[Link] This is the Pre-Published Version.

1 Bridging the Gap between Sustainable Housing and Affordable Housing: The Required
2 Critical Success Criteria (CSC)
3
4 Albert P.C. Chan 1 and Michael Atafo Adabre 2
5
6
7 Abstract
8 Studies on specific critical success criteria (CSC) for performance measurement of sustainable
9 affordable housing projects are limited. This study aims to identify and classify the various
10 CSC from the views of affordable housing experts around the world. 21 CSC were identified
11 from a comprehensive literature review followed by a questionnaire survey on the identified
12 21 CSC. With 51 responses, the data were analyzed. Factor analysis indicated that the various
13 CSC can be grouped into six components: household satisfaction CSC, stakeholders’
14 satisfaction CSC, house operation cost CSC, time measurement CSC, location affordability
15 cost CSC and quality-related CSC. Practically, the findings of this study can serve as a guide
16 for assessing the performance of affordable housing projects as well as serving as a guide to
17 developers, NGOs and government agencies in the allocation of resources for the provision of
18 sustainable affordable housing. Future study would investigate the interrelationship between
19 critical success criteria and critical success factors for sustainable affordable housing.
20
21
22 Keywords: Critical Success Criteria; Affordable housing; Sustainable housing; Affordability
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42

1
Co-author’s address: Prof. Albert P.C. Chan, Building and Real Estate Department, Hong
Kong Polytechnic University, Hong Kong. E-mail address: [Link]@[Link]
2
Corresponding author’s address: Michael Atafo Adabre, Building and Real Estate
Department, Hong Kong Polytechnic University, Hong Kong: Tel: +85266450743
E-mail address: 17902405r@[Link]

© 2019. This manuscript version is made available under the CC-BY-NC-ND 4.0 license [Link]
43 1. Introduction
44 Housing is among the basic social conditions which define the quality of life and wellbeing of
45 the citizens of any nation. However, in a constantly changing and urbanizing world, housing
46 supply has not been able to adequately meet demand (Gan et al., 2017). Corollaries of the
47 acceleration in urbanization are increasing affordability challenges among low income earners
48 noted in both developed and developing countries (Dezhi et al., 2016). For instance, it has been
49 estimated that the number of poor people living in shantytowns and sub-standard housing in
50 developing countries is 828 million. Speculations are that this number will increase to 1.4
51 billion by 2020 (Al-Saadi and Abdou, 2016; Desai, 2012; Gan et al., 2017). A survey conducted
52 among some developed countries such as USA, Australia, Singapore, Hong Kong (China),
53 New Zealand and Ireland revealed that out of 293 housing markets surveyed, only 63 were
54 considered affordable (Cox et al., 2017). In general, the anticipation of the world’s population
55 growth from 3.6 billion to 6.3 billion in 2050 is an indication that more housing will be required
56 to meet the mounting housing needs (Golubchikov and Badyina, 2012). Accordingly,
57 sustainable affordable housing remains a priority for all governments and other policy makers.
58
59 Many affordable housing policies have been initiated. However, whether the housing
60 affordability of low-income earners has been improved remains a debate (Gan et al., 2017).
61 Study by Stone (2006) has focused on the economic measure - price affordability - for
62 accessing the success or improvement of housing policies. Conversely, by solely focusing on
63 the economic measure, real estate developers, planners, architects and governments have
64 encountered challenges of low demand and abandonment of housing in the provision of
65 affordable housing (Susilawati and Armitage, 2005; Adabre and Chan, 2018). For example, in
66 a developing country China, it was stated that the average housing price-to-income ratio for
67 many major cities was 10.2 in 2013, which situated China in a group of severely unaffordable
68 housing market (Zhang et al., 2016). However, public rental housing which were less than 30%
69 of market rents were abandoned by applicants in Shenzhen, Wuhan, Nanjing, Zhengzhou and
70 Shanghai (Lin, 2012). Consequently, 90% vacancy rate was reported in the case of Shenzhen
71 (Yuan et al., 2018). In Malaysia, a study indicated the need for affordable housing for low and
72 middle-income earners (Abdul-Aziz and Kassim, 2011). Yet, affordable housing that were
73 supplied to these income categories were left vacant leading to housing overhang (Teck-Hong,
74 2012). A Similar situation of housing abandonment has been reported in a developed country
75 United Kingdom (Mulliner et al., 2013). In all these cases, the abandonments of the houses
76 were attributed to other criteria beyond price affordability. Thus, these paradoxes of housing
77 needs amidst housing overhangs buttress the fact that not all that is affordable is sustainable!
78 Therefore, bridging the gap between sustainable housing and affordable housing is exigent.
79
80 Successively, in a study conducted by Mulliner et al. (2013), it was concluded that in addition
81 to economic measures, there are non-economic criteria associated with evaluating success of
82 sustainable affordable housing projects. These economic and non-economic criteria or standard
83 are termed critical success criteria (CSC). CSC are the set of principles or standards through
84 which judgement can be made whereas critical success factors (CSF) are the set of
85 circumstances, facts or influences which affect / contribute to the results or CSC (Lim and
86 Mohammed, 1992 p.243). For instance, ‘accessibility to shops’ and ‘access to health services’
87 are examples of CSFs (factors) whereas ‘reduced commuting cost or time’ could be used as a
88 CSC (criterion / outcome) which is influenced by the CSFs. Furthermore, ‘availability of green
89 public space’ is a CSF whereas ‘household / stakeholders’ satisfaction’ and ‘quality housing’
90 could be used as CSC (Torbica and Stroh, 2001; Ahadzie et al., 2008). Moreover, ‘the
91 construction method for a housing facility’ and ‘materials used for construction’ are CSFs
92 which could influence CSC such as ‘maintainability of a housing facility’; ‘technical
2
 93 specification of a housing facility’ and ‘environmental performance of a housing facility’
94 (Torbica and Stroh, 2001; Rankin et al., 2008). Finally, ‘the type of communication among
95 project stakeholders’ could be a CSF which influences criteria such as ‘reduced occurrence of
96 disputes and litigation among project stakeholders’ and ‘technology transfer’ in construction
97 projects (Adinyira et al., 2014).
98
99 In addition to the lack of consensus on CSC (Gan et al., 2017), studies on CSC for sustainable
100 affordable housing projects are limited. As such, an investigation on CSC for sustainable
101 affordable housing projects is important for the following reasons. Knowledge on CSC is
102 required for the development of sustainable and affordable housing policies to improve the
103 current and anticipated affordability crises. Besides, real estate developers, governments and
104 international organizations need to be apprised of the effective and appropriate CSC to identify
105 affordability challenges and innovate measures for successful housing delivery. Moreover,
106 CSC serve as measures to guide developers and governments to enhance efficient allocation of
107 the limited resources to meeting the residential needs of the household (Chua et al., 1999).
108 Finally, the categorization of the various CSC will help governments and international policy
109 makers on strategies required to bridge the gap between sustainable housing and affordable
110 housing.
111
112 In the light of the background above, the main objective of this study is to identify the CSC
113 which are required to evaluate success in sustainable affordable housing projects. Therefore, a
114 literature review was conducted to identify the potential set of CSC for sustainable affordable
115 housing projects, which forms section two of this study. Then, Section Three presents a
116 thorough description of the research methodology adopted for the study. Furthermore,
117 statistical analysis of the survey responses together with discussion of results was conducted
118 in the penultimate section, Section Four. Finally, some concluding remarks are stated in Section
119 Five.
120
121 2. Literature Review
122 The identification of key project CSC is important so that construction managers, project
123 managers and policy makers can appropriately plan resource allocation (Chua et al., 1999).
124 Irrespective of the type of construction projects, the iron triangle of time, cost and quality have
125 been widely recognized as the fundamental CSC in many studies (Atkinson, 1999; Bassioni et
126 al., 2004; Chan and Chan, 2004). However, it is a fact that some determinants of success are
127 likely to be distinctive among projects. Moreover, studies have revealed that the iron triangle
128 criteria are non-exhaustive (Lim and Mohamed, 1999; Pinto and Pinto, 1991; Pocock et
129 al.,1996). Therefore, studies have been conducted to comprehensively identify CSC for project
130 monitoring and control in the construction industry (Lim and Mohammed, 1999; Baccarini,
131 1999; Ahadzie et al., 2011; Al-Tmeemy et al., 2011).
132
133 In general construction project, Lim and Mohamed (1999) explored the criteria of project
134 success from different perspectives of stakeholders. The identified criteria were grouped into
135 two categories. These included the macro and micro perspectives. Project completion and
136 satisfaction were the criteria that defined the macro viewpoint of project success while the
137 micro viewpoint was solely defined by the completion criterion. Thus, the classification by
138 Lim and Mohamed (1999) highlighted an overlap between the categories. For instance, the
139 completion criterion was common to both the macro and micro viewpoints. The other criterion
140 – satisfaction – was more focused on the owner and user of the project. Therefore, they failed
141 to provide detail criteria for construction companies or contractors (Al-Tmeemy et al., 2011).
142 In Baccarini (1999), the criteria of project success were grouped into product success and
3
143 project management success based on the goal, purpose, output and input. The product success
144 deals with goals and purpose while the project management success deals with output and
145 inputs. Although Baccarini (1999) flagged some key criteria applicable to construction
146 companies and contractors in the project management success criteria, contractors’ goals such
147 as revenue and profit, market share and competitive advantage were not explicitly stated. Based
148 on this knowledge gap, Al-Tmeemy et al. (2011) conducted a study on developing a framework
149 to categorize project success for building projects from contractors’ perspectives. While
150 maintaining the classification of Baccarini (1999), Al-Tmeemy et al. (2011) added another
151 category of success – market success. Therefore, three classes of project success were
152 identified from the study of Al-Tmeemy et al. (2011). These included: the project management
153 success which consists of adherence to quality targets, schedule and budget; the product
154 success such as customer satisfaction, functional requirement and technical specification;
155 market success such as revenue and profit, market share, reputation and competitive advantage.
156 The market success criteria emphasised on the strategic goals of construction companies.
157
158 Although the identified criteria from previous studies (Lim and Mohammed, 1999; Baccarini,
159 1999 and Al-Tmeemy et al., 2011) are comprehensive and applicable to most construction
160 projects, not all might be relevant for housing projects due to differences in project
161 characteristics. For instance, according to Ahadzie et al. (2008) on mass housing, housing
162 projects involve the construction of domestic residence. Moreover, mass housing projects are
163 speculative in nature since decisions on land acquisition, design and construction of such
164 houses are mostly made without a specific customer in mind. Therefore, with regard to housing
165 projects, Ahadzie et al. (2008) developed four clusters of CSC for mass housing projects:
166 environmental impact, customer satisfaction, quality and overall cost and time. These CSC
167 could be appropriate for affordable housing projects based on the similarities between mass
168 housing and affordable housing. Like mass housing, affordable housing projects involve the
169 construction of domestic residence and are also speculative in nature. Despite the similarities
170 in project characteristics, definitional difference between them suggests that the CSC for mass
171 housing are not comprehensive CSC for affordable housing projects. In Ahadzie et al. (2008 p.
172 678), mass housing is defined as “the design and construction of speculative standardized
173 house-units usually in the same location and executed within the same project scheme.”
174 However, “affordable housing is housing that is reasonably adequate in standard and location
175 for a lower or middle-income household and does not cost so much that such a household is
176 unlikely to be able to meet other basic living costs on a sustainable basis (National Summit on
177 Housing Affordable, 2006). The rule-of-thumb is that housing is affordable if low income
178 household spent less than 30% of their income on housing. Therefore, mass housing projects
179 are affordable provided they meet the affordability criteria / requirements. Otherwise, mass
180 housing cannot be considered affordable housing and therefore different CSC maybe required
181 for assessing the sustainability of affordable housing.
182
183 Findings of the study by Ahadzie et al. (2008) cannot be considered as complete CSC for
184 affordable housing projects. For example, price of housing and rental cost of housing in relation
185 to household income which are important criteria for affordable housing (Mulliner et al., 2013)
186 were not considered among the criteria in their study. Besides, transportation cost in relation
187 to the income of households (Isalou et al., 2014) was also not listed among the criteria identified
188 in their study. Based on these caveats, it is necessary to find out the exclusive CSC for
189 sustainable affordable housing projects. Studies have been conducted on identifying these
190 specific criteria. The traditional ratio criterion measures affordability in terms of the ratio of
191 housing cost to income. However, Chaplin et al. (1994) and Bogdon and Can (1997) stated that
192 though the ratio approach is simple to compute and widely used, it is not adequate enough to
4
193 assess the affordability situation of household. Affordability must involve whether a household
194 has enough income left over for other needs of life after paying housing bills. If the household
195 cannot meet their non-housing needs such as food, medical care and clothing at some minimum
196 level of adequacy after paying for housing bill, then the household is ‘shelter poor’. Thus,
197 unlike the ratio criterion which looks at housing affordability only as a matter of housing cost,
198 the ‘shelter poor’ or ‘residual’ approach takes into account the full amount required for housing
199 and other basic needs (Stone, 2006). However, the residual income approach and the shelter
200 poverty concept have a practical challenge of being translated into an operational affordability
201 scale. It is a problem setting the minimum standard of adequacy for non-shelter items (Bogdon
202 and Can, 1997). Moreover, the conventional ratio and residual approaches focus more on the
203 economic issues of price affordability of housing. This solely does not bridge the gap between
204 sustainable housing and affordable housing. For example, though the prices of a housing
205 facility might be affordable, it is not truly affordable if it located in a remote area with high
206 transportation cost (Golubchikov and Badyina, 2012). In a study conducted by Isalou et al.
207 (2014), it was found out that suburban household spent about 57% of their income on housing
208 and transportation which was significantly higher compared to 45% of housing and
209 transportation expenditure spent by households in the urban areas.
210
211 Yet, the price of a housing facility and transportation cost do not give a complete view of the
212 required CSC for measuring the success of sustainable affordable housing projects (Mulliner
213 et al., 2013; Gan et al., 2017). According to Mulliner et al. (2013 p. 270), to improve quality
214 of life and community sustainability, aside the economic assessment criteria, “the
215 environmental and social sustainability of housing must be taken into consideration”. Using
216 the COPRAS method of Multi-Criteria Decision Making (MCDM), twenty-one criteria were
217 used to assess the affordability of an area. These criteria in descending order of their mean
218 scores include: house price in relation to income, rental costs in relation to income, interest
219 rates and availability of mortgages, social and private rented accommodation availability,
220 homeownership products availability, access to employment opportunities, public transport
221 services accessibility, quality school accessibility, access to shops, access to health services,
222 access to child care, open green public space accessibility, quality of housing, energy efficiency
223 of housing, availability of waste management facilities, appeal of neighborhood area,
224 deprivation in area and presence of environmental problems. It was concluded that considering
225 social and environmental criteria can critically influence the estimation of the affordability in
226 an area as compared to focusing solely on the financial criteria. Although Mulliner et al. (2013)
227 broadened the scope of sustainable affordable housing criteria and contributed significantly,
228 they failed to differentiate critical success criteria (CSC) from critical success factors (CSFs).
229 Out of the twenty-one criteria, only five criteria namely, house price in relation to income,
230 rental costs in relation to income, safety (crime), quality of housing and energy efficiency can
231 be termed as critical success criteria. However, the other 16 criteria are critical success factors
232 (Lim and Mohamed, 1999).
233
234 Similarly, Gan et al. (2017) aimed at identifying key sustainability performance indicators
235 (KSPIs) from three stakeholder groups such as developers, government and academics. Using
236 the fuzzy set theory and variance analysis, 24 KSPIs were conclusively highlighted from 42
237 sustainability indicators of affordable housing. Among the KSPIs, some of the CSC include
238 affordable price / rent, reduced transport cost, cost effectiveness and energy efficiency.
239 However, like in previous study by Mulliner et al. (2013), some of the 24 identified indicators
240 are possibly critical success factors rather than critical success criteria. For instance, ‘providing
241 human resource for economic development’, ‘ensure balance housing market’, ‘availability of

