of residents and their reactions or the residents’ subjective
perception of their housing and poverty situation. Twenty residents, who currently live or
have previously lived in subdivided flats, are interviewed. In addition, cases are selected
considering maximum variation and critical and deviant sampling with various backgrounds
of age, sex, family composition and housing situation. The interviews aim at exploring their
living experience on the basis of the research theoretical framework.
In this chapter, firstly, the background of qualitative analysis, including profile of cases,
analysis method and overview of themes, is presented (Section 7.1). Secondly, three stories of
the housing poverty are revealed as an overview of the lives of the housing of poor residents
(Section 7.2). Third, the themes discovered from the interviews are illustrated with the
wordings from the residents. Discussions follow in each session (Sections 7.3 to 7.7).
289
7.1 Background
7.1.1 Profile of Informants
Twenty informants were interviewed. All of them currently live or have lived in private
rental cubicles or subdivided flats in the past five years in Hong Kong before the interviews.
They either belong to income poor and/or housing poor categories. The income poor category
indicates that the household income of the family is below the official poverty line, which is
50% of the median household income by household size. Housing poor indicates that the
household spends more than 30% of their income on housing cost. Among them, eight were
males and 12 females. Fifteen informants lived with other family members, while five were
single. In terms of housing type, 12 lived in subdivided flats and five in cubicles. Contrarily,
three lived in subdivided flats in the past five years but have currently been allocated public
housing. More than half of them were income poor. Fifteen were housing poor. Seventeen
cases faced housing issues, such as increased rent or shifting home (Table 7-1). The
background of each case is listed in Appendix C.
Table 7-1 Profile of informants
N %
Sex
Male 8 40
Female 12 60
Age
18-40 8 40
41-59 6 30
60 or above 6 30
290
Family composition
Singleton 5 25
Couple without children 1 5
Couple with children 11 55
Single mother with children 3 15
Education
Primary or below 5 25
Secondary 14 70
Tertiary or above 1 5
Place of Birth
Hong Kong 3 15
China (immigrated less than 7 years) 7 35
China (immigrated 7 years or more) 10 50
Housing type
Cubicle/other 5 25
Subdivided flats 12 60
Public housing 3 15
Income poverty (<50% of medium of household income)
Income poor 11 55
Not income poor 9 45
Housing poverty (>30% housing cost per income)
Housing poor 15 75
Not housing poor 5 25
Rent increased in the past 5 years
Increased 17 85
No change 3 15
Moved housing in the past 5 years
Yes 17 85
No change 3 15
7.1.2 Data Collection and Analysis
The researcher interviewed the informants via home visits or meetings in local
community centres. A semi-structured interview guide was developed to study the impact of
housing factors on the daily lives of informants (Appendix E). Interviews were conducted
between March and May 2019 in Cantonese. Each interview lasted for 60–90 minutes and
291
was audio recorded. The audio-recorded data were transcribed in Chinese by research
assistants and inputted into the software NVivo version 12. Thematic analysis was conducted
by the researcher and guided by the suggestions of Miles and Huberman (1994). The
transcribed data were initially coded under open coding to capture the meanings of the
wordings from the cases. The codes were then summarised into meaningful themes
concerning the conceptual framework derived from the literature review. The final themes
and coding used were translated into English by the researcher.
7.1.3 Overview of Themes
Five main themes, namely, understanding of housing and poverty, critical life events, the
impact of poverty on the housing situation, the impact of housing on the poverty situation,
and the response and coping strategies of residents, were identified from the data of the
qualitative interviews. Figure 7-1 shows how they are linked to other themes conceptually.
The social system and policies such as ideology, governance, housing, and welfare policies
represented the analysis background. These concepts influenced the housing and poverty
situation experienced by the residents. Furthermore, critical life events such as marriage,
divorce, illness, and eviction were influenced by their housing and poverty situation. From
the life experience of the informants, housing and poverty mutually influenced each other.
The response and coping strategies of residents as agents were found against the structures,
such as resistance, advocacy, and adaption, and these reactions may further change the macro
292
system and social policy (Figure 7-1).
Figure 7-1 Relationship between themes
293
7.2 Three Stories of Housing Poor
Before studying the common themes revealed from the experience of residents, three
stories are initially presented to demonstrate the hardship faced by the housing poor residents.
Their experience was full of sadness, struggle, and resilience.
7.2.1 ‘I want to jump down to the street … I am exhausted and depressed …’
Case R experienced the housing path from public housing to a subdivided flat and then
back to public housing. His family suffered a lot when living out of public housing. Case R
even wanted to commit suicide under extreme stress when living in the subdivided flat. Case
R, a 36-year-old male, lived with his wife and two children. He migrated from mainland
China to Hong Kong 20 years ago and lived with his mother and father in a public housing
estate for many years. He got married in 2010. His friends told him that households can
choose to live with their original families instead of moving out to live independently if they
were poor. However, sometimes, residents were forced to move out, as they could no longer
live together in a small flat, especially when the relationship between the mother-in-law and
daughter-in-law was not good.
Case R and his nuclear family needed to move out, but it was difficult to find a flat. In
2010, Case R worked in a warehouse and earned around 5000–6000 HKD; his wife was
pregnant at that time. They could only afford a very small sub-divided flat in Kwun Tong that
would cost them more than 3000 HKD. However, the environment there was extremely bad.
294
Later, with the enforcement of the legislation on minimum wages in 2011, the wages of Case
R increased to 11,000 HKD in 2017 and he moved to a larger flat of around 200 square feet.
However, the rent also increased to 8,000 HKD. He attempted to work two jobs, requiring
more than 14 hours every day.
‘I would wake up at 6:30 am and drop my children to school. Then I would go to my
office at 9 am and work until midnight, almost 15 hours per day … I would sleep for a
while and then came the next day again … It was harsh…’ (Case R)
With long working hours and low income, Case R found it difficult to meet his friends
and felt socially excluded. He could not afford social activities and needed to reserve money
to pay rent. Moreover, he was exhausted living in a small subdivided flat after working from
morning till night and found it extremely stressful returning to his small living flat. Once, he
wanted to commit suicide.
‘… to be frank…I think I went into depression when I was living in the subdivided
flat … Once I told my wife I want to jump down to the street … I was really in trouble; it
was too difficult to continue living there … It was like a rubber band under high tension,
going to break.’ (Case R)
All that he could do was wait for public rental housing to change the housing situation.
He applied for public housing many years ago. He had noted the ‘3-years’ waiting time
promise of the government years ago but waited for more than nine years. He was very
295
anxious during the waiting period. In the allocating stage, even the residents were worried
that the location of the public housing might not be favourable to them. Moreover, they
would be required to wait for several months or more if they rejected the first chance.
However, there was no guarantee the second chance would be better than the first. Informants
found it difficult to decide if the first chance was not favourable to them.
Case R admitted that getting public housing was a destination for him in some sense and
he planned to live there until old. He believed that, in the future, he might be unable to buy a
flat by himself and needed to stay in public housing for a very long period.
‘I think public housing is a destination for me. I mean, we need public housing, as we
cannot afford to buy a flat, considering my future economic situation. Moreover, I have
two children, with many expenses. How can I save money to buy a house?’ (Case R)
After getting into public housing a year before the interview, Case R felt much relaxed,
as he now had some more money to consume. He changed to part-time work and could
accompany his children and walk around the community to meet other people.
7.2.2 ‘I found houses more than 30 times … I have no choice to move …’
Case J, aged 29 with tertiary education, lived in a small subdivided flat with her baby. She
and her boyfriend had a baby around two years ago. However, her boyfriend deserted her,
making her a single mother. Thus, she needed to rely on CSSA to support her and her baby.
Within the tight budget of the rent allowance of CSSA, she found it very difficult to find a
296
suitable flat.
Case J was frustrated about the low level of CSSA subsidy, both the standard rate and
rental subsidy. She was living in a flat with a rent of 5,800 HKD, but the rental subsidy was
around 3000. After the housing cost, she is left with only 4000 HKD to spend. She and her
baby live in deprivation. She can hardly afford to buy milk powder for her baby or fresh food
from the market. She had to take second-hand clothing from her friend to save money.
‘Basically, housing is what I need the most … but after paying rent, how can I afford to
buy food? For clothing, should I wait for second-hand clothes all the time? My friends
sometimes rebuked me for being unable to buy milk powder for my baby, but who would
pay for me? It costs more than one thousand for a month. How can we afford it?’ (Case J,
Female, 29, subdivided flat)
Moreover, she neither had savings nor hardly any money to buy anything besides daily
necessities. She even gave up her desire to dress better. Besides deprivation, Case J found that
her friends gradually left her, and she kept getting socially excluded. She explained that,
sometimes, she needed to borrow money from friends or live in their home, as she cannot
afford the rent. However, she said it was difficult for her friends to support her financially for
a long time.
297
‘… I need to borrow money from my friends … How can I make other friends? How
will your friends allow you to live with them for a long time? I have no friends at all
now … They left me one by one …’ (Case J)
When the interviewer asked her whether she felt she was living in poverty, she answered
bluntly she was poor. She felt she had lost her freedom due to living with low income and
high housing costs. Her choices were limited, making her feel poor.
Moreover, Case J found it very difficult to find accommodation, not only due to her low
income but also because she has to live with her baby. Landlords often rejected her due to her
baby. Further, despite lowering their expectation for housing the poor, households still had
limited choices. Case J had tried more than 30 times to find a flat in the past five years.
Moreover, Case J had no bargaining power to negotiate with the landlord. When she
attempted to complain to the agent, she was reprimanded in foul language. She became
emotional and extremely angry. She was helpless and had to scold back.
‘Before moving here, the housing agent scolded me a lot. He even used foul language. I
got emotional and scolded back. I was angry at that moment. I thought if I had to take
care of my baby, I needed to fight back.’ (Case J)
Concerning CSSA, the amount of the allowance was not enough for her living. Case J
could not perceive any progress in her future. She simply hoped to get public housing as soon
as possible, to go out to work thereafter, and enjoy greater freedom.
298
7.2.3 ‘I cannot sleep at my home … I voice out and hope the policies can be
changed…’
Case P, a 42-year-old male, migrated from mainland China to Hong Kong. He lived with
his wife and two children in a small subdivided flat with a dreadful living environment. He
worked as a cook with a low salary and long working hours. Nevertheless, he actively joined
a community group and participated in civic actions in policy advocacy.
Case P came to Hong Kong seven years ago. He liked living there, except for the
housing situation. He stated that many conditions and services in Hong Kong, such as human
rights, liberty, the judicial system, and medical service, were better than mainland China,
except housing. He initially sold fish in a wet market, but his hands were strained after a year.
Then he began working as a security guard for 12 hours per day. However, he could neither
sleep well at home nor concentrate during work. Thus, he started working as a cook in a
hostel with a better salary. Nevertheless, as the income was just around 15,000 HKD, he
continued living in the small subdivided flat.
As his income was low and employment was precarious, Case P could not save much.
He seldom dined outside with family or participated in gatherings with friends. He had to
spend more than 200 HKD for a meal – the total food expense for his family for one day. He
had to dine at home, even on his wedding anniversary, as eating outside was very expensive
for his family.
299
Despite facing numerous difficulties, Case P actively participated in an action group of an
NGO and engaged in many social actions, including meeting government officials and
petitions aimed at reforming the housing policy in Hong Kong.
‘I am concerned about the housing problem … I met many government officials and told
them the reality … I also went to the government headquarter and many other places to
demonstrate … The government talks about transitional housing this year was what we
advocated in 2013 …’ (Case P)
His expectation from the government was one pushing factor for him to actively join the
action for policy advocacy. He thought it was the government’s duty to enhance the living
conditions of residents, and he urged the government officials for more home visits to
understand the livelihood of people at grassroots. Moreover, he believed society would not
change if no one raised their voice. Furthermore, the social policy may change if people
persist to participate in the action to change it.
‘It is not easy. We are not paid. I have to pay for transportation to join social actions. We
neither get benefits nor money. So, why do I join? I do not know how to put this
across … We need to raise our voice for the government and the public to know … If you
do nothing, nothing will be changed. We should persist, not for days and months but
years.’ (Case P)
On the other hand, he found support and was more relieved on joining the social group
300
in the community centre. He was frustrated with the housing situation at home and working
for long at the workplace, but he found people in the action group were in the same boat, and
he could share his sorrow freely there.
‘I believe I have made many friends here. I am very glad we stay in touch with each
other. We can share our common worries … It is really like a home for me. With not
much pressure and a depressed mood, I feel happier.’ (Case P)
301
7.3 ‘To Me, Housing is … Poverty is …’: Multi-Dimensions of Housing and Poverty
The essence of housing and poverty for residents should be understood before
investigating the relationship between them. This section defines housing and poverty from
the perspective of residents.
7.3.1 ‘Housing is…’: 5S of housing need
A house is not merely a place with walls but means much more than a physical place
(Stone, 1993). When asked what the most important thing in life is, residents claimed it is not
money but housing. Some of them thought housing was their basic need. Some highlighted
housing provided them with privacy, security, stability, and a space to build relationships.
Furthermore, it provided a place to satisfy their basic needs. If housing problems are not
solved, people cannot claim they have good lives. On being asked what they cared most
about housing, informants highlighted several key dimensions of the importance of housing,
which have been summarised into a 5S framework that includes security, sanitation, space,
stability, and separateness.
a) Security
Why is it basic? Housing provides families security to live with stability. Moreover,
most of them treasured privacy in their home (King, 2015; Stone, 1993). On being asked why
housing was important for them, some of the informants cited privacy and security as
reasons, without which they felt the house was not home.
302
Case T, who used to share a flat with more than 30 people, realised he could not live
while feeling insecure every day. He felt unsafe in his living place because his flatmates
would cause a lot of disturbance and steal others’ stuff.
‘My neighbour would come to my side at night. If I woke up and found him next to me, I
would scold him; if I kept sleeping, there was a possibility of my things getting stolen.
There were many drug abusers, and thus, I did not dare to sleep. Even the landlord told
me to take care of my stuff and not call the police if it was stolen.’ (Case T, Male, 45,
singleton hostel)
b) Sanitation
Hygiene was one of the main concerns of the residents. While hunting for flats, they
prioritised a sanitised building and community environment.
Interviewer: ‘What concerns you most about living?’
Case H, female, 71, subdivided flat: ‘A clean and good environment! I want a clean and
safe environment.’
Interviewer: ‘What is your concern when finding a flat?’
Case O, male, 63, cubicle: ‘I consider a clean and hygienic environment and to keep
everything simple.’
c) Space
All residents cared about the living space, especially poor families with a low income to
303
afford larger housing. The impact of living space on the daily lives of residents is illustrated
in the following excerpt from one interview.
‘Money is not the most important, frankly speaking. The living environment is the most
crucial part. If a four-person family lives in a flat of only 10 square metres, how can they
continue living?’ (Case P, Male, 42, subdivided flat)
d) Stability
Stability is crucial for personal and family planning. Families are afraid of moving
around or being evicted by landlords. Finding flats is difficult and costly for them. Case G
was afraid of being evicted merely because the landlord said the rent was relatively cheap.
This one statement made her worried for many days.
‘Yes, I am afraid of eviction. You never know when the landlord will increase the rent or
kick you out. Either way, he keeps saying the rent is very reasonable, and so, I am
worried about moving to another house.’ (Case G, Female, 35, subdivided flat)
In Hong Kong, there is a common practice of signing the rental contract as ‘one year fix,
one year flexible’. In the fixed year, the tenant cannot end the contract or must pay the
remaining rent for that year. Likewise, the landlord can also not evict the tenant without
reasons. However, in the flexible year, the landlord can end the contract with a month’s
notice.
e) Separateness
304
Most residents wanted independent units if they could afford them. Each family member
desired privacy. Some could not tolerate a shared living environment. Privacy was one of the
most important aspects of a living arrangement for them. Despite greater housing costs they
chose a smaller flat with greater privacy.
As Case K, male, 54, cubicle, pointed out, ‘It lacks privacy! Someone keeps talking
beside you while you are asleep. Sometimes, they suddenly knock on the door to chat with
you. How do I deal with it? There is no privacy here!’
Moreover, parents are concerned about the privacy of their children, especially while
growing up. Despite living with their family, privacy was important for each member. Case
G, with a son and a daughter, said the bathroom was very small. The sight of her son coming
out naked after bathing makes her daughter feel uncomfortable.
‘My son is 11 years old now and it is inconvenient to live in a small flat … Sometimes,
he walks out naked after taking a shower. However, my young daughter does not want to
see that. She needs more individual space while growing up.’ (Case G, Female, 35,
subdivided flat)
7.3.2 ‘Poverty is …’: Material, social and subjective dimensions
Resident experience of poverty is not only limited in the monetary dimension but also in
the material, social, and subjective dimensions. Although ‘lacking money’ is one of the most
common descriptions of poverty, the meaning of poverty extends beyond that. Using the
305
framework of deprivation (Townsend, 1979b), social exclusion (Gordon et al., 2000; Silver,
1994) and subjective poverty (Colasanto et al., 1984), the resident experience of poverty is
revealed more comprehensively.
a) Deprivation – experience of material suffering
In Hong Kong, the poor may not suffer from starvation, but given the high living
standard, low level of income, and welfare of the poor, they suffer from a grave deprivation
problem.
Some informants had to squeeze their daily expenses. Case E had to pick old
clothes from the rubbish collection station or accept second-hand clothes from NGOs.
Case E, female, 63, cubicle, stated, ‘Sometimes, I got the clothes from rubbish
stations or social workers. The people in the service centre are nice; they sometimes donate
clothes to the needy.’
When Case I, male, 33, rooftop housing, was asked how frequently he bought clothes
or shoes, he said he just bought one or two items a year. He was a cleaner getting minimum
wage and could not afford much in terms of clothing. He stated, ‘Umm … I buy clothes
around twice every year. Yes, two clothes and maybe a pair of shoes for the whole year.’
Moreover, poor residents found eating outside home highly expensive and used different
strategies to lower their expenses. Case O spent merely 25 HKD daily on food for himself
and his daughter. He would buy some meat, mix it with rice, and eat it for an entire day.
306
Despite a tight expense, most of them had no savings and particularly felt insecure at the end
of the month. Case I said he could not save most of the time, especially after paying the
monthly rent.
‘I cannot save any money! I end up spending all of my money most of the time … Most
of the money goes into paying rent. Normally, I give one thousand to my mum … and
that is all. I cannot save anything.’ (Case I, Male, 33, Rooftop housing)
Sometimes, he had to borrow money from his friends to pay rent. Being in a precarious
job, he would not get his salary on time, but paying rent was a must for him, so he had no
choice but to borrow money from friends.
b) Social exclusion – Experience of standing at the margin of society
The poor were excluded from society for various reasons. Some wanted to lower their
expenses and thus did not meet friends or join social activities, while others felt embarrassed
to meet others.
Friends sometimes invited Case C to meet and eat out, but he always rejected, as he had a
low income and could not afford to go out for gatherings.
‘I seldom go out to play. It requires money, and I am not interested in it. If I have money,
I will go out with my friends to eat and play. I even rejected when my friends dated me;
yes, it is true.’ (Case C, Male, 69, subdivided flat)
Case K, a singleton receiving social welfare, also lived in a similar situation. He found it
307
difficult to make friends when he had no money.
Interviewer: ‘Do you think taking CSSA may affect your social life?’
Case K, male, 54, cubicle: ‘Yes, I have no friends now. We have nothing to talk about.
All of them have become very realistic and materialistic. As a man, it is difficult to
borrow money from friends, right? I have no money, and it is difficult to find friends.’
Case E felt sad and said she did not have friends and relatives because neither did she
have any money nor did people want to meet her. Moreover, she lived far from the town
centre to save money and so seldom went out to meet others to lower transportation expenses.
Thus, she became more socially excluded from society.
‘I have no friends and relatives. Why do I say so? When I was sick, they stayed away
from me … If you are Lee Ka-Shing (the rich tycoon), you will have many friends … I
usually walk to the service centre, as taking a bus is very expensive. Sometimes, I walk
for an hour.’ (Case E, Female, 63, cubicle)
c) Subjective feeling poor – Shame, low status, and no dignity
Being materially or physically deprived does not mean subjectively feeling poor.
Residents living in a deprived situation with low income felt they were not living in poverty,
especially when they compared themselves with the poorer and their worse experiences.
When the interviewer asked the residents ‘Do you think you are living in poverty?’ they
responded with different answers despite their common background of income poverty.
308
Moreover, subjective poverty is in the essence of comparison. When residents compared
themselves with someone in a better situation, they felt bad. Contrarily, they felt they were
living quite well when compared with the poorer.
A comparison happens in everyday life. For example, Case A felt she was living in
poverty when she saw the branded clothes others wore. Another example pertained to tutorial
classes for children. Normally, poor families cannot afford tutorial classes and can merely
apply for free supplementary classes provided by the school. There is keen competition for
academic achievement in school, and parents feel they are poor when they know children
from other families can attend many tutorial classes.
‘You can see the brands they use, their handbags, and the places their children go for
tutorial classes … Sometimes I compare with them … I think they look down upon me …
I just look like an old woman from a rural area.’ (Case A, Female, 28, subdivided flat)
On the other hand, some residents, despite living in a deprived situation, found they
were living well when they compared themselves with others with worse livelihood or their
experiences in the past. For example, Case D thought her living standard was fine when she
watched the news about the poorer housing situation in Hong Kong.
‘Yes … Those people are really in a poor situation … I think, my situation is fair, not the
poorest … When I see four-person families living in smaller flats, I feel very sad … It is
309
the government that has created the situation and pushed people into depression.’ (Case
D, Female, 50, subdivided flat)
Case I, who lived in a rural area in China for many years, believed his well-being had
improved immensely compared to his experience. The same repairing job in China merely
earned him around 2000 to 3000 dollars, but he could earn more than 15000 in Hong Kong.
He was happy about that.
‘Umm … my income and living standard improved … compared with my living in a rural
area. I was a repair worker in the mainland but only earned 2000 to 3000 dollars per
month. Now, I can earn 15,000. It is a huge difference.’ (Case I, Male, 33, Rooftop
housing)
7.3.3 ‘Housing and poverty are…’: the relationship between housing and poverty
The relationship between housing and poverty is complicated. Both dimensions cannot be
simply defined without each other but are mutually defined. Moreover, housing and poverty
exerted an impact on each other and acted as an interactive multiplier. According to the views
of residents experiencing both housing and poverty problems, both dimensions and their
experience cannot be described separately.
When the interviewer asked the informants ‘What is poverty?’ some of them highlighted
housing was the core element in defining poverty. Case C replied that as he was poor, he
could not make a down payment to buy a house. He thought owning a house implies stability
310
in wealth and not living in poverty.
Case C, male, 69, subdivided flat: ‘Poverty? I think if you do not have one million
dollars, you are poor in Hong Kong because you cannot buy a house yourself.’
Interviewer: ‘You mean you are poor if you cannot buy a house?’
Case C: ‘Yes, the ability to buy a house implies you have a stable asset, doesn’t it?’
Case F, female, 46, subdivided flat, believed those with a higher income or rich do not
live in subdivided flats with a poor living environment and very small living place, thereby
directly claiming that living in a subdivided flat means living in poverty. She stated, ‘Well,
what is poverty? I think it means living in a subdivided flat. People who are not poor with
enough assets will never live in a subdivided flat.’
Moreover, residents living in poor housing experience the interactive impact of housing
and poverty. When they were poor, without enough income, their only choice was poor
housing. At the same time, they had to pay high rent for private housing that used a large
amount of their income, making them poorer.
‘I think in Hong Kong, poverty means no place to live. If you do not have a good place to
live, you are poor, for example, if you are not renting public housing but a private one. I
have to pay 6000 HKD for rent out of my 10,000-HKD salary’ (Case S, Female, 35,
subdivided flat to public housing)
Most of the time, housing and poverty do not impact the livelihood of households
311
individually but together. The combined effect results in many negative consequences for
poor families, such as ruined family relationships, harmful consequences for health and
mental health, loss of freedom in daily life, and low quality of life.
7.3.4 Short discussion
In this section, the voices of informants reflect their understanding of housing and
poverty. It presents the diverse nature of housing and poverty. Residents with different
backgrounds show varied concerns for their living situation.
The 5S dimensions provide a new framework to understand the housing concerns of
residents. This approach enriches the existing literature about the nature of housing via a
subjective perception of low-income families in Hong Kong. The framework demonstrates
the concern of residents about housing. For low-income families, the functions of housing
focus on the physical and material level (Stone, 1993) but not as a commodity (Harvey, 2014)
or capital accumulation (Piketty, 2014). The latter concept of housing is far from grassroots in
Hong Kong. Almost all informants view housing as a basic need, and the 5S framework
concludes their definition of basic necessity.
The concept of deprivation helps capture the material inadequacy of residents. It
echoes the idea of Townsend (1962), that is, ‘subsistence’ is an inadequate criterion of
poverty. Few citizens suffer from starvation, especially in developed cities, but this situation
does not indicate the absence of poverty. Relative deprivation helps show the situation where
312
people lack socially perceived essential items in their daily lives (Bradshaw & Finch, 2003).
Stories of residents reveal the experience of deprivation in Hong Kong and how they feel
when deprived. Deprivation is a result of lack of money, and material inadequacy is a major
concern of the informants, compared with social exclusion and subjective poverty. This
finding supplements the results of the quantitative analysis, which cannot reveal the
subjective experience of informants.
For social exclusion, stories of residents show how they are marginalised out of
society due to the lack of resources. This finding is far from the original definition of social
exclusion, which concerned exclusion from the labour market (Silver, 1994). The informants
feel excluded from their friends and social participation. The conceptualisation of social
exclusion in the latter years concerning social relation, social network (Levitas, 2006a), and
cultural activities (Wagle, 2002) is more applicable in Hong Kong than in other cities.
Nevertheless, the term ‘social exclusion’ is not commonly used in Hong Kong.
Subjective poverty is a complicated phenomenon. The need and comparison theories
are usually used to analyse the situation (Liao, 2009). From the interviews, it is found that the
subjective poverty situation of the residents is more coherent with the comparison theory.
Residents usually feel poor not due to the lack of necessity but because of comparing
themselves with their significant others. The background of households is crucial in
determining subjective poverty (Mahmood, Yu, & Klasen, 2018). For example, those who
313
experience poverty as children may show resilience to the current situation. Contrarily, the
adaptation theory (Foye, 2017) may not work for certain residents. Although some of them
lived in poor housing for years, they cannot adapt to the situation, given that the environment
is extremely undesirable.
The stories of residents demonstrate that housing and poverty are reciprocal nexus and
mutually defined. This definition echoes the literature that shows housing and poverty are
highly associated. The previous research studied the relationship between housing and
poverty from a macro perspective (Crisp, Eadson, & While, 2016; Madden & Marcuse, 2016;
Stephens & vanSteen, 2011). The present study supplements the micro perspective with
subjective views from residents living under poor housing conditions. To several informants
living in poor housing, the meanings of housing and poverty are interchangeable. Housing is
a part of the definition of poverty and vice versa. Nevertheless, the impact of poverty on
housing and that of housing on poverty are illustrated in the following section.
314
7.4 ‘I am Poor, because…’: Critical Life Events
Why did families get into housing poverty? The cross-sectional quantitative analysis is
limited to studying the snapshot of housing poverty. However, understanding how people
went into poverty, further affecting their housing situation, is important. The concept of
housing pathway (Allen, 2007; Clapham, 2002; Opit, Witten, & Kearns, 2019) reminds us to
view the life situation through a pathway approach with change and dynamics. This section
reveals the housing and poverty pathway of residents. Many informants have experienced
critical life events in the past, such as marriage, migration, illness, and significant changes in
their lives. These critical incidents may have led residents to suffer from poverty.
7.4.1 Family change, marriage, and broken marriage
Normally, in Chinese society, after marriage, most people move out to form new
families. Case L, male, 60, subdivided flat, got married in mainland China many years ago.
He lived independently in a small flat with low rent before but has to move to a bigger house
with high rent after his wife came to Hong Kong. He stated, ‘I needed to move to a bigger
house after my wife came to Hong Kong five years ago. When I was single, I could live
anywhere. However, you need a large place for a family, right?’
On the other hand, some cases needed to find a new flat to live due to a broken marriage.
Case G, a single mother with three children, lived with her husband and his parents, but the
experience was a tragedy for her. Her mother-in-law treated her very badly and she felt her
315
status was lower than even that of a domestic worker. She eventually left because of the
broken marriage but said she got more freedom than before.
‘We applied for divorce two years ago. He would drink all the time. Of course, the living
space is much smaller now, but I feel more relaxed. I never get scolded when I come back
home. Those seven years of living with my parents-in-law were disastrous …’ (Case G,
7.4.2 Migration from China
Several families migrated from mainland China and experienced a substantial change in
Hong Kong, especially in terms of the housing situation. Case I said he could pay 1,000
dollars to rent a larger house in the mainland but now had to pay more than 5,000 HKD for a
small subdivided flat. He liked most things in Hong Kong, except the housing and asked for
more support for new immigrants.
‘The flat here is much smaller than that in the mainland, you know. In the mainland, I can
rent a flat with a sitting room for around 1,000 dollars. However, now I can only put a
small television near my bed ...’ (Case I, Male, 33, Rooftop housing)
Moreover, Case A shared her children were not used to the living environment and
sometimes wanted to go back to mainland China for a living.
‘I immigrated to Hong Kong in 2017. When I arrived, I felt the flat was tiny, much
different than that in China. My children and I were not accustomed to it. Sometimes,
316
they wanted to go back to China … The living condition was harsh here ...’ (Case A,
7.4.3 Getting sick, the whole family changed
The sickness of family members could cause huge changes for families, leading to
poverty. It could be unexpected and uncontrollable, and no one simply has responsibilities for
the suffering. Case E lived with her sons for years and received financial support from them.
However, her older son developed a mental illness several years ago and her family was
broken down. She was forced to live outside and get social welfare to support a living.
‘My older son fell sick several years ago. It shocked me immensely. He visited a doctor
and was recovering. However, he cannot support me anymore. My family was broken …
Last year, my older son fell mentally sick, and I could not sleep. I had to move out.
Everything is broken …’ (Case E, Female, 63, cubicle)
Case T once argued with his flatmates, went to sleep outside, and suffered a stroke. This
changed his whole life. He could not work and had to merely wait for public housing to
improve his living standard.
‘My body is worsening. I had a stroke recently. I cannot think of any way out. My social
worker asked me to apply to a singleton hostel; at least, now, I do not have to walk
upstairs when back home …’ (Case T, Male, 45, singleton hostel)
7.4.4 Unstable employment, unstable housing
317
Case L, male, 60, subdivided flat, was a construction worker several years ago but
needed disability allowance after hurting his leg. His income dropped from more than 30,000
to around 2,000, and thus, he had to rely on his wife to take up a part-time job. He stated, ‘I
now rely on my wife’s income … I seriously hurt my leg at work several years ago. I am now
getting the allowance for the disabled, around 2,000 dollars monthly. It is sad …’
The husband of Case S faced employment insecurity and sometimes needed to change
his job. However, they cannot stop paying rent in the transition period. Case S shared they
had to keep savings to move to another house. Once the landlord increased the rent to an
unaffordable level, they needed their savings to find a new flat.
‘We saved a little money when my husband’s work was stable. If you were forced to
move out, you needed to prepare two to three months of rent. So, you used the saving
when you moved to a new house.’ (Case S, Female, 35, subdivided flat to public housing)
What are the causes of poverty and housing poverty? How did people get into poverty
and housing poverty? A continuous debate exists in the literature about whether poverty was
caused by individual or structural factors. In the present study, the quantitative analysis
showed the relationship between housing and poverty with statistics. However, the causes of
or the pathway to poverty were not studied. In the qualitative interviews, the stories of
residents revealed they usually face different critical events in their lifespan and fall into
318
poverty. Examples of critical life events included family change, marriage and broken
marriages, migration, getting sick, and change of employment and housing situation. Thus,
the causes of poverty and housing poverty were complicated and interlocked.
One group of researchers focussed on the individual cause of poverty. They argued
that people fell into poverty because of laziness (Murray, 1984; Somers & Block, 2005). If
people work hard, then they will not be poor and can be rescued from poverty in a short
while. It not only belongs to the rich people or those saved from poverty by personal effort
but also those living in poverty. Certain informants in this study denoted a similar belief.
However, this school has several limitations. Firstly, individual behaviour does not imply
individual fault (Calnitsky, 2018). For instance, in the change in the family setting, if
someone married and moved out to live independently and fell into housing poverty, then he
cannot be blamed for an individual fault. Marrying to build new families is customary.
Secondly, structural factors were overlooked. Separating individual incidents from the social
structure was difficult. Structural force can exist behind individual behaviour (Brady, 2019).
The same example indicates households fell into housing poverty after marriage because the
cost of housing was very high. Therefore, if the housing cost is affordable, then families may
not fall into poverty after marriage. Viewing housing poverty as individual responsibility is
unreasonable. Thirdly, even when the causes of poverty are individual, can the government
ignore the responsibility to help the poor? The government plays a role in enhancing
319
individual well-being and social welfare, especially for the underprivileged, regardless of
how they fell into poverty. What is the benefit of focussing on the causes of poverty as an
individual deficit? Will it help individuals get out of poverty, given their laziness or inability?
The answer is probably no. Even when an individual causes poverty, the government should
implement macro intervention to alleviate the problem in general.
The stories of informants help us understand the causes of poverty deeply and
comprehensively. Life is full of dynamics, accidents, and critical events. Individuals’ failures
cannot be simply viewed as their fault. People fall into poverty for different reasons, some
intentional and some accidental. Housing poor is a result of numerous factors. The
individualistic approach lacks the explanation for poverty; it not only deals with a large
population suffering from poverty but also fails to explain the structural force behind
personal behaviour. Further discussions on social policy and social structure will be revealed
in the following section.
320
7.5 ‘I am Poor, so…’: The Impact of Poverty on Housing
The poverty of residents is not only limited to their housing choice and information
access but also causes instability in their housing situation. Moreover, buying a flat is not a
choice for poor families. They lack the freedom to choose their place, location, and housing
quality.
7.5.1 No money, no housing choice
Can residents live in a better place? Yes, if they have much more money. However, many
of them have low income or cannot afford a high housing cost. If they have more money;
they will not make this choice. Case D’s only concern about housing was a cheap price. She
wanted to live in a larger flat with a lift that would not require her to walk upstairs but had no
choice.
‘I think, I only want a flat with low rent. It is good to have a lift in the building, it is good
to have a sitting room, and it is good if I do not have to walk upstairs, but I do not have
that choice’ (Case D, Female, 50, subdivided flat)
Moreover, it was common for residents to choose a flat on a higher floor but at a cheaper
price; for example, Case O had to walk up to seven floors daily.
‘I saw two flats with lifts before but nearly with double rent than this one. Then I viewed
this flat on the seventh floor; I have to use the stairs every time. Well … the air quality is
good on the upper floor, so I chose this. (Case O, Male, 63, cubicle)
321
7.5.2 Lack of social capital and poor information access
Choosing a house was not only about income but also the social capital and information
access. As the housing market was not transparent and information was limited, some
residents may have chosen without enough comparison. Case H choose her house based on