5
242 green public space and adequate living space within small size unit’ are critical success factors
243 (Lim and Mohamed, 1999).
244
245 It can be concluded from the above literature review that studies on CSC for bridging the gap
246 between sustainable housing and affordable housing are limited. Therefore, a comprehensive
247 investigation of CSC for performance assessment of sustainable affordable housing and for
248 bridging the gap between sustainable housing and affordable housing is worthwhile.
249
250 3. Research Methodology
251
252 3.1 Establishing Potential CSC for Sustainable Affordable Housing
253 To establish the relevance of the various CSC for sustainable affordable housing, a thorough
254 review of the literature on CSC was first conducted. Consequently, a set of 20 CSC that are
255 apposite for sustainable affordable housing was developed. Then a pilot study was conducted
256 by sending out the list of CSC to affordable housing experts with sufficient research and / or
257 industrial experience. This was carried out to review the completeness and clarity of the CSC.
258 Both experts from academia and industry confirmed the comprehensiveness of the CSC with
259 minor corrections on the appropriateness of the words to convey meaning without ambiguity.
260 Moreover, the criterion - waiting time of applicants before being allocated a housing unit – was
261 suggested by one of the experts and upon further consultation with other experts and reading
262 of the literature, it was added. Therefore, a total of 21sustainable affordable housing CSC were
263 established. These are presented in Table 1 with their respective references.
264
265
266
267
268
269
270
271
272
273
274
275
276
277

6
278 Table 1: Potential CSC for Sustainable Affordable Housing
No. CSC for Sustainable Affordable Housing References
CSC01 Timely completion of project Chan and Chan (2004); Bassioni et al. (2004); Ahadzie et al. (2008)
CSC02 Construction cost performance of housing facility Al-Tmeemy et al. (2011); Osei-Kyei and Chan (2017)
CSC03 Quality performance of project Atkinson (1999); Lim and Mohamed (1999); Cox et al. (2003)
CSC04 Safety performance Wai et al. (2012); Kylili et al. (2016); Ngacho and Das (2014)
CSC05 End user's satisfaction with the housing facility Torbica and Stroh (2001); Bryde and Robinson (2005)
CSC06 Project team satisfaction with the housing facility Yan et al. (2018)
CSC07 Environmental performance of housing facility (Eco-friendly) Lim and Mohamed (1999); Atkinson (1999); Rankin et al. (2008)
CSC08 Reduce life cycle cost of housing facility Wai et al. (2012); Ahadzie et al. (2008)
CSC09 Maintainability of housing facility Wai et al. (2012)
CSC10 Energy efficiency of housing facility Wai et al. (2012); Ahadzie et al. (2008)
CSC11 Reduced occurrence of disputes and litigation Osei-Kyei and Chan (2017)
CSC12 Reduced public sector expenditure on managing housing facility Osei-Kyei and Chan (2017)
CSC13 Functionality of housing facility Chan and Chan (2004); Chan et al. (2002)
CSC14 Technical specification of housing Chan and Chan (2004); Osei-Kyei and Chan (2017);
CSC15 Aesthetically pleasing view of completed house Chan and Chan (2004)
CSC16 House price in relation to income Mulliner et al. (2013); Ahadzie et al. (2008)
CSC17 Rental cost in relation to income Mulliner et al. (2013)
CSC18 Commuting cost from the location of housing to public facilities Hamidi et al. (2016)
CSC19 Technology transfer Ahadzie et al. (2008)
CSC20 Waiting time of applicants before being allocated a housing unit Chiu (2007)
CSC21 Take up rate of housing facility (marketability of housing facility) Pullen et al. (2010)
279
280
281
282

7
283 3.2 Data Collection
284 A questionnaire survey was conducted for data collection from affordable housing experts
285 around the world. Questionnaire survey has been used to seek professionals’ views in
286 construction related research (Chan et al., 2018). These experts were selected based on two
287 major criteria as used in previous studies (Ke et al., 2011; Osei-Kyei and Chan, 2017).
288
289 1. Respondents who had broad research and / or industrial experience in affordable housing
290 were selected
291 2. Respondents who have in-depth knowledge on affordable housing projects were contacted
292 to participate in the survey.
293 Considering the selection criteria for experts, it is believed that these experts will offer insight
294 on the relevance of the CSC for sustainable affordable housing projects.
295
296 The targeted respondents for this survey included experts in academia, contractors or
297 developers and consultants. Experts were sourced and identified from affordable housing
298 related publications in top-tier academic refereed journals and databases (member directories)
299 of affordable housing experts. Like snowballing, potential respondents of the questionnaire
300 were implored to forward the questionnaire to any affordable housing expert they deemed
301 suitable to answer the questionnaire. Therefore, it will be a herculean task to state the exact
302 number of questionnaires administered. However, approximately 200 questionnaires were
303 administered. Emails were sent to the participants with the questionnaire attached together with
304 a web-link option for responding to the questionnaire through a “survey monkey”. These
305 flexibility options provided convenient means for experts to respond to the questionnaire to
306 enhance the response rate. Experts were asked to rate on a five-point Likert scale (1= not
307 important, 2= less important, 3= neutral, 4= important, 5= very important) the level of
308 importance of each CSC in measuring success in sustainable affordable housing projects. Fifty-
309 three responses were received. However, two respondents skipped most of the questions on the
310 CSC and were therefore excluded from the number of responses, lowering the number of
311 responses to 51 with a response rate of 26%. Despite the low response rate, the sample size is
312 deemed appropriate for further analysis when compared with the response rate of previous
313 study (Zhang, 2004). Besides, low response rate is not unusual with online questionnaire
314 surveys. For instance, Osei-Kyei and Chan (2017) received 42 responses out of 310 participants
315 (a response rate of 18%). As argued in Chan et al. (2018), a minimum sample size of 30 is
316 regarded as representative of the population. Moreover, despite the small sample size, the aim
317 of the study could be achieved. Table 2 shows the number of responses received from various
318 countries. It shows that most of the responses are from the United States of America, Australia,
319 Malaysia and Italy
320
321
322
323
324
325
326
327
328
329
330
331
332
8
333 Table 2. Responses from Various Countries
Countries Number of Response
USA 12
Australia 5
Malaysia 5
Italy 4
Hong Kong 3
Sweden 3
China 3
Canada 3
Ghana 2
New Zealand 2
Singapore 2
Brazil 1
India 1
Spain 1
South Africa 1
Japan 1
Norway 1
Papua New Guinea 1
Total 51
334
335 3.3 Profiles of Respondents
336 Table 3 is a summary of the profile of respondents. Most of the respondents (72.5%) are in the
337 category of academia / research institute followed by respondents in the consulting firms
338 (9.8%). About 5.9% and 3.9% of the respondents are public sector agencies and private
339 developers / contractors, respectively. With regard to profession, most of the respondents are
340 researchers (54.9%) as shown in Table 3. Many of the respondents (41.2%) had over 20 years
341 of experience in affordable housing projects. Generally, all the respondents indicated that they
342 have been involved in affordable housing research and /or have industrial experience with
343 affordable housing projects.
344 A summary of the research framework for the study is shown in Fig. 1. It reveals the stages,
345 research process, research methodology and findings for the study.
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
9
365
366 Table 3: Respondents’ Category, Profession, Years of Experience and Housing Type Handled
Category, Profession, years of experience
and housing type handled Number of Response Percent
Category
Academia / research institute 37 72.5
Consulting firm 5 9.8
Public sector agency / department 3 5.9
Private developer /contractor 2 3.9
Others 4 7.8
Profession
Academic / researcher 28 54.9
Architect 9 17.6
Quantity Surveyor 3 5.9
Project / Construction manager 2 3.9
Engineer 1 2.0
Others 8 15.7
Years of Experience
1-5 years 9 17.6
6-10 years 11 21.6
11-15 years 6 11.8
16-20 years 4 7.8
> 20 years 21 41.2
Housing Type Handled
Social housing 37 40.2
Public housing 35 38.0
Cooperative housing 14 15.2
Others 6 6.5
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
10
 392

Elaborated in Sections Elaborated in Elaborated in Elaborated in

1 & 2. Sections 3 - 3.3 Sections 4.1 - 4.5 Sections 4.1 - 5.

Data analysis
Research Problem Data Collection & Ranking,
(Ranking analysis,
Identification and Preliminary scrutiny Categorization &
Factor Analysis &
literature review of data correlation of CSC
Pearson correlation)
Stages

[Link] and
[Link] Ranking
comparison of CSC
[Link] review of [Link] design [Link] test
between developed &
the literature and & Piloting [Link] test
developing countries.
identification of [Link] collection and [Link] method
[Link] of
possible CSC scrutiny for suitability [Link] method
CSC groupings
Research
Process

[Link] consistent
Descriptive (Cronbach’s Alpha)
statistics – [Link] sampling 1. Identification of
[Link] review
indicating profile of adequacy test CSC
of research techniques
respondents 3. Bartlett’s test of 2. Development of
of previous study.
(Respondents’ sphericity six components of
2. Lessons were
category of 4. Principal CSC
drawn from previous
profession; Years of component analysis 3. Pearson
studies on forms of
experience and the 5. Varimax with correlation of the
appropriate methods
type of affordable Kaiser CSC to identify
and analysis to be
housing project normalization how CSC correlate
Research adopted for this study
handled) and mean 6. Pearson- with one another
methodo- ranking of CSC
logy correlation