her friend’s suggestion but later found there were cheaper choices in other locations.
‘I am not sure where to find a good house. It seems there are places at a cheaper price, but
I do not know where. My relative once told me there are cheap flats in Ngau Tau Kwok,
but when I went there, I found none. Then I found an agent who referred me to Kwun
Tong. I finally chose one in a hurry. I do not know where to find a cheaper one indeed.’
(Case H, Female, 71, subdivided flat)
Some residents just rely on the information from housing agents but have to pay the high
agent fee from their limited savings. Case N, female, 72, subdivided flat, stated, ‘You do not
know how much is reasonable. When the agent said 5,000, you can say it is very expensive,
but you cannot find a reference.’
7.5.3 Buying a house is not my choice
Promoting homeownership was one of the agendas of the government. Government
officials usually claimed they helped citizens to own houses. Moreover, real estate companies
made many advertisements for selling private housing. The atmosphere put much pressure on
the household to buy a house in Hong Kong. However, it was not a choice for the poor.
322
When the interviewer asked informants whether they would buy a house in the future,
not only were most of the answers negative but some were even angry about it.
‘No! I never think of that … It is because my income is not at that level. My education
and cultural levels are not that high enough … I go to work and earn merely 10,000
dollars per month. It only covers my expenses … I really cannot think of buying a house.’
(Case A, Female, 28, subdivided flat)
Buying a house was far away from the grassroots. Even those who thought of owning a
house thought of their children first. They used the money to educate their children. The
housing price was very high for them, and they did not want to be the ‘slave of house’ that
would require them to pay an expensive mortgage for years.
Previous studies show that the poverty situation causes significant housing circumstances
for residents (Tunstall et al., 2013). However, the micro mechanism of the impact is
insufficiently revealed. This section demonstrates how low-income families have limited
housing choices not only because of lack of money but also because of inadequate social
capital and network. In Hong Kong, salaries of workers increased in previous years especially
after the legislation on minimum wages. Nevertheless, the salary level at grassroots remains
low, and the working poor population is still large (Cheung & Chou, 2016). Besides, housing
323
prices and rent increase at a faster pace than salary. Thus, low-income families cannot easily
find a flat or are forced to live in a smaller flat.
The stories of informants remind us that the housing poor not only lack money but also
social capital. The concept of social capital helps us understand the importance of social
networks and relationships. This concept can be applied to explain why grassroots suffered
from poor housing. It is because they do not have sufficient information to find flats. They do
not have sufficient time or friends to gain better access to housing market information.
Besides providing rental subsidy, social and housing services may help the housing poor to
obtain better housing by providing more information or better agency service.
324
7.6 ‘Housing Makes Me Poor’: Impact of Housing on the Non-Income Poverty Situation
Housing is very crucial in influencing the daily lives of individuals. It causes a critical
impact on the multidimensional poverty situation, including deprivation, social exclusion,
and subjective poverty. Moreover, residents frequently voiced concerns on physical and
mental health, seriously affected by housing circumstances.
7.6.1 Housing is one dimension of deprivation that is magnified: Impact of housing on
deprivation
The deprivation situation is usually connected with income poverty meaning that if
families earn less, they do have enough money to pay for necessities. However, housing
factors should not be ignored. The poor housing condition is a type of deprivation, including
indoor or outdoor problems. Moreover, due to a small living area, the deprivation situation is
worsened.
Air condition. Issues of air pollution and smell are usually overlooked when assessing
the quality of life of concerned residents. As Case A mentioned, despite living on the seventh
floor, the smell from the restaurant on the ground made her uncomfortable.
Case C, male, 69, subdivided flat, echoed the air condition issue. Ventilation was so bad
that he was affected by the toilet smell of the neighbour while eating. He said, ‘The
ventilation is extremely poor here, and the air pumps from the neighbour’s exhaust fan affect
our trachea and eyes. The toilet is very near to the eating place, and hygiene is really bad.’
325
Bedbug. In subdivided flats and cubicles, being disturbed by insects was critically
troublesome to residents. Besides mosquito and rats, bites by bedbugs were quite common
and made residents lose sleep.
Case K, male, 54, cubicle, stated, ‘I cannot sleep well and keep itching every night ...
When you switch off the light, the bedbugs come out, but you cannot turn on the light for the
whole night.’
No place for furniture. The housing is very small to place basic furniture. As Case B
said, she had no television and refrigerator due to the limited living space.
‘I have not watched television for a long time … and I do not even have a refrigerator
because I cannot go inside my home if I get them. Despite having found a refrigerator
costing $300, I have no space to put it.’ (Case B, Female, 59, subdivided flat)
For the singleton Case K, male, 54, cubicle, living in a 30 square feet cubicle, the
situation was worse. He could just put a fan in his home. He said, ‘The space is only around
20–30 square feet. I can only put a single bed here. That is all. Or I can just put a fan behind.’
Difficult to cook at home. Some grassroots families choose to cook at home to save
money from eating outside, but the housing is very small for them to cook. For example,
Case Q did not have a kitchen at home and could just use an electric cooking device. Some of
them can cook, but they had to cook in the toilet.
326
Case Q, female, 38, from subdivided flat to public housing, said, ‘I do not have a kitchen,
but I cannot always go outside to eat, it is very expensive, and I want to cook for my child.
So, I can only buy a small electric cooker to cook here.’
7.6.2 Is housing related to social exclusion? It depends. The impact of housing on
social exclusion
The impact of housing on social exclusion is ambivalent. Some people are socially
excluded because the housing cost is very high, or the location is very far, and they feel
embarrassed to meet friends. However, others presented their social life is not affected
wherever they lived.
Housing cost is too high. Several residents reflected that social activities were very
expensive for them, as they already spent too much on rent.
Case A, female, 28, subdivided flat, said, ‘I do have not enough money to join social
gatherings. Even when my friends call me to go out, I reject them and stay at home to save
money.’
When you spend a lot on housing costs, you have lesser savings. However, sometimes my
daughter wants to take a short trip and the savings are gone. How do I then pay for the
rent in the coming months?’ (Case S, female, 35, subdivided flat to public housing)
Housing location. Some residents living in remote areas need more time to travel to
urban areas, and the transportation fee makes them hesitate to go out. For example, the
327
children of Case F want to go out to study and play, but it is very costly for them.
‘Living in Tuen Mum, it is not convenient to go to the urban area, but my son usually
wants to go to the library in the city centre. Then I need to spend more than a hundred to
go back and forth, and it also requires a lot of time. So, it is very inconvenient.’ (Case F,
Feeling shame and sad to meet friends. The self-image of residents may be affected by
their housing situation, which further influences their social life.
‘It is difficult to make friends here. I seldom chat with my neighbour. We do not open our
door to talk. We have chatted only once to twice in these three years. I have not seen
anyone visit her in these years.’ (Case H, Female, 71, subdivided flat)
Despite moving to public housing, some of them think they belong to the lower class
and are looked down upon by the rich.
‘The mothers of my child’s classmates are arrogant and think they are rich, as they live in
private housing … living in a public house means being from a lower class, or residents
with a lower status …’ (Case S, Female, 35, subdivided flat to public housing)
Exception: Social life is nothing related to housing. Several factors affect the social
interaction of residents. For example, some residents gain support from their neighbours and
friends living nearby in a similar housing situation.
328
‘My neighbour sympathised that I am poor and was initially afraid to talk to me.
However, later, we became friends, and she now volunteers to take care of my children when
I am busy’, as stated by Case A, female, 28, subdivided flat.
These cases highlighted that despite living in a poor situation, they can make friends with
a similar background. They may have limited resources for joining social activities but can
meet friends via other means.
7.6.3 ‘I do not want to tell others where I live’: Impact of housing on subjective
poverty
Subjective poverty concerns self-assessment of the poverty level and relative status of
their friends and neighbours. The housing status of residents crucially affects their self-image
and self-evaluation of the level of their life quality.
I feel embarrassed to live in such a house. Living in subdivided flats makes residents
feel embarrassed and they resist telling others where they live.
Case H, female, 71, subdivided flat, said, ‘I do not want others to know where I live,
but if they want to know, I tell them. It is not my fault that I live here. However, I do not want
to disclose it.’
Case S, female, 35, subdivided flat to public housing, said, ‘If I tell others I live in a
subdivided flat, people will think of a shared toilet, dirty place, inconvenience … and I care
about what others think.’
329
I am poor because I live here. Residents lose their dignity for living in poor housing and
feel poverty when they face difficulties in paying rent.
As Case L, male, 60, subdivided flat, stated, ‘Living here, I already lost my dignity, didn’t
I? You tell others you are living in a subdivided flat, it already implies your status, doesn’t it?’
Case A, female, 28, subdivided flat, stated, ‘The landlord urges me to pay rent, but I do not
have enough money. I need to borrow money from my friend. I feel poor when I am living
here.’
Although most of them urge for public housing, it may not change their image of living in
poverty. Some residents said, as the mean-test public housing policy, it implies most of the
residents in public housing are poor families.
Compared with others, whether good or bad? Interestingly, subjective poverty may not
directly be coherent with the objective living situation of residents. Their self-assessment
may change due to comparison with others. For example, the subjective feeling worsens
when they compare themselves with those living in a better place.
Case F, female, 46, subdivided flat, stated, ‘My son would sometimes tell me his
classmate lives in an expensive apartment next to us and that he wants to live there too. I
would tell him to study hard.’
More residents may compare their situation with those in a worse situation. It serves as a
common coping strategy for the poor population to face difficult circumstances.
330
Case G, female, 35, subdivided flat, stated, ‘I have seen people living in a smaller area
and worse housing conditions than us in the news. I am satisfied after comparing my situation
with them, even though it is not good here.’
7.6.4 ‘Poor housing, poor health’: Impact of housing on physical and mental health.
The health dimension of poverty is overlooked in the literature review of the proposal;
however, in the interviews, informants showed sincere concern about their physical and
mental health situation when talking about the impact of housing.
Too small, too dangerous. Case B, female, 59, subdivided flat, attempted to cook to save
money but cut her hand in the kitchen, as it was very small. She said, ‘I cut my hand in the
kitchen, as it is too small … I was so afraid, I started crying. Blood was flowing and I was
running on the street … It was horrible.’
Cannot sleep well. Another common problem is that residents cannot sleep well because
of the living environment and the noise from indoor and outdoor. For example, residents lived
very close to neighbours and easily woke up by their noise. Some parents, wanting to give an
independent flat to their children, chose to sleep on the floor but experienced bad quality of
sleep and further worsening of emotions.
‘I want to leave space for my daughter to study, and so, I sleep on the floor in the sitting
room … I cannot sleep well, and this makes me emotional and irritable’, said Case D, female,
50, subdivided flat.
331
Easy to get sick. People found they easily fell sick when living in a poor environment.
The situation of Case T was bad. He got into a quarrel with his flatmates because of noise and
was forced to sleep outside, but when he woke up, he got a stroke in the cold weather.
Case T, male, 45, singleton hostel, said, ‘My roommates were playing and gambling at
midnight. I could not sleep and asked them to keep quiet … We quarrelled, and I went down
to the street to rest … It was very cold outside, and I got a stroke …’
Bad mood and poor mental health. The poor living area also caused much pressure on
the members. Persons found themselves in a depressive mood after moving from a large
house to a small flat in Hong Kong. Many cases developed a depressive mood after being
evicted or moving to another flat with a smaller living area or high housing cost; Case D is
one example.
‘Living here makes me depressed and puts me in a bad mood … It was completely
different when I was living in Shenzhen. I did not feel depressed here’, said Case D, female,
50, subdivided flat.
7.6.5 Short Discussion
The stories of informants show that housing influences their poverty situation in many
dimensions, including deprivation, social exclusion, and subjective poverty. Moreover, many
residents are concerned about the impact of housing on their health. This result is one of the
key findings of the interviews. Researchers initially did not intend to explore the health
332
situation of informants, but many of them shared their concerns when talking about housing.
Impact of Housing on Deprivation
The mechanism of the impact of housing on deprivation was revealed by the
individual interviews. Many cases had no housing choice because of their low income and
were forced to live in subdivided flats, normally with a small living area and different indoor
housing problems. Most residents do not have sufficient place for their necessities. For
example, some do not have washing machines because the flat was very small. They needed
to wash clothes by hand instead. The informants living in cubicles faced an even worse
situation. Their most essential items could not be placed in their flat because the living area
was only about 30 square feet. This finding demonstrated how a small living area is linked to
deprivation. Besides, informants living in a subdivided flat usually face different housing
environment problems, such as poor air conditioning, insects, rats, and poor hygiene. These
housing problems also caused deprivation in the views of residents. This result is coherent
with the finding of the quantitative analysis and previous studies (Guio & Maquet, 2007;
Napiorkowska-Baryla, & Witkowska-Dabrowska, 2018). The stories of residents extend the
previous work by showing how deprivation occurs in tiny subdivided flats in Hong Kong.
People suffered from deprivation not only because of the lack of money to buy stuff but also
because of the lack of space for essential items at home.
In terms of housing cost, the result of the regression analysis shows a significant but
333
small impact of housing cost on deprivation. The result of structural equation modelling
(SEM) analysis proposes that the impact of housing cost on deprivation is mediated through
the living area and indoor housing problems. Qualitative interviews show that informants
earned only a little income, and some of them could not spend much on housing costs after
paying their other daily necessities. Some families attempted to squeeze the housing cost to
save money for their children. Therefore, the amount of housing cost may not alone reflect
the deprivation situation faced by residents. However, housing cost is a crucial concern for
most residents. This finding echoes previous studies that highlight the influence of housing
affordability on deprivation (Daniel et al., 2018; Warren, 2018).
Housing type is an important factor in deprivation. The SEM model reveals that the impact of
the housing factor is enlarged for private housing residents but does not show how this effect
occurs. The informants allocated from subdivided flats to public housing demonstrate the
positive effect of changing housing type. After moving to public housing, they paid less
housing costs and had a larger living space. Both factors contributed to reducing the risk of
deprivation. This finding is consistent with the argument of Borg (2015) that the enlargement
of the private rental market worsens the deprivation situation. On the contrary, for housing
location, no significant relationship with deprivation is found in the interviews.
Impact of Housing on Social Exclusion
For social exclusion, in the quantitative analysis, most housing factors were
334
associated with social exclusion, but the strength of the impacts was insignificant. From the
qualitative interviews, the impact of housing factors on social exclusion was found to be
ambivalent, and the mechanism was found to be complicated.
In terms of housing cost, certain cases showed that the housing cost was a large
burden to them and attempted to limit expenses on social activities. Besides, they do not have
much spare money after spending on housing costs and daily necessities. This condition
limits their social lives. Although financial constraint limits the social activities of certain
cases, many other cases were not remarkably affected. Instead, the level of social interaction
and participation was more influenced by the personal characteristics and demographic
background of the informants. For example, young females were more willing to seek help
and actively make friends with others, but the males and the elderly were more reluctant to
meet others and were easily socially excluded. This result is similar to the quantitative
analysis that shows that social exclusion is more affected by demographic background than
housing factors.
Housing location is not a crucial factor in social exclusion for most informants. Most of
them make friends and build a social network in their workplaces or local communities.
However, the social life of informants who live in a remote area away from the city centre is
also affected. For example, one elderly who lives in a rural area in Tuen Mun needs to spend
half an hour walking from home to the city centre. She is not eager to leave home to meet
335
friends or join social activities in the community centre; thus, she is more socially excluded.
The impact of housing location on social exclusion was more serious for the elderly group
because they were more reluctant to travel far away from their local areas. The housing
location is divided into only three levels in the quantitative analysis. This approach may not
demonstrate the influence of housing location for those living in remote areas (Wong, 2011).
The qualitative cases supplement this limitation. On the contrary, few cases mention that
living in subdivided flats made them reluctant to meet friends. The housing type also
demonstrated an impact on the social life of certain residents. This finding was not clearly
indicated by the quantitative analysis. Generally, the qualitative interview data showed
consistent results with the quantitative analysis.
Impact of Housing on Subjective Poverty
In terms of subjective poverty, housing type is the most important housing factor in
terms of their perception of the poor. Several informants feel shameful for living in the
subdivided flat. The housing type implied specific meanings for residents. In Hong Kong,
certain informants living in subdivided flats simply indicate living in poverty. The majority of
poor people were forced to live in subdivided flats, given that the waiting time of public
housing was increasing as well as the rent of private housing. Moreover, the public, including
the media and political parties, usually used subdivided flats as a symbol of poverty to raise
public awareness of the problems. This finding may explain why those living in subdivided
336
flats felt poorer. In many cases, subdivided flats are a symbol of a small living area with a
poor housing environment. This finding echoes the result of quantitative analysis that
suggests that living density and indoor housing problems are significantly associated with
subjective poverty.
On the other contrary, the comparison was one of the main causes of being subjectively
poor. Similar to the previous discussion about subjective poverty, informants felt poor by
comparing themselves with their significant others (Foye, 2017; Liao, 2009). In the
interviews, several informants shared they feel poor because they compare their living
situation with that of their friends or relatives. They feel poor when compared with others
who own a house or live in a large house. Moreover, certain informants compare their living
situation with other rich people in Hong Kong. The housing inequality was serious in Hong
Kong, the housing of which is one of the most expensive in the world (Demographia, 2017).
The real estate industry was well developed in the country. Many news and advertisements
were about luxury housing and speculation of housing every day. People easily compare their
living situation with others in better housing. A comparison with someone who lives in large
and luxury housing or someone who owns houses with extremely high value may make
people feel poor. The increasing housing cost and number of subdivided flats may widen the
gap between the rich and the poor, and it also increases the risk of subjective poverty of
grassroots.
337
Dynamics of Housing Choice and Interaction among Housing Factors
The quantitative analysis focuses on the impact of individual housing factors on the non-
poverty situation. The SEM models attempt to study the interaction among housing factors.
Nevertheless, they cannot reveal how residents choose among different housing elements,
including housing cost, living area, and location. The qualitative interviews fill this gap. The
relationship among housing elements is non-linear and varies by the preference of the
residents. Most cases initially consider their income and housing cost to determine how much
they can afford for renting flats. However, many low-income families cannot afford adequate
housing with reasonable living space and environment. They need to squeeze the housing
cost and ‘choose’ flats with a small area and basic environment. That is a possible reason
housing cost was not directly associated with deprivation and social exclusion. For housing
location, many informants initially consider the community they are familiar with, such as the
district where they lived before or where their children studied. The rent difference among
districts is not their concern. Moreover, the numbers of informants did not show much
concern about the living location. For them, other housing factors such as living environment
was more important to their daily life. This also echoed the quantitative findings in which
living location was not associated with non-income poverty and it also attempted to explain.
The availability of community facilities is also not a concern of the informants when finding
houses; it is normally the result of their choice. The housing choice of residents and
338
interaction among housing factors were complicated and beyond the scope of this thesis.
Further research and theoretical discussion are needed in the future.
Housing is a Core Factor of Multidimensional Poverty
In general, housing influences livelihood in multiple aspects, including physical, mental,
material, social, and subjective. Physically, the small living area and noise cause poor sleep
and induce potential physical harm to residents. Mentally, the high housing cost and dense
environment generate high stress and tension for members. Materially, poor housing is a type
of deprivation in terms of dimensions of air ventilation, hygiene, and infrastructure of
buildings. Socially, residents may hesitate to meet friends or join social activities because
they have too little spare money or live far away from the city centre. Subjectively, the self-
assessment of being poor or self-image is influenced by the housing types of people despite
their income level. The housing condition can create a negative image of the residents and
hurt their self-esteem. Besides income, housing factors demonstrated a critical impact on
various types of poverty situations. These findings contributed to the conceptualisation of
multidimensional housing poverty, which highlighted the importance of the roles of housing
in the analysis of multidimensional poverty.
339
7.7 ‘Life is Full of Struggles’: Response, Struggle, and Coping Strategies of Residents
How did the residents respond to different critical events and the ongoing housing
poverty situation, or how did they cope with them?
7.7.1 Resistance: Useful versus Useless
The social actions are useless, and I did not want to participate
The residents were mostly under a poor living situation; however, many of them felt
hopeless about changing the current situation. Case B joined an action group of an NGO for
several times, but she said the change was too far away from her.
‘In the centre, the workers always invite us to voice out but, you know, the landlord may
kick you out if you voice out … Sometimes, the worker called me to meet and I came.
However, I laughed at them sometimes…’ said Case B, female, 59, subdivided flat.
Case N, who met the legislative councillor and the chief executive in the previous social
action, also felt hopeless to see policy change.
‘No, we are small potatoes. You can chat here, but no one is listening to you. You cannot
resist. Go for a petition? Not my business. We are helpless. I have seen the chief
executive Carrie Lam and have gone to the legislative council before, so what?’ (Case N,
I want to make change and actively participate
Not many of the residents stood out to attempt to make a change, but there were
340
exceptions. Case N and Case O joined the social actions organised by NGOs and social
workers, and they have seen policy change in the past.
‘I joined the petitions every time. I shouted, “more public housing” in each action. Last
Saturday, you see, there were some new measures for the poor relief … Sometimes, the
government may respond…’ said case N, female, 72, subdivided flat.
‘The government may respond to the demands sometimes. If the public raises an issue,
the government needs to react, depending on more or less. They need to respond to a certain
extent’, said Case O, male, 63, cubicle.
Participating is a relief
Joining social groups for advocacy or social actions not only provides policy change but
also gives relief or social support to the participant. Some of the residents found it was
difficult to make a real change, especially in a short period, but they attempted to voice out.
Case Q said she needs to come out and tell others about the living situation of the poor, at
least to let the public know. She found voicing out the problems to give her relief.
‘I think we must voice out. It is better than keeping silent at least. Who can understand
you if you do not say anything? And sometimes. I think we cannot control how the others
respond, but at least we feel relieved after voicing out. No one will respond to you
immediately, isn’t it?’ (Case Q, Female, 38, subdivided flat to public housing)
341
Case P, male, 42, subdivided flat, echoed similar thoughts. He was quite relieved after
voicing out the housing problems and felt at home in the action group, as residents were
facing similar housing situation and had the same targets of advocacy. He said, ‘I was much
happier after I joined the action group … at least I can vent here … release my pressure … it
is a good way to relax…’
7.7.2 Resilience: To adapt versus Not to adapt
Facing a poor housing situation, some of the residents chose to adapt to the living
condition, even if it was difficult to do so. But the stories behind ‘adaptation’ were
complicated. Did they feel good? Or was it some kind of suppression? How could they adapt?
Attempt to adapt
Case A, who claimed her family was living in deprivation but always told herself it was
not a problem and tried to get used to it, sometimes felt helpless.
‘We have no choice. I need to get used to this … it is helpless, to be frank … after many
years, sometimes, I tell myself not to get depressed for this … There is no alternative, but
we need to continue living, right? We are already adults, we need to face the reality …’
(Case A, Female, 28, subdivided flat)
Case D mentioned she tried to adapt to the situation because she had no choice,
however, she found that it may lead to mental illness. She felt under high pressure even kept
telling herself to adapt to the small living area and poor environment.
342
‘Everyone needs to get used to the environment, right? You should feel uncomfortable in
the beginning…Sometimes I guess I get depressed because of this…Why I should live in
this situation now? It seems I lost everything, no one knows what I feel and how sad I
am… but…I adapted, I adapted…’ (Case D, Female, 50, subdivided flat)
Too horrible, I do never adapt
S Some residents were ambivalent about their situation. For example, when chatting
with Case E, who lived in a very deprived situation, she mentioned she used to live in a poor
situation and felt hopeless. She then talked about her daily life in more detail. She said that
the poor living condition was too bad to adapt to.
‘I will never forget that horrible living experience, cannot forget. You know, in the
evening, many mosquitos would fly around you and bite you. The hygiene was so bad,
and I was afraid of being infected; it was dreadful! Too terrible to adapt to.’ (Case E,
Female, 63, cubicle)
Case T had a similar experience. He once lived in bed space shared with more than 30
singletons, and it was too insecure to live there. He claimed no one can adopt to that type of
living situation. The so-called ‘adaption’ for him was a sense of helplessness because he did
not have other choices.
‘I cannot stand it anymore! Who can adapt to this living environment? No one. There
were different issues and accidents every day. There was a fear of things getting stolen or
343
broken. The neighbours came to use your private stuff … I was just helpless. What can I do?’
(Case T, Male, 45, singleton hostel)
7.7.3 Future: Hopeful versus hopeless
The interviewer asked the residents whether they have hope or expectations for the future.
They answered in different ways.
Several cases said they did not have any hope. They did not want to be disappointed, as
the reality was very harsh. More expectations, more disappointment. Thus, they tried to give
up hope and live in reality.
Case L further elaborated that she was not the only one without hope; the next generation
too felt hopeless. She blamed the government project, which proceeded too slowly, further
stating that people needed to wait for several years for new housing and more than 10 years
for the land. Moreover, the housing price was too high and kept increasing, so she thought her
children would be unable to afford it after growing up.
‘You see, policy and planning in these years are talking about seven, 10, and 20 years of
planning. I will die before the projects are finished and allocated to public housing. How
will the new generation survive?’ (Case L, Male, 60, subdivided flat)
When the residents shared their hopes, most of them were concerned about the housing
situation initially. Cases H and T shared the only hope was getting public housing soon.
344
‘A better housing is my only hope at this moment. I do not think of any other aspects. I
only expect to be allocated public housing this year, that’s it’, said Case H, female, 71,
subdivided flat.
‘I do not have hope, what can I expect? I need to move out if the landlord wants to kick
me out. So, I can only wait for public housing; it is the happiest thing’, said Case T, male, 45,
singleton hostel.
Informants react differently to the current living situation. Interesting observations reveal
the concern of residents. The reaction of the resident depends on their background, living
experience, and nature of issues.
Adaption versus resistance. Adaption is one of the common coping methods of
residents. They asked themselves to adapt to the poor housing and poverty situation because
many of them believed that changing their current living situation was difficult. In particular,
those who experienced poor living situations tend to adopt the current housing and poverty
situation, even with satisfaction. However, adaption is a dilemma. If everyone could easily
adapt to a poor living environment, then society would have no voice, and the overall living
standard would decline continuously. However, the living condition may not change in a
short time. If residents continue to complain or get angry about the situation, then their
mental health might be affected. Someone can adapt to the situation personally but still resist
345
the current situation socially. Some informants react similarly. However, residents cannot
easily voice their opinions or join advocacy activities. Thus, the role of community or social
workers is important. They need to shift their role from social control to social change. Social
workers may provide social service and support to the poor, but they can also line up
residents for policy advocacy.
Level of adaption and minimum living standard. Some informants shared a situation
of extremely tiny living areas, poor hygiene, and a high level of rent. This finding reminds us
of the rationale for setting a basic living standard. If the market mechanism fails to provide
adequate housing for the general population, then the government should intervene. Similar
to the legislation on minimum wage, we cannot tolerate a certain level of wage, which is too
low for a living. Nevertheless, whether the government should tighten the regulation of living
space was debatable, for example, a lower limit of living area per capita was set for a
subdivided flat, setting a basic living standard is essential to protect the housing right of
residents. However, some argued that this condition may reduce the number of subdivided
flats and increase the rent. The setting of the minimum living standard should complement
rent control and tenancy security protection to avoid rent increase or eviction after
enhancement of the living environment.
Hopeful versus Hopeless. Being hopeless is a common reaction of informants in facing
the poor housing situation because they cannot foresee a change in a short period. However,
346
for families with children, the informants were usually hopeful in the future; they do not
focus on the current situation. They focus on their children’s wellbeing, including education,
social activities, and living environment. Moreover, the parents are more motivated to
improve their living standards than singletons. Children instil hope in families that face
poverty and hardship. Certain informants wait for their children to get a job to improve their
living, even when they need to wait for more than 10 years. Other families without children
normally do not expect an increase in income but wait for the government to enhance the
current situation.
347
7.8 Conclusion of Qualitative Analysis
In this chapter, the stories of residents revealed the mutual impact of housing and poverty
on their daily lives. The findings of quantitative analysis, which only provided a static view
of the housing poverty situation, were broadened. The interviews revealed that the residents
fell into poverty because of different critical life events, such as migration, marriage, divorce,
or unemployment. Their housing choice was very limited, not only because of their low
income and lack of social capital and information but also due to the loose regulation of the
housing market and poor welfare protection. Nevertheless, they struggle with housing cost
and quality; some of them limited their housing cost and lived in houses with poor conditions.
This finding answered the questions raised by the quantitative analysis.
Low-income families were forced to live in subdivided flats, with small living area and
numerous housing problems. Furthermore, their stories demonstrated how poor housing
circumstances further influence their non-income poverty situation, including deprivation,
social exclusion, and subjective poverty. This finding supplements the quantitative analysis in
addressing the impact of housing elements on the non-income poverty situation. Nonetheless,
the social exclusion and subjective poverty situation of individuals are influenced by their
style and life history. Moreover, the qualitative findings supplement the quantitative analysis
in studying multidimensional housing poverty. The mechanism of how housing factors
influenced deprivation, social exclusion, and subjective poverty was revealed by the stories of
348
cases; however, this mechanism needs to be further studied for more detailed investigation. In
summary, the life path of subdivided flat residents fell into a poverty spiral, and the residents
suffered from the continuous mutual impact of housing and poverty. Most of them wait for
public housing or for their children to grow up; this gives them hope to live. They are
struggling with adaption and resistance and hopefulness and hopelessness every day. To
enhance the situation, progressive change in housing and antipoverty policy is required.
349
Chapter 8 Conclusion
In recent years, the concern over the worsening problems of housing and poverty in Hong
Kong has been growing. However, the dominant income/poverty approach in defining and
conceptualising poverty has failed to show the suffering of the poor, particularly the impact
of the high-cost housing and deteriorating non-income poverty situation.
First, this thesis uses three housing poverty lines and census data sets to examine the
change in the housing poverty situation in Hong Kong from 1996 to 2016 to highlight the
worsening problem faced by the residents of rented private houses. The increase in housing
prices and rentals in Hong Kong mostly influences residents living in private rental housing,
new couples, and those who aim to start new families. Therefore, this analysis further looks at
the situation from the perceptions of ‘housing problem’ by the poor and ‘housing profit’ by
the rich. The establishment of official housing poverty lines facilitates the monitoring of the
housing poverty situation and evaluating the effectiveness of housing policy. These will be
crucial for solving the housing problem of the poor.
Second, the impacts of housing factors on non-income poverty were assessed using the
SPPR data set. By using logistic and multiple linear regression methods, living area per capita
and indoor housing problems were found to cause the most significant impact on non-income
poverty. Moreover, these two factors also showed the mediating effects amongst income and
non-income poverty in the SEM analysis. Amongst the non-income poverty, subjective
350
poverty was the most influenced by the housing factors, and social exclusion was the least
impacted. This result revealed the importance of housing in affecting the subjective
assessment and well-being of residents.
Third, the interaction between housing situation and poverty was revealed by interviews
with residents living in subdivided flats. The qualitative data showed that these residents fall
into poverty because of various critical life events, such as marriage, divorce, unemployment,
and migration. With a limited choice of housing, the residents were forced to squeeze their
cost of housing and live in small flats with numerous housing problems. Additionally, these
housing problems deepened their poverty causing deprivation, social exclusion, and
subjective poverty. Many residents experienced a ‘poverty spiral’ and faced hardship and
struggle in their everyday life.
This section presents the answers to the research questions (Section 8.1), highlights the
theoretical contributions (Section 8.2) and empirical significance of this research (Section
8.3). The personal reflections are highlighted (Section 8.4) to illustrate the concern and
visions of the researcher. These are followed by a discussion of the limitation of the
quantitative and qualitative study (Section 8.5), and suggestions for further research (Section
8.6).
351
8.1 Conclusive response to the research questions
In this research, four main research questions are proposed. After the data analysis and
discussion, the conclusive responses are as follows.
RQ1. What is the housing poverty situation in Hong Kong?
The census data from 1996 to 2016 reveals the housing poverty situation by the
construction of different ‘housing poverty line’. Generally, the ratio of the housing poverty
rate to the overall population decreased from 2006 despite the increasing price and rent of
housing. The increased proportion of households that finished the mortgage payments was
the main reason that contributed to this situation. However, the housing poverty situation of
private rental residents has been seriously deteriorating over the past 10 years. One of the
reasons is the shortage of the supply of public housing, which results in prolonging the time
the residents have to stay in the private rental market.
The housing poverty lines help in assessing the poverty situation of various groups of
housing types. For example, the use of the ‘ratio housing poverty line’ showed that private
rental housing residents are is the worst suffering group, and the situation has been
deteriorating during the past 10 years. The use of the ‘residual housing poverty line’
highlighted the condition of those who lived in public rental housing, and privately-owned
housing, for whom, although the housing cost is lower, affording a basic living standard is not
possible.
352
The ‘housing poverty’ analysis reminds us that the ‘housing problem’ of certain groups
in the population is the ‘housing profit’ for other groups. House owners, particularly those
who have completed mortgage payments, benefitted from the increasing housing prices and
rents. By contrast, the low-income private rental residents, and those who want to start a new
family, such as new couples, suffer from the ‘housing problem’ in Hong Kong.
Overall, the ‘housing poverty’ measurement enhances the understanding of the poverty
situation of Hong Kong residents because the housing cost forms a very large portion of their
expenses. The income approach, which was adopted by the Hong Kong government, may not
reveal the housing cost burden faced by the poor families who were under the ‘housing
poverty’ line.
Table 8-1 summarises the merits and limitations of the income poverty line and various
housing poverty lines with reference to the principles of poverty line suggested by the
Commission of Poverty (HKSAR government, 2019) and poverty approach comparison
framework proposed by Laderchi, Saith, and Stewart (2003).
Table 8-1 Comparison of poverty lines and their application in Hong Kong
Income poverty Ratio Housing Residual Housing

line poverty line housing poverty poverty (30/40)
line
Ready Easy to Relatively easy; Measurement of Relatively easy;
measurability measure depend on basic living depend on
definition of standard is definition of
housing cost arguable housing cost
International Easy to Easy to Hard to Relatively easy
353
comparability compare compare compare basic
living standard
Regular data Normally Normally data less Normally
availability available available regularly available
collected, but
could
easily be
improved.
Amenability to Easy to Easy to A bit Concrete
compilation and understand understand complicated to measurement
interpretation understand for the poor
Major Not count Count those Definition of The threshold is
conceptual expense choose high basic living arbitrary
weaknesses housing quality standard is
or wealthy unclear
house owners
Application in Hong Kong
Sensitive to Cannot count Sensitive to Sensitive to Sensitive to
housing poor those with high private housing groups suffered low-income
housing cost residents from high families with
housing cost high housing
cost
Assist in policy Focus on Good at assess Good at finding Good at
formulation or monetary policy affordability out the targeting the
Assess policy stress underprivileged poorest group
effectiveness after paying
housing cost
Major Not count those Not count those Not easy to The threshold is
limitation with high income poor define basic quite high and
essential without housing living threshold overlook other
expense cost underprivileged
RQ2. How are the interactions between housing situations and poverty in Hong Kong?
This study is also focused on the impact of housing factors on non-income poverty
situation. The SPPR data set was used after controlling the demographic factors and income.
The housing factors reveal different levels of impact on deprivation, social exclusion, and
354
subjective poverty. Amongst the housing factors, indoor housing problems have the strongest
impact on non-income poverty, whilst the impacts of living location are not significant in
various models. Amongst the three types of non-income poverty, subjective poverty is most
affected by housing factors, whilst social exclusion is the least influential factor.
The SEM model analysis illustrates the impact of housing on income and non-income
poverty. Amongst the three housing factors, housing cost, living area, and indoor housing
problems; the living area is found to be the common mediator between income and non-
income poverty. By contrast, the mediating effects of housing cost on income and non-
income poverty are not significant (Figure 8-1).
Figure 8-1 The mediating impact of housing factors on non-income poverty
Note. * = fully mediated; V = partially mediated; X = not mediated.
355
RQ3. Why does housing poverty exist, persist, or change in Hong Kong?
The quantitative result shows that the ‘housing poverty’ situation in Hong Kong has
persisted for many years. The circumstance cannot be explained by any single reason.
Numerous political-economic factors underlie it. Hong Kong is a global city with a rapid flow
of capital. Under the influence of globalisation and neoliberalism, and the financialisation of
housing catalogues, prices and rents of housing have been increasing (Fernandez & Aalbers,
2016; Harvey, 2005; Smart & Lee, 2003; Yip & LaGrange, 2006). At the policy level, the
uneven distribution of land, privatisation of housing, lack of public housing, and loose private
housing market regulations are all the factors that are worsening the ‘housing poverty’ situation
for low-income families. At the cultural level, the speculative culture of housing in Hong Kong
also contributes to the ballooning housing prices (Chan, 2000). The public, which includes
even poor households, generally believe that buying a house is a crucial means of capital
accumulation or retirement protection. This situation keeps up the demand for homeownership.
At the social and political level, continuing social conflicts in recent years have been related to
housing and land issues in Hong Kong. The residents of subdivided flats, NGOs, social workers,
and legislative councillors raised public concern over the subdivided flats and poor housing
situation faced by low-income families. The government responded to some of the concerns by
such measures as building social housing and increasing the supply of public housing. However,
the current effort by the government is quite insufficient for alleviating the housing poverty
356
situation.
RQ4. What is the experience of people living in poverty and housing poverty
situations?
The qualitative individual interviews further reveal the interaction between housing
and the poverty of residents (Figure 8-2). Families trapped in the income poverty situation
said that the causes were various life events, such as marriage, illnesses, migration, or
unemployment. Apart from the low income, poor households also lacked social capital and
market information, which further limited their choice of housing. Shortage of public housing
and inadequate regulation of the private rental housing market, low-income families are
forced to live in subdivided flats characterised by costs, poor quality of housing, and many
problems in their living environment. Deficiencies in housing reveal their critical and
negative influence on the situation of deprivation, social exclusion, and subjective poverty in
which these households find themselves. From the interviews, the utmost influence on the
situation of deprivation was exerted by the living area because the flats were quite small and
could hold very few possessions of the occupants. Residents were socially excluded not
because of the location of their houses but by the extremely high housing costs as well. They
had to limit their social activities to save money to pay the rent. The subjective poverty was
reflected in the shame that the residents felt for living in subdivided flats. They often
compared themselves with others living in large flats and felt ashamed of having to live in
357
poverty. Moreover, residents often compared their current housing situation with that in the
past. The intensity of the feeling of subjective poverty depended on whether they saw their
current living situation to be improving or worsening. Overall, the households are trapped in
a ‘poverty spiral’, a process set in motion by the reciprocal impacts of housing and poverty.
To get out of the trap, the majority of the households are waiting for the allocation of public
housing and for a better environment for their children to grow up in.
Figure 8-2 The common path of housing poverty
8.2 Theoretical Contributions and Significance
This thesis enriches the theories and concepts of poverty in a housing perspective and
broadens the understanding of housing poverty. The literature review indicated the theoretical
gaps in for this research to fill, namely, (1) disregarding the housing perspective while
conceptualising poverty, (2) missing the contribution housing factors in explaining various
non-income dimensions of poverty, and (3) the process of the interaction of poverty situation
358
with the housing factors. This research aimed to fill in these theoretical gaps.
In the past, the housing perspective was overlooked in poverty study, and the
conceptualisation was dominated by monetary or income approach (JRF, 2015; Saunders,
2017; Stephens & van Steen, 2011; Thalmann, 2003). In the current study, the concept of
‘housing poverty’ is presented and applied in analysing the poverty situation in Hong Kong.
The result and discussion demonstrate the importance of housing perspective in poverty
study. The financialisation and privatisation of housing have become a serious problem
throughout the world, particularly under the influence of globalisation (Sassen, 2011;
Valença, 2015). The housing cost has been rising and the housing environment has
deteriorated. The current poverty situation can no longer be explained by the theories of the
past that disregard housing. The mainstream definitions of poverty, such as low income,
deprivation, and social exclusion, are insufficient to describe and explain the suffering of
families and individuals. This study, which uses quantitative and qualitative analysis,
illustrates the use of the concept of ‘housing poverty’ to understand, describe, and explain the
difficulties faced by the underprivileged sections of the society.
This research also investigated the idea of ‘multidimensional housing poverty’.
Accordingly, distinct types of housing-induced poverty are revealed (Kutty, 2005; Stone,
2006a), including housing-induced deprivation (Daniel, Baker, & Lester, 2018; Guio &
Maquet, 2007), housing-induced social exclusion (Arthurson & Jacobs, 2003; Munch, 2012)
359
and housing-induced subjective poverty (Clapham, Foye, & Christian, 2018; Foye, 2017). In
quantitative analysis, the result showed that material factors attaching to housing, such as
housing environment and living area, have substantially more impact on non-income poverty
than other housing factors, such as location and the cost of housing. Moreover, subjective
poverty is considerably more influenced by housing factors than by social exclusion and
deprivation. Additionally, the current research broadened the mainstream understanding of
housing, which focuses on the monetary and economic dimensions. The social function and
subjective understanding of housing also are important for the livelihood of residents. The
results from SEM analysis also reveal the complicated interaction amongst housing factors.
The housing factors not only impact non-income poverty but also interact with each other.
This finding has enriched the original theoretical framework for studying the interaction
between housing and poverty (Figure 8-3). Given that Hong Kong is one of the global cities
with prominent levels of financialisation, urbanisation, and capitalisation of land and housing,
the above new findings from this research in the housing poverty provide a comprehensive
framework for explaining various dimensions of poverty in other cities with a similar
background as Hong Kong.
360
Figure 8-3 Final framework of significant relationship amongst housing and poverty situation
This research also examines the role of housing in the setting in motion of a ‘poverty
spiral’ for families and individuals. From the SEM model analysis, some housing factors have
been determined as the mediators between income and non-income poverty. From qualitative
individual interviews, the stories of residents show how their poverty situation has worsened
owing to poor housing. These results highlight the interaction between housing and poverty
in a global city with high housing costs and bring to light the struggles of the underprivileged
who have limited housing choices. The housing circumstance affects the situation of
deprivation faced by the residents but not the level of their social exclusion. The subjective
poverty situation of residents depends on the housing situation and their life history and
relatives. This research provides direction for further research on ‘multidimensional housing
poverty’ and the role of housing in the ‘poverty spiral’.
361
8.3 Empirical and Policy Contributions and Significance
To recap, this study attempts to fill several gaps in empirical research including, (1) the
‘housing poverty line’ has not existed so far in Hong Kong; (2) the impact of housing factors
on deprivation, social exclusion, and subjective poverty in Hong Kong has not been studied
so far; (3) the dynamics amongst housing poverty, housing pathway and subjective
experience of residents are the missing pieces in the literature on poverty in Hong Kong; (4)
there has been a lack of comprehensive discussion on the existence of the housing poverty
This research is a pioneer study for measuring the ‘housing poverty’ situation in Hong
Kong. Three ‘housing poverty lines’ are constructed based on the previous literature and
applied to measure the ‘housing poverty’ situation using census data from 1996 to 2016.
Moreover, this study compared the use of income poverty line and various housing poverty
lines to reveal their merits and disadvantages for describing the poverty situation in Hong
Kong.
This study is also the first research that examines the impact of housing factors on the
different non-income poverty situations in Hong Kong. Studies in Hong Kong that focus on
deprivation, social exclusion, and subjective poverty are limited. The current research
broadens the description of poverty in Hong Kong and explains the poverty situation from a
housing perspective. This research will assist the policymakers and the public to understand
362
the various dimensions of the suffering of individuals including the material, social, and
subjective aspects of the suffering.
The voices and stories of the residents of subdivided flats are included in this study.
Although numerous news reports on subdivided flats have been provided in recent years, the
existing theoretical framework is inadequate to analyse their situation. This study proposes
the ‘poverty spiral’ and ‘multidimensional housing poverty’ as frameworks on which to base
their living experience. This research finds that most of them were pushed into poverty by
various critical life events and further suffered from deprivation, social exclusion, and
subjective poverty owing to poor housing circumstances. They were continuously struggling
to cope with poverty.
This research is important for anti-poverty policy design and its implementation. The
Hong Kong government claimed it is concerned with the poverty situation and set the official
poverty line at 50% of the median monthly family income of 2013. The poverty line aims to
monitor the poverty situation and evaluate the effectiveness of poverty relief measures
annually. The ‘housing poverty lines’ proposed in this study facilitate the monitoring of the
poverty situation in a comprehensive manner, taking into consideration the housing costs and
types. The housing poverty lines will also assist in formulating an anti-poverty policy that
corresponds with a housing perspective.
By studying the impact of housing factors on the different dimensions of poverty, this
363
research provides an additional approach, apart from the income approach, for the
government to focus on the reduction of the level of poverty. For example, more emphasis
needs to be placed on the quality, living area, and cost of housing. The research on the
subjective experience and living histories of residents also provide further hints for poverty
eradication. By contrast, the political-economic explanations of the persistence of housing
poverty in Hong Kong enable us to trace the problems to their roots. Progressive reforms in
the housing market, land policy, and political system are needed to improve the housing
Overall, the following specific policies are suggested to alleviate the housing poverty
situation in Hong Kong:
■ Establish measurements of housing poverty lines and update them regularly to monitor
the poverty situation to implement targeted anti-poverty policy and evaluate the
effectiveness of housing policies. Residual housing poverty approach is suggested as it
measures how many people cannot maintain a basic standard of living after paying
housing cost;
■ Initiate large-scale and regular surveys to examine the non-income poverty situation of
Hong Kong residents, including deprivation, social exclusion, and subjective poverty;
■ Progressively increase the supply of public housing, particularly public rental housing,
with a stable living arrangement and low housing prices and rent;
364
■ Besides implementing rent control and strengthening the legal protection of rights of the
tenants, tightening housing market regulation, and the legislating a minimum living area
will effectively protect the housing for the poor;
■ Provide social housing with the regulated living area, housing cost and indoor facilities
to reduce non-income poverty caused by housing factors, and
■ Provide housing services, including the maintenance of the indoor and outdoor
environment, housing agency service, and neighbourhood supporting service.
8.4 Reflection
At the end of this thesis, I would like to reflect on the research design, research topic,
research process, limits of this research, and personal learning derived from it.
Research process. At the beginning of this research, I planned to study the relationship
between housing and poverty. However, the literature review revealed extensive literature and
numerous theories regarding housing and poverty, which rendered this topic too large to be
managed. Thereafter, it was decided to focus on the study of the impact of housing factors on
poverty. Given the availability of secondary data sets with a variety of variables on housing
and poverty, I decided to use the ‘SPPR’ data set for the core analysis of my thesis. Additional
data sets were likewise obtained from the Census and Statistics Department of Hong Kong,
which facilitated the expansion of the scope of this research to transform it from a snapshot of
2016 to a panorama of the period from 1996 to 2016. This expansion provided further
365
evidence to study the change in the housing poverty situation in Hong Kong and necessitated
an additional comprehensive discussion to explain the phenomena. Given the increasing
availability of secondary data sets worldwide, making good use of data appropriately is
necessary to facilitate the research process.
In the process of qualitative research, I gained an additional understanding of the unique
role that a researcher plays. As a social worker in NGOs that serve poor families and
individuals for years, this researcher found the situation of the informants familiar. In the
qualitative interviews, the researcher discovered that the informants were quite willing to
share their opinions that were different from the agenda, if any, of the NGOs. For example,
the informants may be inclined to say that they were poor when expecting social support
from NGOs or not to share their ideas on social policies that the NGOs did not support. When
they were informed that the interviews were confidential and that the interests of the
researcher did not conflict with those of the interviewees, the latter felt safe to share their
thoughts. Since the interviews often lasted for an hour, the informants had enough time to
comfortably share their ideas. Social workers lack this advantage because their working
environment is restricted by the constraints of resources and high service demand. These
constraints restrict the time social workers can spare for listening to their service users.
Research with the cases in housing poverty. Though the statistics may lack the
subjectivity of social agencies, they provide a generalised image of the poverty situation. The
366
stories of the informants inspired this researcher to rethink the influence of housing and the
poverty situation on individuals. One elderly subject who lives in the remote area of Tuen
Mun pointed out that people are living in a situation of deprivation and social exclusion in
Hong Kong despite the increased wages and anti-poverty measures implemented in recent
years. Another subject who has been working poor for many years and attempted had
attempted suicide and reminded me that the poor housing situation impacts the mental health
of individuals and even threatened their lives. Moreover, such an impact can be a long-term
one. Once families suffer from housing poverty, their long-term well-being and family
relationships are affected. Nevertheless, one interviewee who actively joined social actions
for policy advocacy impressed me by the resilience and resistance of the poor. Although they
may express their hopelessness and feeling of uselessness under the social structure, some of
them react proactively and fight for social justice for others. These stories reassured this
researcher about the subjectivity and agency power of individuals. Although macro forces,
such as globalisation and financialisation, are beyond the control of an individual, they
should find ways to live and strive to improve their lives.
Personal reflection. The mission of this researcher is to eradicate poverty and the
primary aim in pursuing a PhD is to enrich the understanding of theories on poverty,
specifically in terms of its description and the explanation of and change in poverty situation.
This PhD thesis facilitated the revision, reflection, and renewal of the researcher’s work
367
experience on housing and anti-poverty campaign conducted during the past years.
The theoretical review of poverty provided a historical perspective for understanding the
multidimensional nature of poverty. Different frameworks with various features, such as
material, social, cultural, and political aspects or perspectives have been suggested to
describe poverty. Initially, the plethora of frameworks had confused the researcher till the
complexity of the situation was understood. Given that poverty broadly means ‘absence of
well-being’ and that it is difficult to encompass human well-being in a single definition, this
review broadened the researcher’s concepts of housing and brought on the realisation that,
apart from housing cost, housing quality and community are also crucial in affecting the
poverty levels of families. However, housing has use and exchange values. Under the impact
of globalisation and neoliberalism, housing problems and poverty situations are deteriorating
rapidly and tremendously. The empirical review of Hong Kong echoed this analysis. To
address poverty, an in-depth study needs to be conducted on the relationship amongst
poverty, housing, and various macro factors.
A review of the social science philosophy brought clarity to the researcher’s world view
on social issues and human behaviour. However, an elaborate review in this thesis may not be
necessary. As a PhD student, this researcher hopes to tackle the philosophical questions in
social science, such as the existence of social reality, the nature of knowledge, and the ways
of knowing. Although the philosophical debate is occasionally endless, the arguments are
368
worth reviewing for the eventual discovery of one’s own position on the various arguments
being debated. The historical review of social science philosophy may not reveal the absolute
truth, but the process shows that which is judged as wrong and needs to be eliminated. A
mixed-methods model was used in this research to answer the research questions
comprehensively and to practice several research skills, such as advanced statistical analysis
and conducting in-depth interviews. This researcher will treasure this period as a PhD
student. The experience will be the basis on which to build further research experience for
evolving as a researcher.
Understanding is, without doubt, only the first step. After this research, there is a reason
for optimism that a comprehensive diagnosis of the social situation could suggest more
modes for addressing poverty and contribute to fighting inequality and social injustice.
8.5 Limitations of this study
This study, which covers the quantitative and qualitative aspects of the subject, has
several limitations.
Quantitative study.
Use of secondary data. The major result of the thesis theme is based on the quantitative
analysis of the secondary data (i.e., ‘census’ and ‘SPPR’ data set). This research aimed to
examine the housing poverty situation in Hong Kong. Although it is not feasible, in the time
available for a doctoral study, to conduct a large-scale random survey of a representative
369
sample of the entire population in Hong Kong, the use of the secondary data sets facilitated
making representative and generalisable findings that answered the research questions.
However, the variables in the ‘census’ data set are inadequate. Information on housing
conditions and non-income poverty is limited. For the ‘SPPR’ data set, although a random
sample stratified by living quarters and the living district was used for the survey, the number
of respondents in the sample who lived in subdivided flats was limited. Even though weights
were assigned, the analysis may yet underestimate the impact of housing factors on the
underprivileged.
Causal relationship. The ‘census’ and ‘SPPR’ data sets used are cross-sectional.
Therefore, the capability of these data sets to establish longitudinal causal relationships
amongst the variables of the cases. Although the SEM analysis was used, the data sets with
variables belonging to different periods are needed to establish longitudinal causal
relationships.
For the impact on poverty, socioeconomic status, social support, and some simple
background information were controlled. However, other factors affecting poverty and
deprivation, such as self-esteem, personal history, or in-born intelligence, were not controlled
in this research design. Although the ‘SPPR’ data set has many factors related to poverty, not
all the related factors can be controlled.
Measurement Issue. The housing cost in the measurement was underestimated because
370
‘census’ data excludes such expenses as electricity and water fees, management fees, and
government rates. ‘SPPR’ data excludes decoration and repair fees. Given the limited ‘SPPR’
data, the housing location was divided only into three groups for analysis. This may lead to
an underestimation of the impact of the neighbourhood effect. To make up for this
shortcoming, this additional detailed data such as tertiary planning units can be used in future
studies. Furthermore, the social dimension of social exclusion measurement in quantitative
research was constructed with variables related to social support and social network.
However, the possibility that the concept of social exclusion may include other dimensions,
such as employment and economic exclusion cannot be ruled out. The impact of housing
factors on the other dimensions of social exclusion can be explored in future studies.
The quantitative part focuses on the impact of housing factors on the poverty situation.
Based on the interrelated nature of housing and poverty, further study of the relationship
between the two factors in the reverse direction is recommended.
Direction of the variable relationships. The relationship between housing factors and
poverty situation is shown from the logistic regression perspective. However, the direction of
relationship is not clear. For example, people living under poverty may tend to live in low-
quality housing. The SEM model proposed, with measurement in different time slots, was
used to attempt and solve this problem. However, it requires significant cases of changing
housing situation in this period, which needs to be tested. Additional information can be
371
obtained through qualitative individual interviews, such as those cases moved from sub-
divided housing to public housing.
Effect of housing factors. To show the effect of housing factors on poverty, it is better to
choose people living in their houses for several years (say, 3 to 5 years). Nevertheless, it is
difficult to control this factor in random sampling. Information is limited to the year of living
in the ‘SPPR’ data set. For an improved investigation, in-depth case studies should be
conducted and additional questions in this area should be included in the interview guide.
Qualitative study
Case selection bias. The selection of interview cases may have introduced bias in data
collection. Although the purposeful case selection, aiming to maximise diversity in the
sample, was made, the number of cases, however, was limited. Also, most of the cases were
referred by social workers in NGOs. They may tend to be socially less isolated or may have
additional social support. Since the sample size is small, the qualitative findings based on it
are not generalisable.
Interviewer bias. Given that the researcher is responsible for conducting all interviews,
the involvement may affect the response of cases, such as asking leading questions that get
the expected answers or data. Moreover, coding was mostly done by the researcher, and the
researcher’s interpretation of the terms used by the interviewees was limited by the
experience of the researcher. A certain level of bias was unavoidable, but the researcher
372
maintained a reflexive and critical attituded during the interviews to provide an open space
for the informants to express their subjective experiences freely.
Cross-sectional interview. The informants were interviewed only once. Although the
researcher attempted to ask their experience in housing, family, and employment in the past,
their feelings, thoughts, and replies were constrained at the time of the interview. Moreover,
their memories of emotions and experiences could have been inaccurate. Longitudinal
interviews with the same informants are recommended to study their change of experience
and thoughts.
Observation. It had been suggested that the interview the informants should be
conducted in their respective living spaces to give the researcher to gather additional
information through observation. However, some of the informants preferred to be
interviewed in the service centres because their living places were extremely cramped or
noisy. The informants experienced difficulty in sharing the living experience, particularly the
impact of housing problems on their daily lives.
8.6 Recommendation and further research directions
This study is the beginning of a profound examination and explanation of the ‘housing
poverty’ situation in Hong Kong. The following recommendations are made for further
research.
Firstly, given that the stated focus of this research was on the impact of housing factors on
373
poverty, research in the reverse direction, that is, of the impact of poverty on the housing
situation is may be undertaken in the future.
Secondly, a large-scale or representative survey focused on the residents of subdivided
flats may be initiated in the future. Given that the housing problems they face are evident and
serious, this vulnerable group should be the focus of considerable attention. Additional
dimensions of housing, such as subjective views on housing and life satisfaction, can be
explored in the future.
Thirdly, cohort study can be conducted by using the census data to examine the housing
poverty situation faced by different cohort generations regarding the rising concern of
housing problems of the younger generation.
Fourthly, housing inequality has not been explored in this study. Such questions as to
whether the rising cost of housing pushes up or reduces housing inequality need exploration.
Such research can be conducted using census data for different years.
Fifthly, a longitudinal study will prove useful for exploring the impact of housing factors
on the poverty situation over time. The reason is that the impact of housing factors, such as
eviction, rent increase, and living in small areas may have long-term effects. Moreover, a
longitudinal study may also examine the adaption effect, such as subjective poverty and
living satisfaction. The second wave of the SPPR data set may provide information for such
research.
374
Lastly, the ‘poverty spiral’ and ‘multidimensional housing poverty’ frameworks are
constructed in this research and can be applied in other such research in other cities to
compare the situation there with that in Hong Kong.
8.7 Conclusive Remarks
Housing poverty, specifically the poverty situation induced by housing factors, is the core
concern of this PhD thesis. A mixed-method approach was used in this study to examine the
housing poverty situation in Hong Kong. The first finding was from the analysis of census
data was that the overall housing poverty has not worsened because of the increasing number
of homeowners who have paid off their mortgages. However, the housing poverty situation of
private rental housing residents was aggravated seriously. Second, the analysis of the SPPR
dataset revealed that the living area and indoor housing problems caused a significant impact
on the non-income poverty situation, which included deprivation, social exclusion, and
subjective poverty. Both housing factors acted as mediating factors between income and non-
income poverty. Amongst the three non-income poverty situations, subjective poverty was
influenced the most by housing factors, whilst social exclusion was the least affected.
Thirdly, residents fell into poverty owing to various critical life events and because their
choice of housing was severely limited. High housing costs, small living areas, and poor
housing conditions further worsened their poverty situation. Moreover, they were trapped in a
spiral of poverty and faced multidimensional housing poverty. This thesis contributes to
375
enriching the empirical description and theoretical explanation of poverty by including the
housing perspective in poverty study.
376
References
Aalbers, M. B., & Christophers, B. (2014). Centering Housing in Political Economy.