Fig 1: Research Framework for the Study

393
394 4. Data Analysis and Results
395 The Statistical Package for Social Science (SPSS) version 20 was used to conduct statistical
396 analysis of data. Three statistical analyses were conducted: descriptive analysis, factor analysis
397 and Pearson Correlation, as shown in Fig. 1. Before conducting the statistical analysis, the
398 internal consistency reliability and how well the set of 21 CSC are correlated to one another
399 were checked using the Cronbach’s Alpha (α). A Cronbach’s Alpha of 0.720 was obtained.
400 According to Field (2013), a Cronbach’s alpha of 0.70 is considered acceptable. Therefore, the
401 Cronbach’s alpha gives indication that the 21 CSC are internally consistent and well correlated
402 to one another. Then, the descriptive analysis was used for ranking the various criteria based
403 on their computed means and standard deviation values. Moreover, the Pearson correlation
404 analysis was employed to determine how each CSC correlated with one another. Finally, factor
405 analysis was conducted using all the 21 CSC to reduce them into components or categories of
11
406 CSC that could be measuring the same underlying effect (Ahadzie et al., 2008). The steps in
407 the analysis, findings and discussions are presented in subsequent sections.
408
409 4.1 Mean Ranking of CSC
410 The CSC were ranked based on their mean and standard deviation values (shown in Table 4).
411 The ranking is first based on the mean values of the CSC. However, if two or more CSC have
412 the same mean the CSC with the lowest standard deviation is ranked the highest. The top five
413 CSC for responses from all countries include house price in relation to income (CSC16), rental
414 cost in relation to income (CSC17), maintainability of housing facility (CSC9), end user’s
415 satisfaction with housing facility (CSC5) and functionality of housing facility (CSC 13) with
416 mean scores of 4.833, 4.771, 4.553, 4.417 and 4.333, respectively.
417
418 The high ranking of price and rental cost of housing implies that though the other criteria are
419 necessary for sustainable affordable housing, priority is most centered on price and rental
420 affordability. Similarly, in Gan et al. (2017) price and rental affordability were highly ranked
421 by different stakeholders namely government agencies, developers and academics. Therefore,
422 improvement in any of the CSC, that is likely to increase price and rental affordability of
423 housing could be resisted. This was confirmed by a study conducted by Chan et al. (2018) in
424 which increase in cost was among the main reasons for the low adoption of green building
425 technologies in both developed and developing countries. Therefore, Chan et al. (2018)
426 concluded that cheaper and efficient green building technologies should be adopted to improve
427 the level of success of the other criteria in housing projects (i.e. reduce life cycle cost of housing
428 facility and energy efficiency of housing facility) without increasing price and rental cost of
429 housing. The five least ranked CSC from all responses include: reduced public sector
430 expenditure on house management (CSC12), reduced occurrence of disputes and litigation
431 (CSC11), project team satisfaction (CSC6), technical specification of housing (CSC14) and
432 technology transfer (CSC 19) which all had mean values below 3.700. Similarly, in Ahadzie et
433 al. (2008) technology transfer was the least ranked critical success criterion.
434
435 Furthermore, the means, standard deviation and ranking were calculated separately for both
436 developed and developing countries. Classification of countries into developed and developing
437 countries was done with reference to data from Mandelli et al. (2016). China, Malaysia, Ghana,
438 Papua New Guinea, South Africa, India and Brazil were grouped as developing countries while
439 USA, Australia, Italy, Hong Kong, Sweden, Canada, New Zealand, Singapore, Spain, Japan
440 and Norway were classified as developed countries. Among the developed countries, priority
441 was given first to rental cost of housing and then house price. However, in developing
442 countries, price of housing was ranked first while rental cost was ranked forth. In Gilbert
443 (2016), it was stated that the privatization of public housing due to abysmal low rents, self-help
444 housing and the cultural preference for ownership among developing countries could be the
445 reasons for the preference of price affordability over rental affordability. From the findings (as
446 shown in Table 4), other CSC such as commuting cost from location of housing facility to
447 public facilities, maintainability of housing facility and reduce lifecycle cost were ranked
448 relatively high among developed countries as compared to their rankings from developing
449 countries. It is not surprising given the disparities in the ranking of these sustainability related
450 criteria. This reflects the high priority devoted to these criteria from developed countries as
451 compared to developing countries (Darko et al., 2018).
452
453
454
455
12
456 Table 4: Ranking of CSC
Code All Countries Developed Countries Developing Countries
Mean SD Rank Mean SD Rank Mean SD Rank
CSC16 4.833 .429 1 4.857 .430 2 4.769 .439 1
CSC17 4.771 .425 2 4.857 .355 1 4.539 .519 4
CSC09 4.553 .503 3 4.559 .504 3 4.539 .519 4
CSC05 4.417 .613 4 4.343 .639 4 4.615 .506 3
CSC13 4.333 .724 5 4.286 .750 6 4.462 .660 8
CSC03 4.313 .689 6 4.171 .707 10 4.692 .480 2
CSC02 4.313 .748 7 4.286 .789 5 4.385 .650 9
CSC04 4.292 .544 8 4.200 .531 9 4.539 .519 6
CSC08 4.250 .700 9 4.229 .690 7 4.308 .751 14
CSC18 4.250 .758 10 4.200 .797 8 4.385 .650 9
CSC10 4.167 .694 11 4.086 .612 12 4.385 .870 12
CSC20 4.167 .883 12 4.086 .951 11 4.385 .650 9
CSC01 4.042 .898 13 3.886 .932 13 4.461 .660 7
CSC21 4.000 .905 14 3.882 .946 14 4.364 .674 13
CSC07 3.854 .684 15 3.800 .719 15 4.000 .577 18
CSC15 3.833 .753 16 3.743 .741 16 4.077 .760 17
CSC12 3.688 1.095 17 3.543 1.146 17 4.078 .862 16
CSC11 3.583 .964 18 3.429 .948 18 4.000 .913 20
CSC06 3.575 .853 19 3.412 .857 19 4.000 .707 19
CSC14 3.521 .875 20 3.286 .789 20 4.154 .801 15
CSC19 3.065 1.020 21 2.971 .937 21 3.333 1.231 21
457
458 4.2 Factor Analysis
459 Factor analysis was conducted to group the 21 CSC into components. This was necessary to
460 identify the underlying structures of CSC for sustainable affordable housing projects. The
461 Principal Component Analysis (PCA) was adopted for the factor analysis. Prior to conducting
462 the analysis, the suitability of the data for factor analysis was assessed. The Kaiser-Meyer-
463 Olkin (KMO) Sampling Adequacy Test and Bartlett’s Test of Sphericity were carried out to
464 determine the data appropriateness. KMO measures the sampling adequacy as a ratio of the
465 squared correlation between the variables to the squared partial correlation between the
466 variables (Field, 2013). KMO value of 0 is an indication of the unsuitability of data for factor
467 analysis while a value of 1 indicates that the data are suitable and will yield reliable and distinct
468 factors in the factor analysis. A KMO value above 0.5 is deemed appropriate (Field, 2013).
469 Table 5 shows the test results. The KMO measure of sampling adequacy was 0.63. Thus, this
470 was considered acceptable. Besides, the Bartlett Test of Sphericity was conducted to check if
471 the original correlation matrix is an identity matrix. For data suitability for factor analysis, the
472 Bartlett’s test of sphericity must be large with a small associated significance level (Pallant,
473 2013). The Bartlett’s test of Sphericity was 483.120 at a significance level of 0.000. This
474 indicates that the population correlation matrix was not an identity matrix (Larose, 2006; Field,
475 2013). Therefore, the test results of the KMO and Bartlett’s Test suggested that the data were
476 suitable for factor analysis.
477
478 With the selection of the Varimax Rotation, the Principal Component Analysis was then carried
479 out to identify the fundamental structures of CSC. Conventionally, only variables with
480 eigenvalue and factor loading at cut-off points of 1.0 and 0.50, respectively, were retained.
481 Since the factor loadings for all the CSC exceeded 0.50 (Shown in Table 7), all the 21 CSC
482 were retained. “The relatively high values of the loading factors (0.6 for more than four
13
483 variables) lend support to the favorability of the sample size for the analysis” (Ahadzie et al.
484 2008 p. 681). Six components were extracted (as shown in Table 7). The total variance
485 explained by each component (as shown in Table 7) is as follows: Component 1 (29.377%);
486 component 2 (13.103%); component 3 (10.317%); component 4 (7.868%); component 5
487 (6.790%) and component 6 (5.271%). In sum, the components explained 72.726% of the total
488 variance.
489
490 Depending on the underlying variables in each component, the components were named as
491 follows: component 1 was named ‘Household satisfaction CSC’; component 2: Stakeholders’
492 satisfaction CSC; component 3: House operation cost CSC; component 4: Time measurement
493 CSC; component 5: Location affordability cost CSC; component 6: Quality-related CSC.
494
495 Table 5: KMO and Bartlett’s Test
Kaiser-Meyer-Olkin Measure of Sampling
Adequacy 0.630
Bartlett's test of sphericity Approximate chi-square 483.120
df. 210
Sig. 0.000
496
497
498

14
Table 6: Correlation Matrix of CSC

CODE CSC01 CSC02 CSC03 CSC04 CSC05 CSC06 CSC07 CSC08 CSC09 CSC10 CSC11 CSC12 CSC13 CSC14 CSC15 CSC16 CSC17 CSC18 CSC19 CSC20 CSC21
CSC01 r 1.000
CSC02 r .392** 1.000
CSC03 r .116 .343* 1.000
CSC04 r .105 .189 .433** 1.000
CSC05 r .161 .174 .441** .521** 1.000
CSC06 r .361* .047 .124 .230 .393** 1.000
CSC07 r -.094 -.200 .370** .231 .402** .365* 1.000
CSC08 r .051 .173 .232 .140 .397** .393** .389** 1.000
CSC09 r -.005 .269 .293* .415** .487** .206 .415** .559** 1.000
CSC10 r .193 .020 .200 .263 .233 .442** .590** .569** .384** 1.000
CSC11 r .414** .155 .072 .358* .192 .365* .132 .284 .156 .297* 1.000
CSC12 r .252 .252 .076 .192 .040 .082 -.062 .271 .208 .098 .479** 1.000
CSC13 r .240 .314* .469** .450** .591** .084 .186 .210 .459** .226 .264 .080 1.000
CSC14 r .540** .429** .430** .389** .301* .597** .272 .339* .352* .449* .364* .204 .324* 1.000
CSC15 r .136 .019 .389** .433** .200 .455** .447** .363* .422** .420** .313* .116 .104 .490** 1.000
CSC16 r -.202 .033 -.108 .030 -.135 -.200 -.157 .142 -.057 .095 .034 .339* -.160 -.274 -.022 1.000
CSC17 r -.253 -.239 -.041 .111 .048 -.219 -.044 .125 .009 .132 -.082 -.112 .115 -.302* -.122 .369** 1.000
CSC18 r .172 .122 .255 .438** 366* -.030 .113 .281 .303* .243 .349* .507** .426** .056 .186 .261 .248 1.000
CSC19 r .263 .291 .254 .206 -.008 .323* .274 .327* .232 .305* .495** .437** .033 .518** .483** .026 -.227 .150 1.000
CSC20 r .045 -.145 .192 .339* .498** .354* .358* .344* .367* .231 .333* .231 .111 .106 .426** .075 .161 .350* .160 1.000
CSC21 r -.045 -.102 .226 .458** .404** .150 .331* .214 .101 .249 .312* .115 .193 .116 .273* .341* .325* .398** .218 .640** 1.000

r =Value for Pearson correlation.

p= Value of the significance
*Correlation is significant at the 0.05 level (2-tailed)
**Correlation is significant at the 0.01 level (2-tailed)