Housing, Theory and Society, 31(4), 373–394.
https://doi.org/10.1080/14036096.2014.947082
Aglietta, M. (1979). A Theory of Capitalist Regulation: The US Experience. London: NLB.
Aglietta, M. (1998). Capitalism at the turn of the century: regulation theory and the challenge
of social change. New Left Review, 232, 41–90.
Alcock, P. (1997). Understanding poverty. (J.Campling, Ed.) (2nd ed.). Hampshire:
Macmillan.
Alkire, S. (2005). Why the Capability Approach? Journal of Human Development, 6(1), 115–
135. https://doi.org/10.1080/146498805200034275
Alkire, S., & Foster, J. (2011). Counting and multidimensional poverty measurement. Journal
of Public Economics, 95, 476–487. https://doi.org/10.1016/j.jpubeco.2010.11.006
Allen, C. (2007). The meaning of housing: A pathways approach. Journal of Urban Affairs,
29(1), 102–104.
Amerigo, M., & Aragones, J. I. (1997). A theoretical and methodological approach to the
study of residential satisfaction. Journal of Environmental Psychology, 17(1), 47–57.
https://doi.org/10.1006/jevp.1996.0038
Ansell, B. (2014). The political economy of ownership: Housing markets and the welfare
state. American Political Science Review, 108(2), 383–402.
https://doi.org/10.1017/S0003055414000045
Arnold, E., & Skaburskis, A. (1989). Measuring Ontario’s increasing housing affordability
problem. Social Indicators Research, 21(5), 501–515.
https://doi.org/10.1007/BF00513458
Arthurson, K. (2004). From stigma to demolition : Australian debates about housing and
social exclusion. Journal of Housing and the Built Environment, 19(3), 255–270.
https://doi.org/10.1007/s10901-004-0692-1
Arthurson, K., & Jacobs, K. (2003). Social exclusion and housing.
Atkinson, A. B. (1987). On the Measurement of Poverty. Econometrica: Journal of the
Econometric Society, 55(4), 749–764.
Atkinson, A. B. (1998). Social exclusion, poverty and unemployment. Exclusion, Employment
and Opportunity. London.
Baert, P. (2005). Philosophy of the social sciences : towards pragmatism. Cambridge, UK:
Polity.
Ball, M. (1986). Housing Analysis: Time For a Theoretical Refocus? Housing Studies, 1(3),
147–166. https://doi.org/10.1080/02673038608720573
Becker, G. S. (1964). Human capital; a theoretical and empirical analysis, with special
377
reference to education. New York: National Bureau of Economic Research.
Bell, C. (1977). On Housing Classes. Journal of Sociology, 13(1), 36–40.
https://doi.org/10.1177/144078337701300108
Bell, D. (1973). The Coming of Post-Industrial Society: A Venture in Social Forecasting.
New York: Basic Books.
Bentler, P. M. (1990). Comparative Fit Indexes in Structural Models. Psychological Bulletin,
107(2), 238–246. https://doi.org/10.1037/0033-2909.107.2.238
Bhaskar, R. (1979). The possibility of naturalism: A philosophical critique of the
contemporary human sciences. New York: The Harvester Press.
Booth, C. (1887). The inhabitants of Tower Hamlets (School Board Division), their condition
and occupations. Journal of the Royal Statistical Society, 50(2), 326–401.
Booth, W. C., Colom, G. G., & Williams, J. M. (2008). The craft of research (3rd ed.).
Chicago: University of Chicago Press.
Borg, I. (2015). Housing Deprivation in Europe: On the Role of Rental Tenure Types.
Housing, Theory and Society, 32(1), 73–93.
Bourguignon, F., & Chakravarty, S. R. (2003). The Measurement of Multidimensional
Poverty. The Journal of Economic Inequality, 1(1), 25–49.
https://doi.org/10.1023/A:1023913831342
Boyer, R. (2000a). Is a Finance-led growth regime a viable alternative to Fordism? A
preliminary analysis. Economy and Society, 29(1), 111–145.
Boyer, R. (2000b). The political in the era of globalization and finance: Focus on some
regulation school research. International Journal of Urban and Regional Research,
24(2), 274–322. https://doi.org/10.1111/1468-2427.00250
Bradshaw, J., & Finch, N. (2003). Overlaps in dimensions of poverty. Journal of Social
Policy, 32(4), 513–525. https://doi.org/10.1017/S004727940300713X
Brady, D. (2019). Theories of the Causes of Poverty. Annual Review of Sociology, 45(1),
155–175. https://doi.org/10.1146/annurev-soc-073018-022550
Bramley, G. (2012). Affordability, poverty and housing need: triangulating measures and
standards. Journal of Housing and the Built Environment, 27(2), 133–151.
Bryman, A. (1988). Quantity and quality in social research. Boston: Unwin Hyman.
Busiol, D. (2016). A review of research on the consequences of living in a high-density city.
International Journal of Child and Adolescent Health, 9(4), 443–453.
Byrne, B. M. (2001). Structural Equation Modeling With AMOS,EQS, and LISREL:
Comparative Approaches to Testing for the Factorial Validity of a Measuring
Instrument. International Journal of Testing, 1(1), 55–86.
378
https://doi.org/10.1207/S15327574IJT0101
Caffaro, F., Galati, D., Zardoya Loureda, M. V., & Roccato, M. (2019). Housing-Related
Subjective Well-Being in Turin (Italy) and Havana (Cuba): Dimensions and Prediction.
Applied Research in Quality of Life, 14(1), 273–285. https://doi.org/10.1007/s11482-
018-9592-5
Cai, W., & Lu, X. (2015). Housing affordability : Beyond the income and price terms , using
China as a case study. Habitat International, 47, 169–175.
https://doi.org/10.1016/j.habitatint.2015.01.021
Calnitsky, D. (2018). Structural and individualistic theories of poverty. Sociology Compass,
12(12), 1–14. https://doi.org/10.1111/soc4.12640
Castells, M., Goh, L., & Kwok, R. Y. W. (1990). The Shek Kip Mei syndrome : economic
development and public housing in Hong Kong and Singapore. London: Pion Ltd.
Castro Campos, B., Yiu, C. Y. Y., Shen, J., Liao, K. H. H., & Maing, M. (2016). The
anticipated housing pathways to homeownership of young people in Hong Kong.
International Journal of Housing Policy, 16(2), 223–242.
Chan, E. H. W., So, H. M., Tang, B. S., & Wong, W. S. (2008). Private space, shared space
and private housing prices in Hong Kong: An exploratory study. Habitat International,
32(3), 336–348. https://doi.org/10.1016/j.habitatint.2007.11.004
Chan, E. H. W., Tang, B., & Wong, W. S. (2002). Density control and the quality of living
space: a case study of private housing development in Hong Kong. Habitat
International, 26(2), 159–175. https://doi.org/10.1016/S0197-3975(01)00041-8
Chan, H. L., Lee, S. K., & Woo, K. Y. (2001). Detecting rational bubbles in the residential
housing markets of Hong Kong. Economic Modelling, 18(1), 61–73.
https://doi.org/10.1016/S0264-9993(00)00030-4
Chan, K. H., Cheung, C. K., Ho, W. C., & Tija, Y. N. (2012). An Evaluation Study on the
Outcome of the Community Investment and Inclusion Fund (CIIF). Hong Kong.
Chan, K. W. (1999). Housing policy: The root of housing problems (in Chinese). In J.Lee &
S.Chiu (Eds.), New Social Policy (pp. 173–194). Hong Kong: The Chinese University of
Hong Kong.
Chan, K. W. (2000). Prosperity or Inequality: Deconstructing the Myth of Home Ownership
in Hong Kong. Housing Studies, 15(1), 28–43. https://doi.org/10.1080/02673030082450
Chan, L. S., & Chou, K. L. (2017). A Survey of Asset Poverty Among Older Adults of Hong
Kong. Social Indicators Research, 1–18. https://doi.org/10.1007/s11205-017-1674-5
Chaplin, R., & Freeman, A. (1999). Towards an Accurate Description of Affordability,
36(11), 1949–1957.
Chau, R. C. M., Yu, S. W. K., & Boxall, K. (2017). Combating Social Exclusion Faced by
Disabled people in the Wage Labour Market in Hong Kong. Social Policy and Society,
379
3, 1–15. https://doi.org/10.1017/S147474641700032X
Chen, K. M., Leu, C. H., & Wang, T. M. (2019). Measurement and Determinants of
Multidimensional Poverty: Evidence from Taiwan. Social Indicators Research, 145(2),
459–478. https://doi.org/10.1007/s11205-019-02118-8
Chen, Y. Y., Wong, G. H. Y., Lum, T. Y., Lou, V. W. Q., Ho, A. H. Y., Luo, H., & Tong, T.
L. W. (2016). Neighborhood support network, perceived proximity to community
facilities and depressive symptoms among low socioeconomic status Chinese elders.
Aging and Mental Health, 20(4), 423–431.
Cheung, C. K. (2013). Morale in Relation to Caring and Social Exclusion in Society. Social
Indicators Research, 113(1), 471–490. https://doi.org/10.1007/s11205-012-0105-x
Cheung, K. C. K., & Chou, K. L. (2016). Working Poor in Hong Kong. Social Indicators
Research, 129(1), 317–335. https://doi.org/10.1007/s11205-015-1104-5
Cheung, K. C. K., & Chou, K. L. (2017). Poverty, deprivation, and depressive symptoms
among older adults in Hong Kong. Aging & Mental Health, 0(0), 1–8.
Cheung, K. C. K., Chan, W. S., & Chou, K. L. (2019). Material Deprivation and Working
Poor in Hong Kong. Social Indicators Research, 145(1), 39–66.
Cheung, K. C. K., & Chou, K. L. (2018). Poverty , deprivation and life satisfaction among
Hong Kong older persons. Ageing and Society, 1–19.
Chiu, R. L. H. (2007). Planning, land and affordable housing in Hong Kong. Housing Studies,
22(1), 63–81. https://doi.org/10.1080/02673030601024614
Chiu, R. L. H. (2010). The Transferability of Hong Kong’s Public Housing Policy.
International Journal of Housing Policy, 10(3), 301–323.
Chou, K.L. (2013). Familial Effect on Child Poverty in Hong Kong Immigrant Families.
Social Indicators Research, 113, 183–195. https://doi.org/10.1007/s11205-012-0088-7
Chou, K. L. (2017). Social exclusion in old age: a validation study in Hong Kong. Aging &
Mental Health, 1–8. https://doi.org/10.1080/13607863.2017.1330870
Chou, K. L., & Chow, N. W. S. (2009). The roles of human capital and social capital in the
economic integration of new arrivals from Mainland China to Hong Kong. Habitat
International, 33(4), 340–346. https://doi.org/10.1016/j.habitatint.2008.08.001
Chou, K. L., & Lee, S. Y. (2018). Superimpose Material Deprivation Study on Poverty Old
Age People in Hong Kong Study. Social Indicators Research, 139(3), 1015–1036.
https://doi.org/10.1007/s11205-017-1740-z
Chuang, C. N., & Wei, C. P. (1998). Explanation and Criticism: On the Critical Scientific
Realism (in Chinese). Taiwanese Journal of Political Science, 9, 121–144.
380
Chung, R. Y., & Wong, S. Y. S. (2015). Health Inequality In Hong Kong. China Review,
15(2), 91–118.
Clapham, D. (2002). Housing Pathways: A Post Modern Analytical Framework. Housing,
Theory and Society, 19(2), 57–68. https://doi.org/10.1080/140360902760385565
Clapham, D. (2005). The meaning of housing A pathways approach. Bristol: Policy Press.
Clapham, D. (2009). Introduction to the Special Issue – A Theory of Housing: Problems and
Potential. Housing, Theory and Society, 26(1), 1–9.
Clapham, D. (2010). Happiness , well-being and housing policy, 38(2), 253–267.
Clapham, D., Foye, C., & Christian, J. (2018). The Concept of Subjective Well-being in
Housing Research. Housing, Theory and Society, 35(3), 261–280.
Clarke, A., Morris, S., & Williams, P. (2014). The role of housing organisations in reducing
poverty. York.
Clarke, S., & Ginsburg, N. (1975). The Political Economy of Housing. Political Economy
and the Housing Question. Papers Presented at the Housing Workshop, London.
Conference of Socialist Economists, 1–19.
Colasanto, D., Kapteyn, A., & van derGaag, J. (1984). Two Subjective Definitions of
Poverty : Results from the Wisconsin Basic Needs Study. The Journal of Human
Resources, 19(1), 127–138.
Coleman, J. S. (1988). Social Capital in the Creation of Human Capital. American Journal of
Sociology, 94, S95–S120. https://doi.org/10.1086/228943
Collier, A. (1994). Critical Realism: An Introduction to Roy Bhaskar’s Philosophy. London:
Verso.
Corbetta, P. (2003). Social research : theory, methods and techniques. Thousand Oaks,
Calif., Calif.: SAGE Publications.
Coyne, I. (1997). Sampling in qualitative research. Purposeful and theoretical sampling;
merging or clear boundaries? Journal of Advanced Nursing, 26(3), 623–630.
https://doi.org/10.1046/j.1365-2648.1997.t01-25-00999.x
Creswell, J. W. (2007). Qualitative inquiry and research design: Choosing among five
traditions (2nd ed.). Thousand Oaks, Calif.: SAGE Publications.
Creswell, J. W. (2014). Research design Qualitative, quantitative, and mixed methods
approaches (4th ed.). Thousand Oaks, Calif.: Sage Publications.
Creswell, J. W., & Plano Clark, V. L. (2007). Designing and conducting mixed methods
research. Thousand Oaks, CA: Sage.
Crettaz, E., & Suter, C. (2013). The Impact of Adaptive Preferences on Subjective Indicators:
An Analysis of Poverty Indicators. Social Indicators Research, 114(1), 139–152.
381
Crisp, R., Eadson, W., & While, A. (2016). Tackling poverty through housing and planning
policy in city regions. London. Retrieved from
https://www4.shu.ac.uk/research/cresr/sites/shu.ac.uk/files/tackling-poverty-housing-
planning-city-regions.pdf
Crossley, S. (2016). Book Review: Underclass: a history of the excluded since 1880. Housing
Studies, 31(1), 128–130. https://doi.org/10.1080/02673037.2015.1127350
Crotty, M. (1998). The foundations of social research : meaning and perspective in the
research process. Thousand Oaks, Calif.: Sage.
CSD. (2015). Thematic Household Survey Report No. 57: Housing conditions of sub-divided
units in Hong Kong. Hong Kong.
CSD. (2016). Thematic Household Survey Report No. 57: Housing conditions of sub-divided
units in Hong Kong. Hong Kong.
CSD. (2017a). Hong Kong 2016 Population By-census - Thematic Report: Household Income
Distribution in Hong Kong. Hong Kong.
CSD. (2017b). Hong Kong by-census 2016 result. Retrieved January15, 2017, from
http://www.bycensus2016.gov.hk/
Cuthbert, A. R. (1991). For a few dollars more: Urban planning and the legitimation process
in Hong Kong. International Journal of Urban and Regional Research, 15(4), 575–593.
Daniel, L., Baker, E., & Lester, L. (2018). Measuring Housing Affordability Stress: Can
Deprivation Capture Risk Made Real? Urban Policy and Research, 36(3), 271–286.
Davies, K. (2008). Housing Poverty: From social breakdown to social mobility. london.
Delang, C. O., & Lung, H. C. (2010). Public Housing and Poverty Concentration in Urban
Neighbourhoods: The Case of Hong Kong in the 1990s. Urban Studies, 47(7), 1391–
1413. https://doi.org/10.1177/0042098009353623
Demographia. (2017). 14th Annual Demographia International Demographia International
Housing Affordability Survey: 2017. New Zealand. Retrieved from
http://www.demographia.com/dhi2018.pdf
Denzin, N. K. (1971). The Logic of Naturalistic Inquiry. Social Forces, 50(2), 166–182.
Denzin, N. K., & Lincoln, Y. S. (2000). Handbook of qualitative research (3rd ed.).
Thousand Oaks: Sage.
Doling, J., & Ronald, R. (2014). Housing East Asia socioeconomic and demographic
challenges. New York: Palgrave Macmillan.
Donnellan, M. B., Trzesniewski, K. H., & Lucas, R. E. (2011). Introduction. In
M.Trzesniewski, K. H., Donnellan & R. E.Lucas (Eds.), Secondary data analysis: An
Introduction for Psychologists (pp. 3–10). Washington: American Psychological
Association.
Drakakis-Smith, D. W. (1979). High society : housing provision in metropolitan Hong Kong,
382
1954 to 1979, a jubilee critique. Hong Kong: University of Hong Kong.
Duclos, J.-Y., Sahn, D. E., & Younger, S. D. (2006). Robust Multidimensional Poverty
Comparisons with Discrete Indicators of Well-Being. The Economic Journal, 116(514),
943–968. https://doi.org/10.2139/ssrn.911224
Dunford, M. (1990). Theories of regulation. Environment and Planning D: Society and
Space, 8, 297–322.
Engels, F. (1975). The housing question (2nd ed.). Moscow: Progress Publishers.
Evans, G. W., Wells, N. M., Chan, H. Y. E., Saltzman, H., & Saltzmand, H. (2000). Housing
quality and mental health. Journal of Consulting and Clinical Psychology, 68(3), 526–
530. https://doi.org/10.1037/0022-006X.68.3.526
Evans, R. (1998). Tackling deprivation on social housing estates in englang: An assessment
of the housing plus approach. Housing Studies, 13(5), 713–726.
Fafard St-Germain, A. A., & Tarasuk, V. (2020). Homeownership status and risk of food
insecurity: Examining the role of housing debt, housing expenditure and housing asset
using a cross-sectional population-based survey of Canadian households. International
Journal for Equity in Health, 19(1). https://doi.org/10.1186/s12939-019-1114-z
Farr, J. (2004). Social Capital. Political Theory, 32(1), 6–33.
Feilzer, M. Y. (2010). Doing Mixed Methods Research Pragmatically: Implications for the
Rediscovery of Pragmatism as a Research Paradigm. Journal of Mixed Methods
Research, 4(1), 6–16. https://doi.org/10.1177/1558689809349691
Fernandez, R., & Aalbers, M. B. (2016). Financialization and housing: Between globalization
and Varieties of Capitalism. Competition and Change, 20(2), 71–88.
Field, J. (2008). Social Capital. New York: Routledge.
Flick, U. (2007). Designing qualitative research. London: SAGE Publications.
Flick, U. (2009). An introduction to qualitative research (4th ed.). London: SAGE
Publications.
Florida, R., Mellander, C., & Rentfrow, P. J. (2013). The Happiness of Cities. Regional
Fok, P. K., & Hu, J. jing. (2012). Paradigms of Curriculum Research: Integrating
Quantitative and Qualitative Research. Education Journal, 40(1–2), 1–14.
Fong, F. M., & Wong, C. (2015). Setting the Poverty Line: Policy Implications for Squaring
the Welfare Circle in Hong Kong. China Review, 15(2), 59–90. Retrieved from
http://www.jstor.org/stable/43710026%5Cnhttp://www.jstor.org/stable/pdfplus/10.2307/
43710026.pdf?acceptTC=true
Forrest, R. (2003). Some reflections on the housing question. In Housing and Social Change:
383
East-West Perspectives (p. 277). New York: Routledge.
Forrest, R. and, & Kearns, A. (2001). Social Cohension, Social Capital and the
Neighbourhood. Urban Studies, 38(12), 2125–2143.
Forrest, R., LaGrange, A., & Yip, N. M. (2002). Neighbourhood in a high rise, high density
city: Some observations on contemporary Hong Kong. Sociological Review, 50(2), 215-
240+311. https://doi.org/10.1111/1467-954X.00364
Forrest, R., LaGrange, A., & Yip, N. M. (2004). Hong Kong as a Global City? Social
Distance and Spatial Differentiation. Urban Studies, 41(1), 207–227.
Forrest, R., & Xian, S. (2018). Accommodating discontent: youth, conflict and the housing
question in Hong Kong. Housing Studies, 33(1), 1–17.
Forrest, R., & Yip, N. M. (2014). The Future for Reluctant Intervention: The Prospects for
Hong Kong’s Public Rental Sector. Housing Studies, 29(4), 551-565.
Foye, C. (2017). The Relationship Between Size of Living Space and Subjective Well-Being.
Journal of Happiness Studies, 18(2), 427–461. https://doi.org/10.1007/s10902-016-
9732-2
Foye, C., Clapham, D., & Gabrieli, T. (2018). Home-ownership as a social norm and
positional good: Subjective wellbeing evidence from panel data. Urban Studies, 55(6),
Friedmann, J. (1986). The World City Hypothesis. Development and Change, 17(1), 69–83.
https://doi.org/10.1111/j.1467-7660.1986.tb00231.x
Friedmann, J. (2001). World cities revisited: A comment. Urban Studies, 38(13), 2535–2536.
Friedrichs, J. (1997). Context effects of poverty neighbourhoods on residents. In
H.Vestergaard (Ed.), Housing in Europe (pp. 141–160).
Friedrichs, J., Galster, G., & Musterd, S. (2003). Housing Studies Neighbourhood effects on
social opportunities: the European and American research and policy context. Housing
Studies, 18(6), 797–806. https://doi.org/10.1080/0267303032000156291
FS Office. (2016). 2015 Study on Earnings Mobility. Hong Kong.
Fung, K. K., & Forrest, R. (2002). Institutional Mediation , The Hong Kong Residential
Housing Market and the Asian Financial Crisis. Housing Studies, 17(2), 189–207.
Fung, Kwok Kin, & Hung, S. L. (2014). Strengthening a community of poverty in an affluent
society: Strategies to build social capital in Tin ShuiWai North in Hong Kong.
Community Development Journal, 49(3), 441–457. https://doi.org/10.1093/cdj/bst051
Fusco, A. (2012). The relationship between income and housing deprivation in Luxembourg :
384
a longitudinal analysis (No. No 2012-10). Luxembourg.
Galster, G. (2002). Trans-Atlantic Perspectives on Opportunity, Deprivation and the Housing
Nexus. Housing Studies, 17(1), 5–10. https://doi.org/10.1080/02673030120105848
Gan, Q., & Hill, R. J. (2009). Measuring housing affordability : Looking beyond the median.
Journal of Housing Economics, 18, 115–125. https://doi.org/10.1016/j.jhe.2009.04.003
García-Vélez, D. F., Contreras-Jaramillo, M. A., Torres-Gutiérrez, T. P., & Correa-Quezada,
R. F. (2020). Social Exclusion in Ecuador with Housing Indicators: A Regional
Analysis. In Contributions to Management Science (pp. 159–176). Springer.
https://doi.org/10.1007/978-3-030-23604-5_10
Garland, D. (2016). The Welfare State: A Very Short Introduction. Oxford: Oxford University
Press.
Gilbert, A. G. (2014). Free housing for the poor: An effective way to address poverty?
Habitat International, 41, 253–261. https://doi.org/10.1016/j.habitatint.2013.08.009
Glaser, B. G. (1963). Retreading research materials: The use of secondary analysis by the
independent researcher. American Behavioral Scientist, 6(10), 11–14.
Glass, G.V. (1976). Primary, Secondary, and Meta-Analysis of Research. Educational
Researcher, 5(10), 3–8. https://doi.org/10.3102/0013189X005010003
Goedhart, T., Halberstadt, V., Kapteyn, A., & vanPraag, B. (1977). The poverty line: concept
and measurement. The Journal of Human Resources, 12(4), 503–520.
Goodstadt, L. F. (2013). Poverty in the midst of affluence: how Hong Kong mismanaged its
prosperity. Hong Kong: Hong Kong University Press.
Gordon, D. (1998). Definitions of concepts for the perceptions of poverty and social
exclusion. In Perceptions of Poverty and Social Exclusion 1998. Report on Preparatory
Research (pp. 5–15).
Gordon, D., Levitas, R., Pantazis, C., Patsios, D., Payne, S., Townsend, P., …Williams, J.
(2000). Poverty and Social Exclusion in Britain. York.
Gordon, D., & Pantazis, C. (1997). Breadline Britain in the 1990s. (D.Gordon &C.Pantazis,
Eds.) (1st ed.). Bristol: Summerleaze House Books.
Gou, Z., Xie, X., Lu, Y., & Khoshbakht, M. (2018). Quality of life (QoL) survey in hong
kong: Understanding the importance of housing environment and needs of residents
from different housing sectors. International Journal of Environmental Research and
Public Health, 15(2). https://doi.org/10.3390/ijerph15020219
Goulden, C., & D’Arcy, C. (2014). A Definition of Poverty: JRF Programme Paper. York.
Gray, M., & Simpson, B. (1998). Developmental social work education: a field example.
International Social Work, 41(2), 227–237.
Greene, J. C., Caracelli, V. J., & Graham, W. F. (1989). Toward a Conceptual Framework for
Mixed-Method Evaluation Designs. Educational Evaluation and Policy Analysis, 11(3),
385
Guio, A.-C., & Maquet, I. E. (2007). Material deprivation and poor housing. Comparative
EU statistics on income and living conditions: Issues and challenge.
Guo, Y., Chang, S.Sen, Chen, M., & Yip, P. S. F. (2017). Do Poorer Areas Have Poorer
Access to Services in Hong Kong? A Small-Area Analysis Based on Multiple Spatial
Accessibility Indicators. Social Indicators Research, 138(1), 1–21.
Guo, Y., Chang, S.Sen, Sha, F., & Yip, P. S. F. (2018). Poverty concentration in an affluent
city: Geographic variation and correlates of neighborhood poverty rates in Hong Kong.
PLoS ONE, 13(2), 1–18. https://doi.org/10.1371/journal.pone.0190566
Gustafsson, B., Shi, L., & Sato, H. (2004). Can a subjective poverty line be applied to China?
Assessing poverty among urban residents in 1999. Journal of International
Development, 16(8), 1089–1107. https://doi.org/10.1002/jid.1127
Hagenaars, A., & deVos, K. (1988). The Definition and Measurement of Poverty. Journal of
Human Resources, 23(2), 211–221. Retrieved from
http://search.proquest.com/docview/202734770?pq-origsite=gscholar
Haila, A. (2000). Real Estate in Global Cities: Singapore and Hong Kong as Property States.
Urban Studies, 37(12), 2241–2256. https://doi.org/10.1080/00420980020002797
Hair, J. F., Black, W. C., Babin, B. J., & Anderson, R. E. (2014). Multivariate data analysis
(7th ed.). London: Pearson Education.
Hakim, C. (1982). Secondary analysis in social research : a guide to data sources and
methods with examples. Boston: Allen & Unwin.
Hancock, K. E. (1993). “Can Pay?Won’t Pay?” or Economic Principles of ’ Affordability’.
Urban Studies, 30(1), 127–145.
Hartman, C. (1998). The Case for a Right to Housing. Housing Policy Debate, 9(2), 223–246.
Harvey, D. (2005). A brief history of neoliberalism. New York: Oxford University Press.
Harvey, D. (2014). Seventeen contradictions and the end of capitalism. London: Profile
Books.
He, S., Liu, L., Yang, G., & Wang, F. (2017). Housing differentiation and housing poverty in
Chinese low-income urban neighborhoods under the confluence of State-market forces.
Urban Geography, 38(5), 729–751. https://doi.org/10.1080/02723638.2016.1139406
Helliwell, J. F. (2003). How’s the job? Well-being and social capital in the workplace.
Economic Modelling, 20, 331–360. https://doi.org/10.1016/S0264-9993(02)00057-3
Herbers, D. J., & Mulder, C. H. (2017). Housing and subjective well-being of older adults in
Europe. Journal of Housing and the Built Environment, 32(3), 533–558.
Hesse-Biber, S. N. (2010). Mixed methods research : merging theory with practice. New
386
York: Guilford Press.
HKCSS. (2011). Hong Kong Deprivation and Social Exclusion Research (in Chinese). Hong
Kong.
HKCSS. (2012). Hong Kong Deprivation and Social Exclusion Research - CSSA recipients,
Families with disabled and Elderly (in Chinese). Hong Kong.
HKCSS. (2014). Social Deprivation of the socially disadvantages (in Chinese). Hong Kong.
HKCSS. (2017). Social Indicators of Hong Kong. Retrieved November6, 2017, from
https://www.socialindicators.org.hk/
HKHA government (2017). Hong Kong Housing Authority Web Page. Retrieved March7,
2017, from https://www.housingauthority.gov.hk/
HKSAR government. (2016). Supplementary Poverty Line Analysis: Expenditure Patterns of
Poor Households in 2015. Hong Kong.
HKSAR government (2017a). CS hails social capital builders. Government News. Retrieved
from
http://www.news.gov.hk/en/categories/health/html/2017/01/20170106_174532.shtml
HKSAR government (2017b). Hong Kong is ranked as the world’s freest economy again.
Retrieved October23, 2017, from
http://www.info.gov.hk/gia/general/201702/16/P2017021600208.htm
HKSAR government (2017). Hong Kong Poverty Situation Report 2016. Hong Kong Poverty
Situation Report. Hong Kong. Retrieved from
https://www.povertyrelief.gov.hk/eng/pdf/Hong_Kong_Poverty_Situation_Report_2018
(2019.12.13).pdf
HKSAR government (2019). Hong Kong Poverty Situation Report 2016. Hong Kong Poverty
Situation Report. Hong Kong. Retrieved from
https://www.povertyrelief.gov.hk/eng/pdf/Hong_Kong_Poverty_Situation_Report_2018
(2019.12.13).pdf
Ho, A. P. ying, & Chan, K. T. (2010). The social impact of work-integration social enterprise
in Hong Kong. International Social Work, 53(1), 33–45.
Ho, K. Y., Li, W. H. C., & Chan, S. S. C. (2015). The Effect of Poverty and Income Disparity
on the Psychological Well-Being of Hong Kong Children. Public Health Nursing, 32(3),
212–221. https://doi.org/10.1111/phn.12147
Ho, M. H. C., & Chiu, R. L. H. (2002). Impact of accessibility on housing expenditure and
affordability in Hong Kong’s private rental sector. Journal of Housing and the Built
Environment, 17(4), 363–383. https://doi.org/10.1023/A:1021153911692
Hofferth, S. L. (2005). Secondary Data Analysis in Family Research. Journal of Marriage
and Family, 67, 891–907.
Holliday, I. (2000). Productivist welfare capitalism: Social policy in East Asia. Political
387
Studies, 48, 706–723. https://doi.org/10.1111/1467-9248.00279
Holliday, I., & Tam, W. K. (2001). Social Capital in Hong Kong. An International Quarterly,
19(1–2), 144–170.
Hollis, M. (1994). The philosophy of social science : an introduction. New York: Cambridge
University Press.
Hooper, D., Coughlan, J., & Mullen, M. R. (2008). Structural equation modelling: Guidelines
for determining model fit. Electronic Journal of Business Research Methods, 6(1), 53–
60. https://doi.org/10.21427/D79B73
Hopkins, K. (1971). Hong Kong: the industrial colony; a political, social and economic
survey. (K.Hopkins, Ed.). New York: Oxford University Press.
Howe, K. R. (1988). Against the Quantitative-Qualitative Incompatibility Thesis or Dogmas
Die Hard. Educational Researcher, 17(8), 10–16.
https://doi.org/10.3102/0013189X017008010
Hoyle, R. H. (2012). Handbook of structural equation modeling. New York, NY: The
Guilford Press.
Hu, F. Z. Y., & Chou, K. L. (2015). The antipoverty effect of public rental housing in Hong
Kong. Habitat International, 46, 206–213.
Hu, L. T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure
analysis: Conventional criteria versus new alternatives. Structural Equation Modeling,
Hui, E. C. M. (1999). Public Housing Rents in Hong Kong: Policy and Structure. Journal of
Urban Planning and Development, 125(1), 36–53. https://doi.org/10.1061/(ASCE)0733-
9488(1999)125:1(36)
Hui, E. C. M. (2001). Measuring Affordability in Public Housing from Economic Principles:
Case Study of Hong Kong. Journal of Urban Planning and Development, 127(1), 34–
49. https://doi.org/10.1061/(ASCE)0733-9488(2001)127:1(34)
Hui, E. C. M., Chau, C. K., Pun, L., & Law, M. Y. (2007). Measuring the neighboring and
environmental effects on residential property value: Using spatial weighting matrix.
Building and Environment, 42(6), 2333–2343.
https://doi.org/10.1016/j.buildenv.2006.05.004
Hui, E. C. M., Wong, F. K. W., Chung, K. W., & Lau, K. Y. (2014). Housing affordability,
preferences and expectations of elderly with government intervention. Habitat
International, 43, 11–21. https://doi.org/10.1016/j.habitatint.2014.01.010
Hulchanski, J. D. (1995). The concept of housing affordability: Six contemporary uses of the
housing expenditure‐to‐income ratio. Housing Studies, 10(4), 471–491.
Ibrahima, S. (2013). The Subjective Approach as a Tool for Understanding Poverty: The
388
Case of Senegal. Procedia Economics and Finance, 5(13), 336–345.
https://doi.org/10.1016/s2212-5671(13)00040-3
Izuhara, M., & Forrest, R. (2013). ‘Active Families’: Familization, Housing and Welfare
across Generations in East Asia. Social Policy & Administration, 47(5), 520–541.
https://doi.org/10.1111/spol.12002
Jessop, B. (2002). The future of the capitalist state. Cambridge, UK: Polity.
Johnson, R. B., & Onwuegbuzie, A. J. (2004). Mixed methods research: A research paradigm
whose time has come. Educational Researcher, 33(7), 14–26.
https://doi.org/10.3102/0013189X033007014
Johnson, R. B., Onwuegbuzie, A. J., & Turner, L. A. (2007). Toward a Definition of Mixed
Methods Research. Journal of Mixed Methods Research, 1(2), 112–133.
Johnston, M. P. (2014). Secondary Data Analysis : A Method of which the Time Has Come.
Qualitative and Quantitative Methods in Libraries, 3(3), 619–626.
https://doi.org/10.1097/00125817-200207000-00009
JRF. (2015). Housing and Poverty. York. https://doi.org/10.2307/1924490
Kakwani, N., & Silber, J. (2008). Introduction: Multidimensional Poverty Analysis:
Conceptual Issues, Empirical Illustrations and Policy Implications. World Development,
36(6), 987–991. https://doi.org/10.1016/j.worlddev.2007.10.004
Kapteyn, A., Kooreman, P., & Willemse, R. (1988). Some Methodological Issues in the
Implementation of Subjective Poverty Definitions. The Journal of Human Resources,
23(2), 222–242.
Kemeny, J. (1992). Housing and Social Theory. New York: Routledge.
https://doi.org/10.1017/CBO9781107415324.004
Kemeny, J. (2001). Comparative housing and welfare: Theorising the relationship. Journal of
Housing and the Built Environment, 16(1), 53–70.
Kemeny, J. (2006). Corporatism and housing regimes. Housing, Theory and Society, 23(1),
Kemeny, J., & Lowe, S. (1998). Schools of Comparative Housing Research: From
Convergence to Divergence. Housing Studies, 13(2), 161–176.
King, P. (2009). Using theory or making theory: Can there be theories of housing? Housing,
Theory and Society, 26(1), 41–52. https://doi.org/10.1080/14036090802704296
King, P. (2015). The Principles of Housing. London: Routledge.
King, P. (2017). Thinking on housing: Words, Memories, Use. New York: Routledge.
Kingdon, G. G., & Knight, J. J. (2006). Subjective well-being poverty vs. income poverty and
capabilities poverty? The Journal of Development Studies, 42(7), 1199–1224.
389
Kline, R. B. (2011). Principles & practice of structural equation modelling (3rd ed.). New
York: The Guilford Press.
Kornbluh, M. (2015). Combatting challenges to establishing trustworthiness in qualitative
research. Qualitative Research in Psychology, 12(4), 397–414.
Ku, H. B. (2006). Body, dress and cultural exclusion: Experiences of Pakistani women
in ’global’Hong Kong. Asian Ethnicity, 7(3), 285–302.
Kuhn, T. S. (1962). The structure of scientific revolutions. Chicago: University of Chicago
Press.
Kuo, C. yeoung. (2011). Mixed research and the qualitative-quantitative debate (in Chinese).
Soochow Journal of Political Science, 29(1), 1–64.
Kutty, N. K. (2002). The house poor - are high housing costs keeping non-poor americans at
poverty standards of living? Austria.
Kutty, N. K. (2005). A new measure of housing affordability Estimates and analytical results.
Housing Policy Debate, 16(1), 113–142.
Kwadzo, M. (2015). Choosing Concepts and Measurements of Poverty: A Comparison of
Three Major Poverty Approaches. Journal of Poverty, 19(4), 409–423.
Kwon, H. (2005). Transforming the Developmental Welfare State in East Asia. Development
and Change, 36(3), 477–497. https://doi.org/10.1111/j.0012-155X.2005.00420.x
LaGrange, A. (1998). Privatising Public Housing in Hong Kong: Its Impact on Equity.
Housing Studies, 13(4), 507–525. https://doi.org/10.1080/02673039883245
LaGrange, A. (1999). Privatizing Public Housing in the Welfare Regime of a Tiger Economy:
A Case Study of Hong Kong. Housing, Theory and Society, 16(1), 17–30.
LaGrange, A. (2011). Neighbourhood and class: A study of three neighbourhoods in Hong
Kong. Urban Studies, 48(6), 1181–1200. https://doi.org/10.1177/0042098010370628
LaGrange, A., &Pretorius, F. (2002). Private rental housing in Hong Kong. Housing Studies,
LaGrange, A., & Pretorius, F. (2016). State-led gentrification in Hong Kong. Urban Studies,
Laderchi, C. R., Saith, R., & Stewart, F. (2003). Does it Matter that we do not Agree on the
Definition of Poverty? A Comparison of Four Approaches. Oxford Development Studies,
Lai, N., & Wang, K. (1999). Land-supply restrictions, developer strategies and housing
policies: the case in Hong Kong. International Real Estate Review, 2(1), 143–159.
390
Retrieved from http://umacweb1.umac.mo/fba/irer/papers/past/vol2_pdf/143-
159hkland.pdf
Lam, C. L. K., Guo, V. Y., Wong, C. K. H., Yu, E. Y. T., & Fung, C. S. C. (2017). Poverty
and health-related quality of life of people living in Hong Kong: Comparison of
individuals from low-income families and the general population. Journal of Public
Health (United Kingdom), 39(2), 1–8. https://doi.org/10.1093/pubmed/fdw046
Lam, C. M. (2011). Psychological Stress and Parenting Behavior Among Chinese Families:
Findings from a Study on Parent Education for Economically Disadvantaged Families.
Social Indicators Research, 100(3), 451–462. https://doi.org/10.1007/s11205-010-9623-
6
Lau, K. Y., & Murie, A. (2016). Residualisation and resilience: public housing in Hong
Kong. Housing Studies, 1–25. https://doi.org/10.1080/02673037.2016.1194376
Lau, M. H. M., & Wei, X. (2018). Housing size and housing market dynamics: The case of
micro-flats in Hong Kong. Land Use Policy, 78, 278–286.
https://doi.org/10.1016/j.landusepol.2018.06.039
Lau, M.. (2010). A globalizing economy, sustainable livelihoods and equality of
opportunities: The case of Hong Kong. Journal of Asian Public Policy, 3(2), 146–159.
Lau, M., Gordon, D., Pantazis, C., Sutton, E., & Lai, L. (2015). Including the Views of the
Public in a Survey of Poverty and Social Exclusion in Hong Kong: Findings from Focus
Group Research. Social Indicators Research, 124(2), 383–400.
Lau, M,, Pantazis, C., Gordon, D., Lai, L., & Sutton, E. (2015). Poverty in Hong Kong.
China Review, 15(2), 23–58.
Lau, M. (2017). Framing processes in planning disputes: analysing dynamics of contention in
a housing project in Hong Kong. Housing Studies, 1–17.
Lavalette, M., & Pratt, A. (2006). Social Policy: Theories, Concepts and Issues (3rd ed.).
Thousand Oaks, Calif.: SAGE Publications.
Lawless, N. M., & Lucas, R. E. (2011). Predictors of Regional Well-Being: A County Level
Analysis. Social Indicators Research, 101(3), 341–357. https://doi.org/10.1007/s11205-
010-9667-7
Lee, J. K. (1999). Housing, home ownership and social change in Hong Kong. Housing,
Home Ownership and Social Change in Hong Kong. Aldershot: Ashgate.
Lee, J. K. (2014). Housing Policy and Asset Building: Exploring the Role of Home
Ownership in East Asain Social Policy. In Social Policy and Change in East Asia (p.
218). Lanham: Lexington Books.
391
Lee, J. K., & Yip, N. M. (2006). Public Housing and Family Life in East Asia: Housing
History and Social Change in Hong Kong, 1953-1990. Journal of Family History, 31(1),
Lee, K. M., Wong, H., & Law, K. yee. (2007). Social Polarisation and Poverty in the Global
City: The Case of Hong Kong. China Report, 43(1), 1–30.
Lee, P. (1994). Housing and Spatial Deprivation: Relocating the Underclass and the New
Urban Poor. Urban Studies, 31(7), 1191–1209.
Lee, S. Y., & Chou, K. L. (2016). Trends in Elderly Poverty in Hong Kong: A
Decomposition Analysis. Social Indicators Research, 129(2), 551–564.
LegCo. (2015). Social mobility in Hong Kong. Hong Kong.
LegCo. (2016). Statistics Highlight - Housing: Public Housing. Hong Kong. Retrieved from
https://www.legco.gov.hk/research-publications/english/1617issh09-public-housing-
20161122-e.pdf
Legislative Council. (2019). Statistics Highlight - Housing: Public Rental Housing. Hong
Kong. Retrieved from https://www.legco.gov.hk/research-
publications/english/1819issh31-public-rental-housing-20190904-e.pdf
Lelkes, O., & Zólyomi, E. (2009). Quality of housing and the link to income.
Lerman, D. L., & Reeder, W. J. (1987). The affordability of adequate housing. AREUEA
Journal, 15(4).
Leu, C., Chen, K., & Chen, H. (2016). Subjective Poverty: Profiles and Determinants.
Journal of Social Sciences and Philosophy, 28(3), 1–36.
Leung, J. T. Y., & Shek, D. T. L. (2013a). Parental beliefs and family functioning in Chinese
families experiencing economic disadvantage in Hong Kong. International Journal on
Disability and Human Development, 12(2), 151–161. https://doi.org/10.1515/ijdhd-
2013-0006
Leung, J. T. Y., & Shek, D. T. L. (2013b). Parenting for resilience: Family processes and
psychosocial competence of Chinese adolescents experiencing economicdisadvantage in
Hong Kong. International Journal on Disability and Human Development, 12(2), 127–
137. https://doi.org/10.1515/ijdhd-2012-0137
Levitas, R. (2006a). The Concept and Measurement of Social Exclusion. In C.Pantazis,
D.Gordon, & R.Levitas (Eds.), Poverty and Social Exclusion in Britain (pp. 123–160).
The Policy Press.
Levitas, R. (2006b). The concepts and measurement of social exclusion. In ChristinaPantazis,
D.Gordon, & R.Levitas (Eds.), Poverty and Social Exclusion in Britain (pp. 123–160).
Bristol: Policy Press.
392
Levitas, R., Pantazis, C., Fahmy, E., Gordon, D., Lloyd, E., & Patsios, D. (2007). The multi-
dimensional analysis of social exclusion. University of Bristol. Retrieved from
http://roar.uel.ac.uk/1781/
Ley, D., & Teo, S. Y. (2014). Gentrification in Hong Kong? Epistemology vs. Ontology.
International Journal of Urban and Regional Research, 38(4), 1286–1303.
https://doi.org/10.1111/1468-2427.12109
Li, J. V. (2014). Recent trends on housing affordability research : where are we up to ? Hong
Kong.
Li, J. V. (2016). Housing Policies in Hong Kong, China and the People’s Republic of China
(ADBI Working Paper 566 No. ADBI Working Paper 566). Tokyo.
Liao, P. S. (2009). Parallels between objective indicators and subjective perceptions of
quality of life: A study of metropolitan and county areas in Taiwan. In Social Indicators
Research (Vol. 91, pp. 99–114). https://doi.org/10.1007/s11205-008-9327-3
Lincoln, Y. S., & Guba, E. G. (1985). Naturalistic inquiry. Beverly Hills, Calif.: SAGE
Publications.
Linneman, P. D., & Megbolugbe, I. F. (1992). Housing Affordability: Myth or Reality?
Urban Studies, 29(3–4), 369–392. https://doi.org/10.1080/00420989220080491
Lister, R. (2004). Poverty. Cambridge, UK: Polity.
Lister, R. (2010). Understanding Theories and Concepts in Social Policy. Bristo: Policy
Press.
Loehlin, J. C. (1992). Latent variable models. An introduction to factor, path and structural
analysis (2nd ed.). New Jersey: Lawerence Erlbaum Associates.
Loury, G. C. (1977). A dynamic theory of racial income differences. In P. A.Wallace & A.
M.LaMond (Eds.), Women, minorities, and employment discrimination (pp. 153–188).
Lexington Books.
Loury, G. C. (1981). Intergenerational transfers and the distribution of earnings.
Econometrica: Journal of the Econometric Society, 49(4), 843–867.
Low, C. T., Stimson, R., Chen, S., Cerin, E., Wong, P. P. Y., & Lai, P. C. (2018). Personal
and Neighbourhood Indicators of Quality of Urban Life: A Case Study of Hong Kong.
3
Lui, H. K. (2007). The Redistributive Effect of Public Housing in Hong Kong. Urban
Lui, H. K., & Suen, W. (2011). The effects of public housing on internal mobility in Hong
Kong. Journal of Housing Economics, 20(1), 15–29.
https://doi.org/10.1016/j.jhe.2010.11.001
Lum, T. Y. S., Lou, V. W. Q., Chen, Y., Wong, G. H. Y., Luo, H., & Tong, T. L. W. (2016).
Neighborhood support and aging-in-place preference among low-income elderly
393
Chinese city-dwellers. Journals of Gerontology - Series B Psychological Sciences and
Social Sciences, 71(1), 98–105. https://doi.org/10.1093/geronb/gbu154
MacDonald, H., Galster, G., & Dufty-Jones, R. (2018). The geography of rental housing
discrimination, segregation, and social exclusion: New evidence from Sydney. Journal
of Urban Affairs, 40(2), 226–245. https://doi.org/10.1080/07352166.2017.1324247
Mack, J., & Lansley, S. (1985). Poor Britain. London: George Allen & Unwin.
Maclennan, D., & William, R. U. (1990). Affordable Housing in Britain and the United
States. York.
Madden, D., & Marcuse, P. (2016). In defense of housing. London: Verso.
Maestri, V. (2015). A Measure of Income Poverty Including Housing: Benefits and
Limitations for Policy Making. Social Indicators Research, 121(3), 675–696.
https://doi.org/10.1007/s11205-014-0657-z
Mahmood, T., Yu, X., & Klasen, S. (2018). Do the Poor Really Feel Poor? Comparing
Objective Poverty with Subjective Poverty in Pakistan. Social Indicators Research,
142(2), 1–38. https://doi.org/10.1007/s11205-018-1921-4
Malpass, P., & Murie, A. (1999). Housing policy and practice. Basingstoke: Macmillan.
Markkanen, S., & Harrison, M. (2013). “Race”, Deprivation and the Research Agenda:
Revisiting Housing, Ethnicity and Neighbourhoods. Housing Studies, 28(3), 409–428.
Marsh, A., & Mullins, D. (1998). The social exclusion perspective and housing studies:
Origins, applications and limitations. Housing Studies, 13(6), 749–759.
Marshall, M. N. (1996). Sampling for qualitative research. Family Practice, 13(6), 522–525.
https://doi.org/10.1093/fampra/13.6.522
McConnell, E. D. (2012). House poor in Los Angeles: Examining patterns of housing-
induced poverty by race, nativity, and legal status. Housing Policy Debate, 22(4), 605–
631. https://doi.org/10.1080/10511482.2012.697908
Milanovic, B. (2016). Introduction. In Global inequality: a new approach for the age of
globalization (pp. 1–9). London: Cambridge University Press.
Miles, M. B., & Huberman, A. M. (1994). Qualitative data analysis: An expanded
sourcebook (2nd ed.). Thousand Oaks, Calif.: SAGE Publications.
Minton, A. (2014). Big Capital: Who Is London For? London: Penguin.
Mitchell, R. E. (1971). Some social implications of high density housing. American
Sociological Review, 36(1), 18–29. https://doi.org/10.2307/2093503
Mitra, S. (2016). Synergies Among Monetary, Multidimensional and Subjective Poverty:
Evidence from Nepal. Social Indicators Research, 125(1), 103–125.
Mohit, M. A., Ibrahim, M., & Rashid, Y. R. (2010). Assessment of residential satisfaction in
394
newly designed public low-cost housing in Kuala Lumpur, Malaysia. Habitat
International, 34(1), 18–27. https://doi.org/10.1016/j.habitatint.2009.04.002
Moisio, P. (2004). A latent class application to the multidimensional measurement of poverty.
Quality and Quantity, 38(6), 703–717. https://doi.org/10.1007/s11135-004-5940-7
Mok, K. H. (2015). Enhancing Global Competitiveness and Human Capital Management:
Does Education Help Reduce Inequality and Poverty in Hong Kong? China Review,
15(2), 119–146. Retrieved from http://search.proquest.com.lib-
ezproxy.hkbu.edu.hk/docview/1735395445?pq-origsite=summon
Monkkonen, P. (2011). Public Housing and Unemployment : Skills and Spatial Mismatch in
Postindustrial Hong Kong (IURD Workinf Paper Series). Hong Kong.
Morgan, D. L. (1996). Focus groups as qualitative research (2nd ed.). Thousand Oaks,
Calif.: SAGE Publications.
Morgan, D. L. (2007). Paradigms Lost and Pragmatism Regained. Journal of Mixed Methods
Research, 1(1), 48–76. https://doi.org/10.1177/2345678906292462
Morgan, D. L. (2014). Pragmatism as a Paradigm for Social Research. Qualitative Inquiry,
Morrow, V. (1999). Conceptualising Social Capital in Relation to the Well-Being of Children
and Young People: A Critical Review. The Sociological Review, 47(4), 744–765.
https://doi.org/10.1111/1467-954X.00194
Morse, J. M. (1991). Approaches to qualitative-quantitative methodological triangulation.
Nursing Research, 40(2), 120–123.
Munch, S. (2012). social exclusion and housing. In International Encyclopedia of Housing
and Home (pp. 377–380).
Munck, R. (2005). Social Exclusion: New Inequality Paradigm for the Era of Globalization?
In M.Romero & E.Margolis (Eds.), The Blackwell Companion to Social Inequalities (pp.
31–49). Washington, DC: Blackwell Publishing.
Murray, C. (1984). Losing Ground: American Social Policy 1950-1980. New York: Basic
Books.
Napiorkowska-Baryla, A., & Witkowska-Dabrowska, M. (2018). Housing Deprivation in
Poland - A Regional Approach. In 2018 Baltic Geodetic Congress (pp. 221–225).
https://doi.org/10.1109/BGC-Geomatics.2018.00048
Narayan, D., Chambers, R., Shah, M. K., & Petesch, P. (2000). Voices of the poor: Crying
out for change.
Neuman, L. W. (2014). Social Research Methods: Qualitative and Quantitative Approaches
(7th ed.). Boston: Allyn & Bacon.
Ng, S. L., Zhang, Y., Ng, K. H., Wong, H., & Lee, J. W. Y. (2018). Living environment and
quality of life in Hong Kong. Asian Geographer, 35(1), 35–51.
395
Nussbaum, M. C. (2000). Women and human development : the capabilities approach. New
York: Cambridge University Press.
Nussbaum, M. C. (2011). Creating capabilities : the human development approach.
Cambridge, Mass.: Belknap Press of Harvard University Press.
O’Brien, M., & Penna, S. (1998). Theorising welfare enlightenment and modern society.
London: Sage.
O’Neill, P. (2008). The Role of Theory in Housing Research: Partial Reflections of the Work
of Jim Kemeny. Housing, Theory and Society, 25(3), 164–176.
Opit, S., Witten, K., & Kearns, R. (2019). Housing pathways, aspirations and preferences of
young adults within increasing urban density. Housing Studies, 0(0), 1–20.
Orshansky, M. (1969). How poverty is measured. Monthly Labor Review, 92(2), 37–41.
Oxfam. (2016). Hong Kong Poverty Report (2011-2015). Hong Kong.
Oxfam. (2017). Oxfam: Our purpose and beliefs. Retrieved June23, 2017, from
https://www.oxfam.org/en/our-purpose-and-beliefs
Oyen, E. (2002). Social Capital Formation as a Poverty Reducing Strategy? In Social capital
and poverty reduction (pp. 11–14). UNESCO.
Padgett, D. K. (2017). Qualitative methods in social work research (3rd ed.). Los Angeles:
SAGE.
Park, S., Cho, J., & Chen, Y. C. (2019). Subsidized housing and geographic accessibility to
neighborhood resources for low-income older people: From later year social exclusion
perspective. Geoforum, 106, 297–304. https://doi.org/10.1016/j.geoforum.2019.09.002
Patterson, M., Moniruzzaman, A., Palepu, A., Zabkiewicz, D., Frankish, C. J., Krausz, M., &
Somers, J. M. (2013). Housing First improves subjective quality of life among homeless
adults with mental illness: 12-month findings from a randomized controlled trial in
Vancouver, British Columbia. Social Psychiatry and Psychiatric Epidemiology, 48(8),
Patton, M. Q. (1990). Qualitative evaluation and research methods. Newbury Park, Calif.:
SAGE Publications.
Pawson, H., & Herath, S. (2017). Sinks of Social Exclusion or Springboards for Social
Mobility? Analysing the Roles of Disadvantaged Neighbourhoods in Urban Australia.
Urban Policy and Research, 35(4), 373–390.
Piketty, T. (2014). Capital in the Twenty-First Century. Cambridge Massachusetts: The
Belknap Press of Harvard University Press.
Policy 21. (2013). Report on survey on Sub-divided United in Hong Kong. Hong Kong.
Popper, K. (1959). The logic of scientific discovery. London: Routledge.
396
Portes, A. (1998). Social Capital: Its Origins and Applications in Modern Sociology. Annual
Review of Sociology, 24(1), 1–24. https://doi.org/10.1146/annurev.soc.24.1.1
Posel, D., & Rogan, M. (2016). Measured as Poor versus Feeling Poor: Comparing Money-
metric and Subjective Poverty Rates in South Africa. Journal of Human Development
and Capabilities, 17(1), 55–73. https://doi.org/10.1080/19452829.2014.985198
Pradhan, M., & Ravallion, M. (2000). Measuring poverty using qualitative perceptions of
consumption adequacy. Development, 82(3), 462–471.
Pryor, E. G. (1983). Housing in Hong Kong (2nd ed.). Hong Kong: Oxford University Press.
Putnam, R. D. (1993). Making democracy work : civic traditions in modern Italy. Princeton,
N.J.: Princeton University Press.
Putnam, R. D. (1995). Bowling Alone: America’s Declining Social Capital. Journal of
Democracy, 6(1), 65–78. https://doi.org/10.1353/jod.1995.0002
Putnam, R. D. (2000). Bowling alone : the collapse and revival of American community. New
York: Simon & Schuster.
Qi, D., & Tang, V. M. Y. (2015). Social Assistance Programs and Child Poverty Alleviation -
A Comparison Between Hong Kong and Mainland China. Asian Social Work and Policy
Review, 9(1), 29–44. https://doi.org/10.1111/aswp.12049
Rating and Valuation Department. (2019). Property Market Statistics 2019. Retrieved
January15, 2017, from http://www.rvd.gov.hk/en/property_market_statistics/index.html
Ravallion, M. (2003). The debate on globalization, poverty and inequality: why measurement
matters. International Affairs, 79(4), 739–753.
https://doi.org/10.1093/acprof:oso/9780199558032.003.0002
Reichardt, C. S., & Rallis, S. F. (1994). Qualitative and quantitative inquiries are not
incompatible: A call for a new partnership. New Directions for Program Evaluation,
1994(61), 85–91. https://doi.org/10.1002/ev.1670
Ren, H., Yuan, N., & Hu, H. (2019). Housing quality and its determinants in rural China: a
structural equation model analysis. Journal of Housing and the Built Environment,
34(1), 313–329. https://doi.org/10.1007/s10901-018-9629-y
Rentfrow, P. J., Mellander, C., & Florida, R. (2009). Happy States of America: A state-level
analysis of psychological, economic, and social well-being. Journal of Research in
Personality, 43(6), 1073–1082. https://doi.org/10.1016/j.jrp.2009.08.005
Rex, J. (1971). The Concept of Housing Class and the Sociology of Race Relations. Race &
Class, 12(3), 293–301. https://doi.org/10.1177/030639687101200303
Ricardo, D. (2001). On the Principles of Political Economy and Taxation. London: Electric
Book Co.
Ringen, S. (1988). Direct and Indirect Measures of Poverty. Journal of Social Policy, 17(3),
351–365.
397
Ritchie, J., & Lewis, J. (2003). Qualitative research practice: A guide for social students and
researchers. London: Sage.
Ritchie, J., Lewis, J., Nicholls, C. M., & Ormston, R. (2014). Qualitative research practice A
guide for social science students and researchers (2nd ed.). Los Angeles: SAGE.
Ritzer, G. (2004). Encyclopedia of social theory. Thousand Oaks: SAGE Publications.
Rodero-Cosano, M. L., Garcia-Alonso, C. R., & Salinas-Perez, J. A. (2014). A Deprivation
Analysis for Andalusia( Spain): An Approach Based on Structural Equations. Social
Indicators Research, 115, 751–765. https://doi.org/10.1007/s11205-012-0226-2
Rojas, M. (2008). Experienced Poverty and Income Poverty in Mexico: A Subjective Well-
Being Approach. World Development, 36(6), 1078–1093.
https://doi.org/10.1016/j.worlddev.2007.10.005
Room, G. (1995). Beyond the threshold : the measurement and analysis of social exclusion.
Bristol: Policy Press.
Rorty, R. (1991). Objectivity, relativism and truth. Cambridge: Cambridge University Press.
Rorty, R. (1999). Philosophy and social hope. New York: Penguin.
Rowntree, S. (1901). Poverty. A Study of Town Life. London: Routledge.
Royce, E. (2015). The Social System and Poverty. In Poverty and Power: A Structural
Perspective on American Inequality (pp. 187–213). USA: Rowman & Littlefield.
Rubin, A., & Babbie, E. R. (2017). Research methods for social work (9th ed.). Boston, MA:
Cengage Learning.
Rudolf, R., & Potter, C. (2015). Housing and happiness : Subjective well-being and
residential environment in Korea. Journal of Korea Planners Association, 50(7), 55.
https://doi.org/10.17208/jkpa.2015.11.50.7.55
Sale, J. E. M., Lohfeld, L. H., & Brazil, K. (2002). Revisiting the quantitative-qualitative
debate: implications for mixed-methods research. Quality & Quantity, 36(1), 43–53.
Sandbæ k, M. (2013). Child poverty in a rich welfare state. Child Indicators Research, 6(1),
Sandelowski, M. (1995). Sample size in qualitative research. Research in Nursing & Health,
18(2), 179–183. https://doi.org/10.1002/nur.4770180211
Sassen, S. (1998). Globalization and its discontents. New York: New Press.
Sassen, S. (2011). Cities in a World Economy (4th ed.). London: Sage.
Sato, H. (2006). Housing inequality and housing poverty in urban China in the late 1990s.
China Economic Review, 17(1), 37–50. https://doi.org/10.1016/j.chieco.2004.09.005
Saunders, P. (1984). Beyond housing classes: the sociological significance of private property
rights in means of consumption. International Journal of Urban and Regional Research,
8(2), 202–227. https://doi.org/10.1111/j.1468-2427.1984.tb00608.x
Saunders, P. (2017). Housing costs, poverty and inequality in Australia. Housing Studies,
398
32(6), 742–757. https://doi.org/10.1080/02673037.2016.1229757
Saunders, P., Wong, H., & Wong, W. P. (2014a). Deprivation and poverty in Hong Kong.
Social Policy and Administration, 48(5), 556–575. https://doi.org/10.1111/spol.12042
Saunders, P., Wong, H., & Wong, W. P. (2014b). Signposting disadvantage - social exclusion
in Hong Kong. Journal of Asian Public Policy, 7(1), 3–17.
Sayer, A. (2000). Realism and social science. London: SAGE Publications.
Sayer, A. (2010). Method in social science: a realist approach (2nd ed.). New York:
Routledge.
Schultz, T. (1961). Investment in Human Capital. The American Economic Review, 51(1), 1–
17.
Schwartz, H., & Seabrooke, L. (2008). Varieties of Residential Capitalism in the International
Political Economy: Old Welfare States and the New Politics of Housing. Comparative
European Politics, 6(3), 237–261. https://doi.org/10.1057/cep.2008.10
Scotland, J. (2012). Exploring the Philosophical Underpinnings of Research: Relating
Ontology and Epistemology to the Methodology and Methods of the Scientific,
Interpretive, and Critical Research Paradigms. English Language Teaching, 5(9), 9–16.
Seidman, I. (2013). Interviewing as qualitative research: A guide for researchers in
education and the social sciences (4th ed.). New York: Teachers College Press.
Sen, A. (1985). Commodities and capabilities. New York: Oxford University Press.
Sen, A. (1992). Poverty and Affluence. In Inequality reexamined (pp. 102–116). Oxford:
Oxford University Press.
Sen, A. (1999). Development as freedom. New York: Knopf.
Sen, A. (2000). Social exclusion : concept, application, and scrutiny. Retrieved from
https://think-asia.org/handle/11540/2339
Shannon-Baker, P. (2016). Making Paradigms Meaningful in Mixed Methods Research.
Journal of Mixed Methods Research, 10(4), 319–334.
Shek, D. T. L. (2003). Economic Stress, Psychological Well-Being and Problem Behavior in
Chinese Adolescents with Economic Disadvantage. Journal of Youth and Adolescence,
32(4), 259–266. https://doi.org/10.1023/A:1023080826557
Shek, D. T. L. (2005). Perceived Parental Control Processes, Parent - Child Relational
Qualities, and Psychological Well-Being in Chinese Adolescents With and Without
Economic Disadvantage. The Journal of Genetic Psychology, 166(2), 171–188.
https://doi.org/10.3200/GNTP.166.2.171-188
Shek, D. T. L., & Lin, L. (2014). Personal Well-Being and Family Quality of Life of Early
Adolescents in Hong Kong: Do Economic Disadvantage and Time Matter? Social
Indicators Research, 117(3), 795–809. https://doi.org/10.1007/s11205-013-0399-3
399
Shildrick, T., & Rucell, J. (2015). Sociological Perspectives on Poverty. York.
Shuey, E. A., Leventhal, T., & Coley, R. L. (2016). Housing Characteristics over Time:
Identifying Patterns for Low-Income Families. Journal of Poverty, 20(1), 102–125.
Silver, H. (1994). Social Exclusion and Social Solidarity: Three Paradigms. International
Labour Review, 133(5–6), 531–578.
Siposné Nándori, E. (2014). Interpretation of Poverty in St. Louis County, Minnesota.
Applied Research in Quality of Life, 9(3), 479–503. https://doi.org/10.1007/s11482-013-
9245-7
Smart, A. (2001). Unruly Places: Urban Governance and the Persistence of Illegality in Hong
Kong’s Urban Squatter Areas. American Anthropologist, 103(1), 30–44.
https://doi.org/10.1525/aa.2001.103.1.30
Smart, A., & Lam, K. (2009). Urban conflicts and the policy learning process in Hong Kong:
urban conflict and policy change in the 1950s and after 1997. Journal of Asian Public
Policy, 2(2), 190–208. https://doi.org/10.1080/17516230903027906
Smart, A., & Lee, J. K. (2003a). Financialization and the Role of Real Estate in Hong Kong’s
Regime of Accumulation. Economic Geography, 79(2), 153–171.
https://doi.org/10.1111/j.1944-8287.2003.tb00206.x
Smart, A., & Lee, J. K. (2003b). Housing and regulation theory. In Housing and Social
Change: East-West Perspectives (p. 277). New York: Routledge.
Smith, D. P. (2012). The social and economic consequences of housing in multiple
occupation ({HMO}) in {UK} coastal towns: geographies of segregation - Smith - 2011
- Transactions of the Institute of British Geographers - Wiley Online Library. Retrieved
from http://onlinelibrary.wiley.com/doi/10.1111/j.1475-5661.2011.00487.x/pdf
Somers, M. R., & Block, F. (2005). From Poverty to Perversity: Ideas, Markets, and
Institutions over 200 Years of Welfare Debate. American Sociological Review, 70(2),
Somerville, P. (1998). Explanations of social exclusion: Where does housing fit in? Housing
Soyinka, O., & Siu, K. W. M. (2018). Urban informality, housing insecurity, and social
exclusion; concept and case study assessment for sustainable urban development. City,
Culture and Society, 15, 23–36. https://doi.org/10.1016/j.ccs.2018.03.005
Stephens, M., & Leishman, C. (2017). Housing and poverty: a longitudinal analysis. Housing
Stephens, M., & vanSteen, G. (2011a). ‘Housing Poverty’ and Income Poverty in England
and The Netherlands. Housing Studies, 26(7–8), 1035–1057.
Stephens, M., & vanSteen, G. (2011b). “Housing Poverty” and Income Poverty in England
400
and The Netherlands. Housing Studies, 26(7–8), 1035–1057.
Stone, M. E. (1993). Shelter poverty: new ideas on housing affordability. Philadelphia:
Temple University Press.
Stone, M. E. (2006a). A housing affordability standard for the UK. Housing Studies, 21(4),
Stone, M. E. (2006b). Housing Affordability : One-Third of a Nation Shelter-Poor. In A Right
to Housing: Foundation for a New Social Agenda (pp. 38–60). Philadelphia: Temple
Uniersity Press.
Stone, M. E. (2006c). What is housing affordability? The case for the residual income
approach. Housing Policy Debate, 17(1), 151–184.
Stone, M. E., Burke, T., & Ralston, L. (2011). The Residual Income Approach to Housing
Affordability: The Theory and the Practice (No. AHURI Positioning Paper No. 139).
Melbourne.
Strauss, A., & Corbin, J. (1990). Basics of qualitative research: Grounded theory procedures
and techniques. Newbury Park, Calif.: SAGE Publications.
Sun, Y., Phillips, D. R., & Wong, M. (2018). A study of housing typology and perceived age-
friendliness in an established Hong Kong new town: A person-environment perspective.
Geoforum, 88, 17–27. https://doi.org/10.1016/j.geoforum.2017.11.001
Sunega, P., & Lux, M. (2016). Subjective perception versus objective indicators of
overcrowding and housing affordability. Journal of Housing and the Built Environment,
Sung, Y. T., & Pan, P. (2010). Applications of mixed methods research in educational studies
(in Chinese). Journal of Research in Education Sciences, 55(4), 97–130.
Tabachnick, B. G., & Fidell, L. S. (2019). Using multivariate statistics (7th ed.). Boston:
Pearson Education.
Tam, M. W. Y., Hui, E., & Zheng, X. (2010). Residential mortgage default behaviour in
Hong Kong. Housing Studies, 25(5), 647–669.
Tang, K.-L. (1998). East Asian Newly Industrializing Countries: Economic Growth and
Quality of Life. Social Indicators Research, 43(1), 69–96.
Tashakkori, A., & Teddlie, C. (1998). Mixed methodology: Combining qualitative and
quantitative approaches. Thousand Oaks, Calif.: SAGE Publications.
Taylor, M. (1998). Combating the social exclusion of housing estates. Housing Studies,
Teddlie, C., & Tashakkori, A. (2003). Major Issues and Controversies in the Use of Mixed
401
Methods in the Social and Behavioral Sciences. In C. T.Abbas Tashakkori (Ed.),
Handbook of mixed methods in social and behavioral research (pp. 3–50). Thousand
Oaks, Calif.: SAGE Publications.
Terry, R. (2011). Improving housing outcomes for young people: practical ideas. York.
Thalmann, P. (1999). Identifying households which need housing assistance. Urban Studies,
36(11), 1933–1947.
Thalmann, P. (2003). “House poor” or simply ’poor’ ? Journal of Housing Economics, 12,
291–317. https://doi.org/10.1016/j.jhe.2003.09.004
Therborn, G. (2013). The killing fields of inequality. Cambridge: Polity.
Thomas, E., & Magilvy, J. K. (2011). Qualitative Rigor or Research Validity in Qualitative
Research. Journal for Specialists in Pediatric Nursing, 16(2), 151–155.
https://doi.org/10.1111/j.1744-6155.2011.00283.x
Tim Wong, L. (2018). Tiny affordable housing in Hong Kong. Indoor and Built Environment,
27(9), 1159–1161. https://doi.org/10.1177/1420326X18792159
Ting, W. F. (2012). The Effectiveness of CIIF Projects in Social Capital Development in Tin
Shui Wai. Hong Kong.
Touraine, A. (1971). The post-industrial society: tomorrow’s social history: classes, conflicts
and culture in the programmed society. New York: Random House.
Townsend, P. (1962). The meaning of poverty. The British Journal of Sociology, 13(3), 210–
227.
Townsend, P. (1979a). Introduction : Concepts of Poverty and Deprivation. In Poverty in the
United Kingdom: a survey of household resources and standard of living (pp. 31–60).
London: Allen Lane.
Townsend, P. (1979b). Poverty in the United Kingdom : a survey of household resources and
standards of living. London: Allen Lane.
Townsend, P. (1979c). Theories of Poverty. In Poverty in the United Kingdom: a survey of
household resources and standard of living (pp. 61–92). London: Allen Lane.
Townsend, P. (1987). Deprivation. Journal of Social Policy, 16(2), 125–146.
Townsend, P. (2002). Poverty, social exclusion and social polarisation: the need to construct
an international welfare state. In World poverty: New policies to defeat an old enemy
(pp. 3–17). Bristol: Policy Press.
Townsend, P., & Gordon, D. (2002). Introduction: The human condition is structurally
unequal. In World poverty: New policies to defeat an old enemy (pp. xi–xxii). Bristol:
Policy Press.
Tse, R. Y. C., & Webb, J. R. (2000). Public versus private real estate in Hong Kong. Journal
of Real Estate Portfolio Management, 6(1). Retrieved from
https://search.proquest.com/docview/197819189/fulltextPDF/741B88A420254E03PQ/1
?accountid=10371
402
Tuli, F. (2011). The Basis of Distinction Between Qualitative and Quantitative Research in
Social Science: Reflection on Ontological, Epistemological and Methodological
Perspectives. Ethiopian Journal of Education and Sciences, 6(1).
https://doi.org/10.4314/ejesc.v6i1.65384
Tunstall, R., Bevan, M., Bradshaw, J., Croucher, K., Duffy, S., Hunter, C., …Wilcox, S.
(2013). The Links between housing and poverty: an evidence review. York.
United Nations. (2017). Sustainable development: poverty. Retrieved June23, 2017, from
http://www.un.org/sustainabledevelopment/poverty/
Valença, M. M. (2015, October 1). Social rental housing in HK and the UK: Neoliberal
policy divergence or the market in the making? Habitat International. Elsevier Ltd.
vanPraag, B., & Ferrer-i-carbonell, A. (2006). A Multi-dimensional Approach to Subjective
Poverty. Amsterdam.
Vartanian, T. P. (2011). Secondary data analysis. Oxford: Oxford University Press.
Vere, J. P. (2010). Special topic enquiry on earnings mobility. Hong Kong.
Wacker, J. (1998). A definition of theory: research guidelines for different theory-building
research methods in operations management. Journal of Operations Management, 16(4),
361–385. https://doi.org/10.1016/S0272-6963(98)00019-9
Wade, R. H. (2004). Is Globalization Reducing Poverty and Inequality? World Development,
32(4), 567–589. https://doi.org/10.1016/j.worlddev.2003.10.007
Wagle, U. (2002). Rethinking poverty: definition and measurement. International Social
Science Journal, 54(171), 155–165. https://doi.org/10.1111/1468-2451.00366
Wallace, A. (2016). Homeowners and Poverty : A Literature Review. York.
Warren, E. J. (2018). Housing affordability and material hardship: Does affordability
measurement matter? Journal of Poverty, 22(3), 228–247.
Warren, M. R., Thompson, J. P., & Saegert, S. (2001). The role of social capital in combating
poverty. Social Capital and Poor Communities, 3, 1–28.
Welshman, J. (2013). Underclass : a history of the excluded since 1880. London:
Bloomsbury.
Whitehead, C., Monk, S., Clarke, A., Holmans, A., & Markkanen, S. (2008). Measuring
Housing Affordability : A Review of Data Sources. Cambridge.
WHO. (2017). Definition of Poverty. Retrieved June23, 2017, from
http://www.who.int/topics/poverty/en/
Wong, H. (2000). The failure of social security in alleviating poverty in Hong Kong. Asian
Pacific Journal of Social Work and Development, 10(2), 86–100.
Wong, H. (2005). The quality of life of Hong Kong’s poor households in the 1990s: Levels of
403
expenditure, income security and poverty. Social Indicators Research, 71(1), 411–440.
Wong, H. (2011). Quality of Life of Poor People Living in Remote Areas in Hong Kong.
7
Wong, H. (2015). Is Poverty Eradication Impossible? A Critique on the Misconceptions of
the Hong Kong Government. China Review, 15(2), 147–169. Retrieved from
http://www.jstor.org/stable/43710029
Wong, H., & Chan, S. M. (2019). The impacts of housing factors on deprivation in a world
city : The case of Hong Kong. Social Policy and Administration, 53(6), 1–17.
https://doi.org/10.1111/spol.12535
Wong, M. M. C., Ma, J. L. C., & Chan, L. C. L. (2017). The impact of poverty on children in
out-of-home care services in a Chinese context and the application of multiple family
group therapy to enrich their family lives. Children and Youth Services Review, 1–9.
https://doi.org/10.1016/j.childyouth.2017.06.013
Woolcock, M. (2002). Social Capital in Theory and Practice: Reducing Poverty by Building
Partnerships between States, Markets and Civil Society. In Social capital and poverty
reduction (pp. 20–44). Geneva: UNESCO.
Wu, W., Stephens, M., Du, M., & Wang, B. (2019). Homeownership, family composition and
subjective wellbeing. Cities, 84(August 2018), 46–55.
https://doi.org/10.1016/j.cities.2018.07.004
Yau, Y. (2012). Ruling out trouble: Unacceptable behaviour and its control in Hong Kong’s
public housing. Habitat International, 36(1), 11–19.
Yi, J., & Zhang, J. (2010). The effect of house price on fertility: Evidence from Hong Kong.
Economic Inquiry, 48(3), 635–650. https://doi.org/10.1111/j.1465-7295.2009.00213.x
Yip, N. M. (1995). Housing affordability in England. University of York.
Yip, N. M. (2013). Challenges to public housing in post-colonial Hong Kong. In The Future
of Public Housing (pp. 65–83). Berlin: Springer.
Yip, N. M., & Forrest, R. (2002). Property owning democracies? Home owner corporations
in Hong Kong. Housing Studies, 17(5), 703–720.
Yip, N. M., & Forrest, R. (2014). Choice or Constraint? Exploring solo- living for young
households in Hong Kong (Urban Research Group CityU on Cities Working Paper
Series No. No.1/2014). Hong Kong.
Yip, N. M., Forrest, R., & LaGrange, A. (2007). Cohort trajectories in Hong Kong’s housing
system: 1981-2001. Housing Studies, 22(1), 121–136.
404
Yip, N. M., & LaGrange, A. (2006). Globalization, de-industrialization and Hong Kong’s
private rental sector. Habitat International, 30(4), 996–1006.
Yip, N. M., & Lau, K. Y. (2002). Setting rent with reference to tenants’ affordability: Public
housing rent policy in Hong Kong. Journal of Housing and the Built Environment,
17(4), 409–418. https://doi.org/10.1023/A:1021118714418
Yip, P. S. F., Wong, J. H. K., Li, B. Y. G., Zhang, Y., Kwok, C. L., & Chen, M. N. (2016).
Assessing the Impact of Population Dynamics on Poverty Measures: A Decomposition
Analysis. Social Indicators Research, 134(2), 531–545. https://doi.org/10.1007/s11205-
016-1434-y
Yiu, C. Y. (2011). A spatial portfolio theory of household location choice. Journal of
Transport Geography, 19(4), 584–590. https://doi.org/10.1016/j.jtrangeo.2010.06.012
Yiu, M. S., Yu, J., & Jin, L. (2013). Detecting bubbles in Hong Kong residential property
market. Journal of Asian Economics, 28, 115–124.
https://doi.org/10.1016/j.asieco.2013.04.005
Yu, S. W. K. (2008). The ideological orientation of the New Dawn Project. The Hong Kong
Journal of Social Work, 42(1/2), 105–117.
Yung, B. (2008). Hong Kong’s housing policy a case study in social justice. Hong Kong:
Hong Kong University Press.
Yung, B., & Lee, F. ping. (2014). “Equal right to housing” in Hong Kong housing policy:
perspectives from disadvantaged groups. Journal of Housing and the Built Environment,
Zhang, F., Zhang, C., & Hudson, J. (2018). Housing conditions and life satisfaction in urban
China. Cities, 81, 35–44. https://doi.org/10.1016/j.cities.2018.03.012
Zheng, L., Zheng, X., & Chen, K. (2017). Incomplete privatization of public rental housing in
Hong Kong. Land Use Policy, 67(May), 537–545.
https://doi.org/10.1016/j.landusepol.2017.07.007
405
Appendix
Appendix A Operationalisation of key housing variables