(CSC01 = Timely completion of projects; CSC02 = Construction cost performance of housing facility; CSC03 = Quality performance of project; CSC04 = Safety performance; CSC05
= End user’s satisfaction with the housing facility; CSC06 = Project team satisfaction with the housing facility; CSC07 = Environmental performance of housing facility (Eco- friendly);
CSC08 = Reduced life cycle cost of housing facility; CSC09 = Maintainability of housing facility; CSC10 = Energy efficiency of housing facility; CSC11 = Reduced occurrence of
disputes and litigation; CSC12 = Reduced public sector expenditure on managing housing facility; CSC13 = Functionality of housing facility; CSC14 = Technical specification of
housing; CSC15 = Aesthetically pleasing view of completed house; CSC16 = House price in relation to income; CSC17 = Rental cost in relation to income; CSC18 = Commuting cost
from the location of housing to public facilities; CSC19 = Technology transfer; CSC20 = Waiting time of applicants before being allocated a housing unit; CSC21 = Take up rate of
housing facility (marketability of housing facility))

15
Table 7: Rotated Component Matrix
Code CSC for Sustainable Affordable Housing Components Loading
1 2 3 4 5 6
Component 1: Household Satisfaction CSC
CSC13 Functionality of housing facility 0.839 ⎻ ⎻ ⎻ ⎻ ⎻
CSC5 End user's satisfaction with the housing facility 0.812 ⎻ ⎻ ⎻ ⎻ ⎻
CSC9 Maintainability of housing facility 0.641 ⎻ ⎻ ⎻ ⎻ ⎻
CSC4 Safety performance (crime) 0.610 ⎻ ⎻ ⎻ ⎻ ⎻
Component 2: Stakeholders’ Satisfaction CSC
CSC1 Timely completion of project ⎻ 0.788 ⎻ ⎻ ⎻ ⎻
CSC6 Project team satisfaction ⎻ 0.688 ⎻ ⎻ ⎻ ⎻
CSC11 Reduced occurrence of disputes and litigation ⎻ 0.607 ⎻ ⎻ ⎻ ⎻
Component 3: Housing Operation Cost CSC
CSC10 Energy efficiency of housing facility ⎻ ⎻ 0.856 ⎻ ⎻ ⎻
CSC8 Reduced lifecycle cost of housing ⎻ ⎻ 0.842 ⎻ ⎻ ⎻
CSC7 Environmental performance of housing facility (Eco-friendly) ⎻ ⎻ 0.530 ⎻ ⎻ ⎻
Component 4: Time Measurement CSC
CSC21 Take up rate of housing facility (marketability of housing facility) ⎻ ⎻ ⎻ 0.802 ⎻ ⎻
CSC20 Waiting time of applicants before being allocated housing unit ⎻ ⎻ ⎻ 0.716 ⎻ ⎻
CSC2 Construction cost performance of housing facility ⎻ ⎻ ⎻ -0.555 ⎻ ⎻
Component 5: Location affordability Cost CSC
CSC12 Reduced public sector expenditure on house management ⎻ ⎻ ⎻ ⎻ 0.818 ⎻
CSC16 House price in relation to income ⎻ ⎻ ⎻ ⎻ 0.649 ⎻
CSC18 Commuting cost from the location of housing to public facilities ⎻ ⎻ ⎻ ⎻ 0.631 ⎻
CSC17 Rental cost in relation to income ⎻ ⎻ ⎻ ⎻ 0.506 ⎻
Component 6: Quality-Related CSC
CSC3 Quality performance of project ⎻ ⎻ ⎻ ⎻ ⎻ 0.686
CSC15 Aesthetically pleasing view of completed house ⎻ ⎻ ⎻ ⎻ ⎻ 0.665
CSC19 Technology transfer ⎻ ⎻ ⎻ ⎻ ⎻ 0.658
CSC14 Technical specification of housing ⎻ ⎻ ⎻ ⎻ ⎻ 0.600

Eigenvalue 6.169 2.752 2.167 1.652 1.426 1.107

Variance (%) 29.377 13.103 10.317 7.868 6.790 5.271
Cumulative Variance (%) 29.377 42.480 52.797 60.665 67.455 72.726
Extraction Method: Principal Component Analysis; Rotation method: Varimax with Kaiser Normalization