The main independent variables (IVs) in this research includes housing location and
housing type and. The operationalization of these concepts in census and SPPR dataset will
be illustrated as below:
Housing location
For analysing housing location, there are different types of location division in Hong
Kong. The location of living space normally divided by 18 districts based on district council
categories. The census and SPPR data sets used this categorization in survey. Scholars (Hui et
al. 2007; Yiu, 2011) used the distance to the Central Business District (CBD) to group district
for analytical purposes. The 18 districts in Hong Kong are grouped as ‘near CBD’, ‘mid-
range from CBD’ and ‘far from CBD’.
■ Near CBD: Central and Western, Wan Chai, Yau Tsim Mong, Sham Shui Po, Kowloon
City, Wong Tai Sin, Kwun Tong, Kwai Tsing
■ Mid-range from CBD: Eastern, Southern, Sai Kung, Tsuen Wan, Sha Tin
■ Far from CBD: Tuen Mun, Yuen Long, North, Tai Po, Islands
Housing type
In the data sets of Census and SPPR, there are different grouping in the variable of ‘type
of housing’ and ‘tenure of accommodation’. New categorizations of housing types are formed
based on the data sets. The groups ‘public rental housing’, ‘public subsidized housing’,
‘private housing’ and ‘other’ are formed based on ‘type of housing’. The housing type
groupings are further divided based on ‘tenure of accommodation’.
Type of housing
New group Census SPPR

Public Rental - Public rental housing units - Public rental housing
Housing - Housing Authority Interim Housing
Units
Public - Housing Authority Tenants Purchase
Subsidised Scheme (TPS) without premium paid
Housing - Subsidised Sale Flats in Housing
Authority/ Housing Society without
premium paid
406
Private - Former Subsidised Sales Flats with
Housing premium paid
- Ordinary Private Residential Flats
- Villas/ Bungalows/ Modern Village
Houses
Other - Simple Stone Structures/ Traditional - Other
Village Houses
- Staff Quarters
- Quarters in Non-residential Building
- Collective Living Quarters
- Temporary quarters
Tenure of accommodation
New group Census SPPR
Owner, with - Owner-occupier – with mortgage - Owner-occupier, with
mortgage mortgage payment or
loan repayment
Owner, - Owner-occupier - without mortgage - Owner-occupier,
without without mortgage
mortgage payment or loan
repayment
Tenant - Sole tenant - Rented
- Co-tenant accommodation
- Main tenant
- Sub-tenant
Other - Provided by employer - Provided by employer
- Other
407
Appendix B Measurement of residual housing poverty
The public rental housing (PRH) was supposed to protect the basic living of residents.
The amount of income limits for applying public housing was consisted of 3 components:
‘housing cost’, ‘non-housing cost’ and ‘contingency’. The ‘non-housing cost’ was be used for
calculation of basic living. Regarding to the mechanism of Hong Kong Housing Authority,
the income limits are derived using a household expenditure approach. The mechanism is
summarized in the following figure.
The non-housing costs are ‘determined with reference to the latest Household
Expenditure Survey (HES) conducted by the Census and Statistics Department (C&SD), and
adjusted by the latest movement in the Consumer Price Index (CPI)(A)(excluding housing
costs), or the change in the nominal wage index as the income factor, whichever is higher’.
Following the established mechanism, the review of income limits adopted the non-
housing expenditure statistics from the latest HES of the lower half expenditure group among
tenant households in the private sector, excluding those households comprising solely of
elderly or non-working members. The ’housing cost’, ‘non-housing cost’ and ‘contingency’
by household size are listed in the following figure.
408
Those households with AH − B < 0 are counted as residual housing poverty cases, where
AH is after housing income, and B is basic living standard. The basic living standard (B) by
household size are count as following:
■ 1-person household: $5503
409
Appendix C Background of individual interview informants
Case Sex Age Family composition Housing Marital Employment Living area Family income Rent
Label Type status Status (square feet) (HKD) (HKD)
A F 28 4-person family, 2 children Subdivided Married Housewife 150 16000 5600
Flat
B F 59 2-person family, couple Subdivided Married Casual 80 5000-6000 3200
Flat worker
C M 69 2-person family, couple, and Subdivided Married Full-time 80 11000 4000
grandson Flat Cleaner
D F 50 4-person family, 2 children Subdivided Married Housewife 200 17000 6200
Flat
E F 63 1-person Cubicle Widowed Retired 40 5000 2500
F F 46 2-person family, single Subdivided Divorced Casual 120 8000 4900
mother Flat worker
G F 35 4-person family, single Subdivided Separated Welfare 120 13000 5600
mother Flat recipient
H F 71 1-person Subdivided Divorced Retired 100 6500 6100
Flat
I M 33 4-person family, 2 children Rooftop Married Maintenance 150 18000 5600
housing worker
J F 29 2-person family, single Subdivided Single Welfare 100 9000 5800
mother Flat recipient
K M 54 1-person Cubicle Single Welfare 30 6700 1800
410
recipient
L M 60 3-person family, couple, and Subdivided Married Casual 100 9000 5200
daughter Flat worker
M F 35 4-person family, 2 children Subdivided Married Housewife 200 14000 7100
Flat
N F 72 1-person Subdivided Widowed Retired 80 6000 4300
Flat
O M 63 3-person family, couple, and Cubicle Married Retired 80 15000 5500
daughter
P M 42 4-person family, 2 children Subdivided Married Full-time 100 15600 3950
Flat Cook
Q F 38 4-person family, 2 children Public Married Housewife 350 10000-20000 1500
housing
R M 36 4-person family, 2 children Public Married Unemployed 400 10000 3200
housing
S F 35 3-person family, couple, and Public Married Housewife 300 15000 2000
daughter housing
T M 45 1-person Singleton Divorced Disabled 60 6500 1500
hostel
411
Appendix D Consent Form
Project Title: The impact of housing on poverty situation in Hong Kong: the study on
deprivation and social exclusion of residents
Name of principal investigator: CHAN Siu Ming
1) This is a study conducted by the PhD student of Social Work Department of the Chinese
University of Hong Kong, Mr. CHAN Siu Ming.
2) This questionnaire will be conducted face-to-face by researcher and the time involved
will be about 60 to 90 minutes.
3) No potential risks on psychological stress or significant discomfort are known.
4) This research aims at examining the impact of housing factors on poverty situation of
residents, which also enable better social policy and practices for the socially
disadvantaged.
5) The treatment of data and personal data arising from the study are kept in strictly
confidential ways and will be used in this research only. No individual or household data
will be released to other parties.
6) The participation of this survey should be voluntary. Participants have the right to refuse
taking part in the interview and to withdraw from the study at any time.
7) Participants have the right to ask any questions during the interview.
8) Should the interviewee has any enquiries, he or she can contact the researcher Mr. CHAN
Siu Ming at 60960921 or email: [email protected]
Statement of Consent
I have read the above information and have received answers to any questions I
asked. I understand the nature of this study and the information provided will be used for
research related purpose only. I agree that the information collected will be kept by the
researcher in strictly confidential ways and for at least five years beyond the end of the
study.
By signing below, I indicate my consent to take part in the study.
Name of Participant: _________________
Signature of Participant: _________________ Date: ___________
Name of Person Obtaining Consent: _________________
Signature of Person Obtaining Consent: ________________ Date: ___________
412
Appendix E Interview Guide
Background
1. Introduction
■ Self-introduction
■ Introduce background and purpose of research
■ Introduce interview process and confidential issue
Living experience and daily living
2. Living experience (in house)
■ Would you describe your daily living experience? How do you describe yourself as a
(housing type) resident?
■ What are the difficulties you face in your living condition? How do you deal with?
■ How you describe your relationships with neighbour, landlord (if any) and community?
■ Where did you live before? Why do you live in this house? How do you describe your
path of housing in the past years? How is the experience?
3. Daily living
■ How do you describe your monthly income and expense?
■ How do you distribute your income and spending? Any priority or criteria?
■ How do you describe your daily living and social life? How do you feel?
■ What are the difficulties you face in daily living? How do you deal with?
Housing and poverty
4. Housing
■ How do you describe your current housing affordability? What is a reasonable housing
affordability? And why?
■ What is the impact of housing affordability on your daily live? How is your experience?
How do you feel?
■ What is the relationship about housing affordability, quality, and community? How do
you make choice among them? What is your consideration?
■ Where did you live before? Why and how did you find this living space? Do you
“choose” to live here? How do you think about choice of living space? What is your
experience?
■ Do you plan to change living space in the future? What is your consideration?
5. Poverty
■ How do you describe “poverty”? Do you describe yourself as living in poverty? Why?
What is your experience of “poverty”, if any?
■ How do you describe “deprivation”? Do you describe yourself as deprived? Why? What
is your experience of “deprivation”, if any?
■ How do you describe “social exclusion”? Do you describe yourself as socially excluded?
413
Why? What is your experience of “social exclusion”, if any?
■ How do you describe the relationship between “being poverty” and “feeling poverty”?
What is your experience of “feeling poverty”, if any?
6. Housing and poverty
■ How do you think about the relationship between housing and poverty, deprivation, and
social exclusion? What is your experience?
■ How do you experience the impact of housing on poverty, such as daily expense, level
of deprivation, relationship with neighbour and community and social exclusion?
■ If you moved to other house before, what are the change and difference? How do you
describe the change to poverty situation? How do you feel?
■ How do you experience the impact of poverty on housing, such as housing cost, housing
quality, and living community?
■ How do you describe yourself in this housing and poverty situation? What is your
experience?
7. Coping Strategies
■ Do you want to change the current housing and poverty situation? Why?
■ What have you done? Do you face any difficulties? What is your experience?
■ What do you want to do in the future? What is your consideration?
Housing and poverty (Macro perspective)
8. Housing and anti-poverty policy
■ How do you view housing policy and anti-poverty policy in Hong Kong?
■ How do you describe the relationship between housing anti-poverty policy and your
current living?
■ Do you have any suggestion for changing current policy? What will you do in making
such change?
Other
9. Other
■ What do you consider when you choose the house? What are the limitations? How do
you make your choice?
■ How do you deal with landlord when you face rental issue or housing problems?
■ Have you applied public housing? If so, why is it important to you? What are the roles
of government?
10. Ending
■ Do you have anything want to ask me?
■ Do you have anything want to share?
■ Give thanks to informants
414