16
 1 4.3 Results of Principal Component Analysis and Discussion
2
3 4.3.1 Component 1: Household Satisfaction CSC
4 The underlying CSC in this component highlight the criteria that lead to household satisfaction
5 in a housing facility. This component is characterized by four main criteria. These four CSC,
6 together with the percentages of their loading in bracket include: functionality of housing
7 facility (83.9%); end user’s satisfaction with housing facility (81.2%); maintainability of
8 housing facility (64.1%) and safety performance (61.0%). This component explains most of
9 the variance among the six components, about 29.377% (please refer to Table 7 for loading
10 and for the variance).
11
12 The correlation matrix (shown in Table 6) revealed significant associations among the various
13 CSC in this component. For example, the correlation between ‘functionality of housing facility’
14 and ‘end user’s satisfaction’ was significant (r=0.591, p=0.01); between ‘functionality of
15 housing facility’ and ‘maintainability of housing facility’ (r= 0.459, p=0.01) and ‘functionality
16 of housing facility’ and ‘safety performance’ (r=0.450, p=.0.01). Therefore, the association
17 among these CSC is coherent since they measure the same factor – household satisfaction.
18
19 Similarly, in Ahadzie et al. (2008), household satisfaction with housing facility emerged as one
20 of the components for mass housing projects. To bridge the gap between sustainable housing
21 and affordable housing, meeting household satisfaction is very important. Household
22 satisfaction is defined as an assessment of the degree to which the current dwelling of the
23 household and quality of the environment are close to the expectations of their favorite one
24 (Galster, 1985). Ensuring functionality of housing according to aspirations, safety performance
25 (i.e. security provision features) and ease of housing facility maintenance are relevant for
26 household satisfaction. Functionality is considered a consequence of the facility. It includes the
27 performance output and the benefits of the facility to the household. Performance output of
28 housing facility measures the quality of the housing while the benefit of the housing
29 functionality is a measure of the household satisfaction (Jusan, 2007). Functionality can be
30 measured by the level of conformance to client’s expectation, with the ultimate goal of
31 achieving fitness for purpose (Chan et al., 2002). Functionality should be assessed at the post
32 construction phase, when the facility is completed and is in use (Chan et al., 2002).
33
34 Moreover, several features of a house ensure residential satisfaction. For instance, separate
35 bedrooms for parents and children contribute to more private space and residential satisfaction
36 (Ren and Folmer, 2017). Similarly, Pearson’s correlation conducted by Mohit et al. (2010)
37 revealed that residential satisfaction is highly and positively correlated with dwelling unit
38 features followed by the social environment, dwelling support services and public facilities.
39 Among planning policies, neighborhood interaction and safety were dominant predictors of
40 residential satisfaction. Moreover, maintainability of a housing facility ensures household
41 satisfaction. In Torbica and Stroth (2001), low-cost maintenance features of house and ease of
42 home maintenance were identified as contributory variables for household satisfaction.
43
44 Although Riazi and Emami (2018) found that design principles on residential satisfaction had
45 a significant value of 0.183, most of the design features were related to safety and security
46 provisions. Some of these features include lighting of public areas, safety of car parking, safety
47 of outdoor parking, safety of indoor space and security for children in public areas. Personal
48 security was identified as a feature that first-time homebuyers look out for in making
49 purchasing decision. Crime rate in the neighborhood and whether a neighborhood is gated are
50 significant factors that influence residential satisfaction and the likelihood of home ownership
17
 51 among first-time homebuyers (Teck-Hong, 2012). Safety community together with good
52 leisure facilities promote residential satisfaction (Ren and Folmer, 2017).
53
54 4.3.2 Component 2: Stakeholders’ Satisfaction CSC
55 This component consists of ‘timely completion of project’ (78.8%), ‘project team satisfaction’
56 (68.8%) and ‘reduced occurrence of dispute and litigation’ (60.7%). These three CSC
57 explained about 13.10% of the total variance (as shown in Table 7).
58
59 The construction of an affordable housing project involves many stakeholders including the
60 targeted households, governments, developers, design team, suppliers and the people in the
61 neighborhoods of the project. Stakeholders receive and execute the success criteria. Therefore,
62 they have the potential to impact the outcome of sustainable affordable housing project (Yan
63 et al., 2019). Findings of the study showed that there is a statistically significant correlation
64 between ‘timely completion of project’ and ‘reduce occurrence of disputes’ (r=0.414, p=0.01)
65 (as shown in Table 6). According to Sambasivan and Soon (2007), most disputes in
66 construction projects are the effects of project delays. Timely completion of projects prevents
67 construction disputes that could arise from construction claims. Besides, decrease in property
68 values due to affordable housing projects is one of the causes of public protest which has caused
69 the failure of many affordable housing projects (Nguyen et al., 2013; Tighe, 2010). Delays and
70 complete abandonment of projects due to political reasons could affect the values of
71 neighboring housing facilities. Such projects are often used as hideouts for criminals. As such,
72 households in the neighborhood might live in fear of insecurity. Therefore, potential tenants
73 and buyers might perceive such surroundings as unsafe. This could lower the rent and price of
74 the neighboring facilities. This leads to dissatisfied neighborhoods who may disrupt and protest
75 against the construction of subsequent affordable housing project. Accordingly, timely
76 completion of affordable housing projects ensures stakeholders’ satisfaction by preventing
77 negative social impacts. It also ensures project team satisfaction (Rashvand and Zaimi Abd
78 Majib, 2013). This is evident in the statistically significant correlation (as shown in Table 6)
79 between ‘timely completion of project’ and ‘project team satisfaction’ (r=0.361, p=0.05).
80 Similarly, ‘reduced occurrence of disputes’ and ‘project team satisfaction’ have a statistically
81 significant association (r=0.365, p=0.05).
82
83 4.3.3 Component 3: Housing Operation Cost CSC
84 The total variance accounted by component 3 is 10.3% (as shown in Table 7). The respective
85 criteria and the percentage of the factor loadings in this component include energy efficiency
86 (85.6%), reduced lifecycle cost of housing facility (84.2%) and environmental performance of
87 housing facility (53.0%) (as shown in Table 7). The criteria showed significant correlation
88 among themselves. The correlation (as shown in Table 6) between energy efficiency and
89 reduced lifecycle cost of housing was significant (r=0.569, p=0.01); the correlation between
90 energy efficiency and environmental performance was also significant (r=0.590, p=0.01).
91 Similarly, reduced life cycle cost and environmental performance of housing facility revealed
92 a significant correlation (r=0.389, p=0.01). These significant associations among these criteria
93 are not surprising because according to Ruparathna et al. (2016), the environmental impact of
94 a housing facility is determined from its lifecycle and its energy consumption. Since all these
95 criteria measure the operation cost or impact of a housing facility (Pacheco et al., 2012), this
96 component was, accordingly, named as housing operation cost CSC.
97
98 For sustainable affordable housing, the operations cost of housing is worth considering due to
99 its cost saving benefits to low income household and the environment. Minimizing the
100 operation cost of affordable housing projects could be achieved through energy efficient
18
101 housing. The fundamental principle of energy efficient housing is to use the minimum energy
102 for operation (such as cooling, lighting, heating etc.) without impacting residents’ health and
103 comfort (Ruparathna et al., 2016). Improving energy efficiency of affordable housing is key to
104 abating the environmental effects – greenhouse effects – due to CO2 emissions. It also reduces
105 the energy use and therefore provides economic benefits such as savings to low-income
106 earners. Moreover, energy efficient affordable housing is a requirement to prevent fuel poverty
107 – low income household spending beyond 10% of their income on domestic energy (Mattioli
108 et al., 2018).
109
110 Studies have been conducted on energy efficient technologies that can be adopted to provide
111 sustainable affordable housing without rendering household shelter poor (Allouhi et al., 2015;
112 Morrissey et al. 2011; Nikolaidis et al., 2009). On the mechanical components of a housing
113 facility, heating, ventilation and air conditioning (HVAC) system is the most energy
114 consumption component of a housing facility (Perez-Lombard et al., 2011). Using thermal
115 solar systems for a substitute of an electric water heater leads to 80% saving of the cost of
116 heating water as well as ensuring environmental protection (Nikolaidis et al., 2009). By
117 changing from air-cooled to water cooled air-conditioning system, substantial electricity
118 consumption could be reduced (Yik et al. 2001).
119
120 With regard to lighting system, about 15% of the total energy of a building is spent on lighting.
121 However, installing better luminous efficacy lamps and linking daylight to lighting systems
122 could reduce electricity consumption on lighting. Moreover, changing to light emitting diode
123 (LED) light system, replacing incandescent lamps with low energy fluorescent lamps and
124 installing automated lighting system can reduce the amount of electricity demanded for lighting
125 (Ruparathna et al., 2016). Another important area for energy efficient housing is the building
126 envelope. Improved Insulation minimizes the heat gain or loss from a building thereby
127 enhancing the thermal performance of the housing facility (Ruparathna et al., 2016). Reflective
128 paint and coating on roofs and walls or insulating paint with low conduction can be used to
129 improve the thermal performance of a building. In a location of high temperature difference
130 between day and night, coating of the external surface of the housing facility provides better
131 thermal function. However, in locations of low temperature difference between daytime and
132 nighttime, housing facilities with interior insulation do better (Huang et al., 2013).
133
134 Building codes set the lowest requirement for energy efficiency in buildings. Notable ones
135 include BREEAM, Leadership in Energy and Environment (LEED) and Green Star. These
136 codes may target one of the following building energy concepts: low energy building, passive
137 houses, zero energy building, zero carbon building (Allouhi et al., 2015). By making building
138 energy code mandatory, it was stated that the yearly electricity consumption, for example in
139 Hong Kong, can be lowered by 7.9% (Lee and Yik, 2002). Therefore, through the development
140 of localized codes or adoption of internationally recognized codes, affordable housing would
141 be energy efficient and thus sustainable.
142
143 The shape of a housing facility affects the amount of solar radiation that the building receives,
144 which consequently influences its total energy consumption (Mingfang, 2002). The higher the
145 solar radiation received by a housing unit, the higher the energy required to cool it (Elasfouri
146 et al., 1991). According to Aksoy and Inalli (2006), 36% of heat energy savings can be obtained
147 by combining the optimization of orientation and shape of a building. For instance, on
148 quantifying the effects of a building shape on the amount of energy required to heat and cool a
149 building, Florides et al. (2002) concluded that the best orientation to maximize the solar
150 benefits of a rectangular building is for the lengthiest wall of the housing unit to face the south.
19
151 The southern orientation is best for heat gain during wintertime and for regulating solar
152 radiation during summer (Pacheco et al., 2012). Shading on buildings also affects the amount
153 of solar radiation gain by a building. For instance, overhangs over windows prevent the direct
154 entry of solar radiation through the window, therefore, it regulates the entry of excessive heat
155 and daylight. However, since overhangs are mostly designed to remain fixed, they could favor
156 energy savings in certain times while hindering energy saving at a different time. Thus, mobile
157 shading devices provide better energy saving benefits than immovable shading devices
158 (Bouchlaghem, 2000). Using the net present value appraisal on a uniform evaluation period,
159 Nikolaidis et al. (2009) found that insulation of the roof of a building provides better
160 intervention concerning heat insulation than with the replacement of windows and doorframes,
161 which yielded low returns on investment.
162
163 Moreover, though mud / baked bricks cannot be used to construct structural elements, its use
164 for the construction of non-loading bearing walls could offer energy saving benefits. According
165 to Chel and Tiwari (2009), internal temperatures of mud houses are moderate throughout the
166 year. This leads to potential energy savings. Mud houses have yearly heating and cooling
167 energy saving of about 1481KWh/year and 1813kWh, respectively. Moreover, mud-houses
168 can alleviate 5.2 metric tons per year of CO2 emission into the atmosphere.
169
170 4.3.4 Component 4: Time Measurement CSC
171 The extracted CSC with their factor loading for this component include ‘take up rate of housing
172 facility (marketability)’ (80.2%), ‘waiting time of applicants before being allocated housing
173 unit’ (71.6%) and ‘construction cost performance of housing facility’ (-55.5%). This cluster
174 explained about 7.87% of the total variance (as shown in Table 7) and was named time
175 measurement CSC.
176
177 The correlation matrix (shown in Table 6) revealed that significant correlations exist among
178 the criteria in this component. For example, the correlation between ‘take up rate of housing
179 facility’ and ‘waiting time of applicant before being allocated a housing unit’ was significant
180 (r=0.640, p= 0.01). Since both criteria measure the time taken for a household to move into a
181 housing facility, the significant correlation between them is logical.
182
183 Aside building affordable housing, it is important to measure how supplied housing is reducing
184 the time spent by low-income earners in the ‘waiting line’ before being allocated affordable
185 housing unit. Besides, assessing how affordable housing supplies are meeting the needs of
186 household is very critical. This can be measured using the take-up rate of housing facilities.
187 Houses that are affordable but not adequate and sustainable are likely to receive low take-up
188 rate by low income earners (Teck-Hong, 2012). Take up rate of an affordable housing facility
189 is significantly associated with household’s satisfaction (r=0.404, p=0.01). The correlation
190 between take up rate and household’s satisfaction indicates that high expectation for household
191 satisfaction leads to high take-up rate of a housing facility. However, high cost of housing
192 facility beyond the affordability range of the household could lead to low take up rate of the
193 housing facility and increase waiting time of applicants for housing unit allocation.
194
195 4.3.5 Component 5: Location Affordability Cost CSC
196 The principal component 5 contains four CSC: reduced public sector expenditure on housing
197 management (81.8%); house price in relation to income (64.9%); commuting cost from the
198 location of housing to public facilities (63.1%) and rental cost in relation to income of
199 household (50.6%) (as shown in Table 7). This component accounted for 6.79% of the total
200 variance (as shown in Table 7). Studies have stated that affordability should be measured as
20
201 location affordability, that is taking into consideration housing affordability cost and cost of
202 transportation or accessibility (Kramer, 2018; Mattioli et al., 2018; Fan and Huang, 2011).
203 Therefore, this component was labelled location affordability cost CSC.
204
205 As shown in the correlation matrix in Table 6, a statistically significant correlation exists
206 between ‘house price in relation to income’ and ‘rental cost in relation to income’ (r=0.369,
207 p=0.01). This association between the two criteria is reasonable since both are used to measure
208 the same item – housing affordability. Similarly, there was a significant correlation between
209 the criteria ‘reduced public sector expenditure on house management’ and ‘commuting cost
210 from the location of housing to public facilities’ (r=0.507, p=0.01).
211
212 Previous studies have elaborated on the importance of housing affordability (Adinyira et al.,
213 2014; Ahadzie et al., 2008). However, an important cost factor which was overlooked in
214 measuring affordability is the cost of transportation. Location affordability incorporates both
215 the cost of housing and transportation. A study conducted by Saberi et al. (2017), revealed that
216 neighbourhoods that seem to be affordable with regard to only housing cost are not definitely
217 affordable when transportation cost is factored in. Housing facilities at the urban peripheral or
218 in low-residential density areas may appear more affordable yet might suffer from inadequate
219 access to various amenities and incur high cost on transportation in order to access the
220 amenities. Thus, the low housing cost is mostly offset by the high commuting cost which leads
221 to transport poverty. A household might be transport poor based on three conditions: if the
222 household spends more than 10% of their income on car running costs, if the household lives
223 more than one mile from the closest bus or station and if it takes more than one hour to access
224 a number of important services by cycling, walking and public transport (Sustrans, 2012 cited
225 in Mattioli et al., 2018). Transportation poverty has many effects. Individuals can be rendered
226 unemployed due to inability to afford ownership of cars / commuting cost. Besides, most
227 households that are able to afford do trade-off transport expenditure against spending on other
228 necessities (Mattioli et al., 2018).
229
230 It is recommended that policies and plans for housing affordability should take into account
231 transportation infrastructure supply (Saberi et al., 2017). Three main factors influence
232 transportation affordability namely the built environment, policy environment and the socio-
233 demographics of households (Fan and Huang, 2011). The socio-demographics of the household
234 defines the influence of household income on transportation affordability. The built
235 environment (defined by the land use) and urban design influence the transportation
236 affordability. There is an association between the built environment and travel behavior. For
237 instance, low residential density and mono-functional use of land are related to more car travel.
238 However, high density areas such urban centers where the buildings are closer, walking and
239 cycling would be encouraged among many households especially low-income earners (Mattioli
240 et al., 2018). It is worth noting that extreme cases of compact city and urbanization could
241 increase traffics on the roads thereby increasing the time spent on travelling. Thus, an
242 affordable house is not sustainable if the cost and time of transportation are very high.
243
244 4.3.6 Component 6: Quality-Related CSC
245 Lastly, the sixth principal component contains four CSC. These CSC together with their factor
246 loading are ‘quality performance of project’ (68.6%), ‘aesthetically pleasing view of completed
247 house’ (66.5%), ‘technology transfer’ (65.8%) and ‘technical specification of housing’
248 (60.0%). This component explains 5.3% of the total variance and is named quality-related CSC.
249