2020 - Chan Siu Ming-2

Uploaded by

Copyright:

Available Formats

2020 - Chan Siu Ming-2

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

2020 - Chan Siu Ming-2

Uploaded by

Copyright:

Available Formats

Impact of Housing on Poverty Situation in Hong Kong:

An Examination on Deprivation, Social Exclusion, and Subjective Poverty

CHAN, Siu Ming

A Thesis Submitted in Partial Fulfilment

of the Requirements for the Degree of

The Chinese University of Hong Kong

Professor TO Siu Ming (Chair)

Professor WONG Hung (Thesis Supervisor)

Professor LAM Ching Man (Committee Member)

Professor YIP Ngai Ming (External Examiner)

poverty among Hong Kong citizens.

首先，本論文分析 1996 至 2016 年的人口普查數據，計算出三條不同的房屋貧窮線，

referred the cases for my study.

supervisor, this study would not have been completed.

I am likewise expressing my appreciation to my thesis committee members, namely,

comments and recommendations to improve my thesis. I am also grateful to my previous

Cheung Wong. Their constructive comments in my proposal formulation enabled me to

clarify my research direction and improve the research methodology.

knowledge of housing and poverty. Evidently, I am deeply indebted to their assistance.

anti-poverty movement and my PhD study. Particularly, I am thankful to Fernando Cheung,

Department of Social Work, The Chinese University of Hong Kong.

I am likewise expressing my gratitude to my parents and family for their unconditional

Her love, support, and acceptance enabled me to complete my PhD study.

I thank God for all the graces that I have received.

1.1 Background and Rationale 16

1.2 Research objectives 25

1.3 Research questions 26

1.4 Overview of the thesis 29

Chapter 2 Theoretical Review of Housing and Poverty 30

2.3 Housing poverty 75

2.4 Conclusion: Theoretical gaps and significance of research 109

Chapter 3 Empirical Review of Housing and Poverty in Hong Kong 111

3.1 Academic literature review 112

3.2 Poverty Situation in Hong Kong 117

3.3 Housing situation in Hong Kong 129

3.6 Conclusion: Empirical study gaps and significance of research 154

Chapter 4 Philosophical Foundation, Research Method and Design 156

4.1 Philosophical foundation of social research 157

4.2 Research method 169

4.3 Research design 182

Chapter 5 Quantitative Analysis: Part 1 192

5.1 Introduction of quantitative analysis 192

5.2 Housing poverty situation in Hong Kong (Census) 193

5.3 Discussion for census data analysis 217

6.1 The impact of housing factors on non-income poverty 236

6.2 The mediating effect of housing factors on non-income poverty 254

6.3 Findings and Discussion 273

6.4 Short conclusion of the quantitative analysis 287

Chapter 7 Qualitative Analysis 289

7.1 Background 290

7.2 Three Stories of Housing Poor 294

7.4 ‘I am Poor, because…’: Critical Life Events 315

7.5 ‘I am Poor, so…’: The Impact of Poverty on Housing 321

7.6 ‘Housing Makes Me Poor’: Impact of Housing on the Non-Income Poverty

7.8 Conclusion of Qualitative Analysis 348

8.1 Conclusive response to the research questions 352

8.2 Theoretical Contributions and Significance 358

8.3 Empirical and Policy Contributions and Significance 362

8.4 Reflection 365