21
250 The findings revealed that some of the four CSC in this component showed statistically
251 significant correlation among themselves. For instance, the correlation matrix (as shown in
252 Table 6) revealed a significant relationship between quality performance of project and
253 aesthetically pleasing view of completed house (r=0.389, p=0.01). Besides, the correlation
254 between quality performance of project and technical specification of housing was significant
255 (r=0.430, p=0.01). Moreover, the correlation matrix revealed a significant association between
256 ‘technology transfer’ and ‘technical specification of housing’ (r=0.518, p=0.01). Likewise, the
257 association between technology transfer and aesthetically pleasing view of completed house is
258 significant (r=0.483, p=0.01).
259
260 The significant association between quality performance and aesthetically pleasing view of
261 completed house could be attributed to the fact that the conventional description of quality is
262 based on issues such as ‘how well a housing facility blends into its environment’, ‘the facility’s
263 psychological impacts on its inhabitants’, ‘the ability of landscaping plan to match the theme
264 of nearby structures’ and ‘the use of intriguing novel design models that capture people’s
265 imaginations’ (Stasiowski and Burstein, 1994). Since the aesthetic definition of quality is
266 subjective, there is often no consensus on whether quality affordable housing has been achieved
267 or not (Arditi and Gunaydin, 1997). However, quality performance of housing facility can also
268 be defined objectively as meeting technical specification of the designer, owner and regulatory
269 organizations (Ferguson and Clayton, 1988).
270
271 Due to the subjective and objective assessment of quality, it is important to differentiate
272 ‘quality in perception’ and ‘quality in fact’. A housing facility that meets client’s and
273 household’s expectation attains quality in perception while a housing facility that meets the
274 technical specification attains ‘quality in fact’ (Arditi and Gunaydin, 1997). ‘Quality in fact’
275 can be achieved by meeting two main requirements: product quality and process quality (Arditi
276 and Gunaydin, 1997). Whereas product quality is ensuring suitable construction materials,
277 equipment and technology required for the construction of a housing facility, process quality
278 involves attaining quality with regard to the design and construction of the housing facility.
279 Achieving both forms of quality is very important. The neglect of quality in perception has
280 often resulted in abandon affordable housing facilities (Arditi and Gunaydin, 1997; Teck-
281 Hong, 2012). Therefore, it is suggested that prior to the construction of any housing facility, a
282 pilot study should be conducted to assess the needs of the intended households. Regular
283 assessment of the needs of the intended households is important since household needs are
284 ephemeral (Adabre and Chan, 2018). This assessment will ensure that the expected quality of
285 a household is met. Though quality is considered a latent variable, it could be achieved based
286 on the housing design features. Design principles such as interior layout (i.e. size of living
287 room, arrangement of rooms, size of kitchen, availability of storage room) and privacy of living
288 space (i.e. number of bedrooms, size of bedrooms and number of bathrooms) are considered
289 very important among low-income households (Opoku and Abdul-Muhmin, 2010). Among
290 interior design features such as number of bedrooms, bathrooms and living rooms, living space
291 was the indicator of quality that had the highest loading and reliability (Ren and Folmer, 2017).
292 Thus, these quality features should be taken into consideration for sustainable affordable
293 housing projects so as to meet household needs.
294
295 The significant positive correlations among technology transfer, technical specification and
296 aesthetically pleasing view of housing (as shown in Table 6) are logical. In Adinyira et al.
297 (2014) technology transfer emerged together with cost of individual units. Accordingly, it was
298 stated that the benefits of technology transfer could improve the price affordability of housing
299 facilities. In this study, technology transfer emerged together with quality performance of
22
300 housing project, aesthetically pleasing view of completed house and technical specification of
301 housing. This implies that aesthetically pleasing view and technical specification could be
302 improved through technology transfer.
303
304 4.4 Relevance and Integration of the Six Components of CSC
305 Based on the empirical data, six main components are essential for making affordable housing
306 sustainable. The attainment of the six components of CSC will ensure that the three main
307 aspects of sustainability – economic, social and environment – are achieved in affordable
308 housing projects. For instance, household satisfaction CSC, time measurement CSC, quality-
309 related CSC and stakeholder’s satisfaction CSC measure success with regard to social
310 sustainability in housing projects. These CSC for social sustainability are essential for ensuring
311 good-quality, secure and healthy housing facilities (Dang et al., 2014). The four components
312 of CSC for social sustainability are interdependent. Quality-related CSC could lead to the
313 achievement of both household satisfaction CSC and stakeholders’ satisfaction CSC. Besides,
314 quality-related CSC could influence high residential take up – one of the time measurement
315 CSC. A high residential take up rate indicates that the housing facilities are appreciated by the
316 target household (Chen et al., 2014) whereas a low residential take up of affordable housing
317 facilities may indicate that the housing facilities need to be improved or other facilities need to
318 be provided to meet household needs. Accordingly, the CSC for social sustainability play an
319 important role in improving the operational performance of affordable housing projects (Yuan
320 et al., 2018).
321
322 Economic sustainability in affordable housing projects could be measured using the location
323 affordability cost CSC (Isalou et al., 2014). These criteria are essential for assessing the success
324 of price / rental and commuting cost affordability of housing facilities. Commuting cost of
325 residents plays a tremendous role on household income (Gan et al., 2017). High commuting
326 cost could cause high cost burden on household income thereby worsening housing
327 affordability of households. The CSC for economic sustainability are interdependent on the
328 CSC for social sustainability and environmental sustainability. For instance, low location
329 affordability cost CSC due to high accessible locations of housing facilities reduces
330 transportation cost which could influence household satisfaction CSC, stakeholders’
331 satisfaction CSC and time measurement CSC (Yuan et al., 2018). Besides, low location
332 affordability cost CSC due to accessible location of housing facilities conserves energy and
333 reduces pollution emission (such as carbon dioxide and carbon mono-oxide), which improves
334 on environmental sustainability (Isalou et al.,2014).
335
336 Housing operation cost CSC (such as energy efficient housing facilities, reduced lifecycle cost
337 of housing facility and environmental performance of housing facility) are essential for
338 environmental sustainability attainment in affordable housing projects. The housing operation
339 cost CSC measure the impacts of housing facilities on the environment. The construction,
340 maintenance and use of housing facilities involves the consumptions of an enormous amount
341 of scarce energy and other resources. These could lead to high environmental depletion.
342 However, using energy efficient technologies and environmental friendly materials could
343 improve environmental sustainability of affordable housing projects (Golubchikov and
344 Badyina, 2012).
345
346 Based on the aforementioned components of CSC, bridging the gap between sustainable
347 housing and affordable housing requires the following six components of CSC: (1) household
348 satisfaction CSC, (2) stakeholders’ satisfaction CSC, (3) housing operation cost CSC (4) time
349 measurement CSC, (5) location affordability cost CSC, (6) quality-related CSC. To summarize,
23
350 a framework of sustainable affordable housing can therefore be established as shown in Fig. 2,
351 which includes the six components of CSC which are essential for sustainable affordable
352 housing. These components are interdependent with important relationships among them.
353
354
355
356
Sustainable Housing CSC 357
358
CSC for Social Sustainability
359
360
Household Satisfaction CSC
361
362
363
Stakeholders’ Satisfaction CSC 364
365
366
Quality-Related CSC
367
368
Time Measurement CSC

Sustainable
Affordable Housing affordable
housing
CSC for Economic Sustainability

Location Affordability Cost CSC

369
370
371
372

CSC for Environmental Sustainability Represents a relationship

between components

Housing Operation Cost CSC

373

Fig. 2: A Framework for Sustainable Affordable Housing

374
375
376
377
378
379
380
381
382
383
384
385
386
24
387 4.5 Application of the CSC for Sustainable Affordable Housing
388 CSC serve as a set of standards by which anything is or can be judged (Lim and Mohammed,
389 1999). The essence of CSC is to develop a set of criteria or standards which serve as guidelines
390 by which policy makers can assess the outcomes of projects (Chan and Chan, 2004). While
391 some of these CSC (such as location affordability cost CSC, housing operation cost CSC and
392 time measure CSC) can be measured objectively through formulae / standard values, the other
393 components of CSC (such as quality-related CSC, household satisfaction CSC and
394 stakeholder’s satisfaction CSC) use personal judgement and subjective opinions of
395 stakeholders of affordable housing projects (Chan and Chan, 2004; Adabre and Chan, 2018).
396 Concerning the objective CSC, a standard for location affordability cost can be established by
397 combining housing and transportation costs. It has been estimated that housing is affordable if
398 the combined housing and transportation cost is less than 45% of household income. Using this
399 standard, policy makers could be guided on determining an appropriate location for siting an
400 affordable housing project. Moreover, it could also serve as a guide to household towards
401 identifying the best affordable location when choosing a home (Isalou et al., 2014). However,
402 one disadvantage of the location affordability cost standard is that it does not account for the
403 housing operation cost (utility bills) likely to be incurred by potential households. Therefore, a
404 new standard which is a combination of location affordability cost CSC and housing operation
405 cost CSC could provide more impact on affordable housing policy formulation. However,
406 providing a separate standard for household energy cost, Mattioli et al. (2018) stated that in
407 energy efficient affordable housing, 10% or less of household’s income on domestic energy is
408 the required standard. Therefore, the energy efficient standard for households will guide policy
409 makers on the appropriate energy efficient strategies to adopt to reduced high energy cost on
410 household income in addition to mitigating environmental impacts such as greenhouse gas
411 emission. Concerning time measurement CSC, one of its aims is to assess the desirability of a
412 dwelling unit among the target households (Pullen et al., 2010). A high take up rate indicates
413 that the housing facilities are desirable to the target household (Chen et al., 2014). However,
414 low take up rate suggests that some aspects of the housing facilities need to be improved.
415
416 Measurement of the subjective CSC – quality-related CSC, household satisfaction CSC and
417 stakeholder’s satisfaction CSC – can be conducted by using a Likert scale (Chan and Chan,
418 2004). For instance, using a five-point Likert scale (1-not satisfied to 5- very satisfied), the
419 satisfaction of households or stakeholders with regard to affordable housing projects could be
420 measured. A low satisfaction score could serve as a guide to policy makers that the housing
421 facility needs to be improved. Besides, reasons for low satisfaction could serve as a guide for
422 the construction of subsequent affordable housing projects.
423
424 5. Conclusions
425 The meaning of success, most often, changes from project to project. Determining whether an
426 affordable housing project is sustainable and therefore a success or a failure is far more
427 complex. This is because there are inadequate studies on identifying a comprehensive list of
428 CSC for assessing the sustainability and success of affordable housing projects. Consequently,
429 affordable housing is mostly assessed based on the price or rental cost, which creates a gap
430 between affordable housing and sustainable housing. Bridging this gap requires sustainable
431 CSC. This study aimed to investigate the CSC required for the provision of sustainable
432 affordable housing. A questionnaire of 21 CSC was administered globally to affordable
433 housing experts. Ranking, factor analysis and Pearson correlation were employed for data
434 analysis.
435

25
436 Findings of this study revealed that though there is high interest on other CSC (such as energy
437 efficiency of housing facility, reduced lifecycle cost of housing facility and environmental
438 performance of housing facility), price and rental cost CSC are the most highly ranked among
439 developed and developing countries. Besides, some of the identified CSC are significantly
440 correlated with one another. Furthermore, six factors were developed for bridging the gap
441 between sustainable housing and affordable housing: (1) household satisfaction CSC, (2)
442 stakeholder’s satisfaction CSC, (3) housing operation cost CSC, (4) time measurement CSC,
443 (5) location affordability cost CSC and (6) quality-related CSC.
444
445 A limitation of this study is that only the opinions of affordable housing experts were
446 considered. The views of households of affordable housing units were excluded. For further
447 studies, it is would be interesting to analyze the views of households on CSC for sustainable
448 affordable together with the views of academics and contractor. Besides, the sample size used
449 for this study is relative small. Therefore, future study with much larger responses could
450 employ statistical analysis such as ANOVA to compare and determine any statistical
451 differences among the views of the various affordable housing stakeholders.
452 Though the study has limitations, there are important contributions of the findings of the study
453 worth stating. The findings of the study have contributed to filling the knowledge gap in the
454 affordable housing literature by providing a comprehensive list of CSC for assessing success
455 in sustainable affordable housing. In addition, the identified CSC are evaluation criteria for
456 bridging the gap between sustainable housing and affordable housing. Moreover, real estate
457 developers, architects, international organizations and government agencies could rely on these
458 CSC for resource allocation in the provision of sustainable affordable housing. The identified
459 CSC could be used by policy makers for identifying suitable locations for affordable housing
460 projects. Furthermore, the CSC could be relevant to potential households in identifying the best
461 affordable location and the most energy efficient housing facility when choosing a home.
462 Moreover, by using the identified CSC from this study, policy makers could be informed of
463 the success level of projects and the possible improvement policies to reduce affordable
464 housing overhang. Finally, the time-measurement CSC could be used to measure the
465 distributional outcome of affordable housing for the achievement of social sustainability. One
466 critical policy implication of this study is that due to high cost of implementing sustainable
467 housing strategies to attaining the identified six components of CSC, incentives from the
468 government could motivate real estate developers to include energy efficient strategies and
469 other sustainability strategies in the design and construction of sustainable affordable housing
470 projects.
471
472 Acknowledgement
473 This paper forms part of a research project entitled “Affordable Housing Supply: A
474 Comparative Study between Developed and Developing Economies”, from which other
475 deliverables have been produced with different objectives but sharing common background
476 and methodology. The authors wish to gratefully acknowledge the Research Grants Council
477 (RGC) and the Department of Building and Real Estate, Hong Polytechnic University, for their
478 financial support in conducting this study. The authors are also thankful to the editors and
479 anonymous reviewers for their useful comments that helped to improve the quality of this
480 paper.
481 References
482 Abdul-Aziz, A. R., & Kassim, P. J. (2011). Objectives, success and failure factors of housing
483 public–private partnerships in Malaysia. Habitat International, 35(1), 150-157.

26
484 Adabre, M. A., & Chan, A. P. (2018). The ends required to justify the means for sustainable
485 affordable housing: A review on critical success criteria. Sustainable Development.
486 Adinyira, E., Botchway, E., & Kwofie, T. E. (2014). Investigating the Underlining Factors of
487 Critical Project Success Criteria for Public Housing Delivery in Ghana. In Proceedings
488 of the 17th International Symposium on Advancement of Construction Management
489 and Real Estate (pp. 527-538). Springer, Berlin, Heidelberg.
490 Ahadzie, D. K., Proverbs, D. G., & Olomolaiye, P. O. (2008). Critical success criteria for mass
491 house building projects in developing countries. International Journal of Project
492 Management, 26(6), 675-687.
493 Aksoy, U. T., & Inalli, M. (2006). Impacts of some building passive design parameters on
494 heating demand for a cold region. Building and Environment, 41(12), 1742-1754.
495 Al-Saadi, R., & Abdou, A. (2016). Factors critical for the success of public‒private
496 partnerships in UAE infrastructure projects: experts' perception. International Journal
497 of Construction Management, 16(3), 234-248.
498 Al-Tmeemy, S. M. H. M., Abdul-Rahman, H., & Harun, Z. (2011). Future criteria for success
499 of building projects in Malaysia. International Journal of Project Management, 29(3),
500 337-348.
501 Allouhi, A., El Fouih, Y., Kousksou, T., Jamil, A., Zeraouli, Y., & Mourad, Y. (2015). Energy
502 consumption and efficiency in buildings: current status and future trends. Journal of
503 Cleaner production, 109, 118-130.
504 Arditi, D., & Gunaydin, H. M. (1997). Total quality management in the construction
505 process. International Journal of Project Management, 15(4), 235-243.
506 Atkinson, R. (1999). Project management: cost, time and quality, two best guesses and a
507 phenomenon, it’s time to accept other success criteria. International journal of project
508 management, 17(6), 337-342.
509 Baccarini, D. (1999). The logical framework method for defining project success. Project
510 management journal, 30(4), 25-32.
511 Bassioni, H. A., Price, A. D., & Hassan, T. M. (2004). Performance measurement in
512 construction. Journal of management in engineering, 20(2), 42-50.
513 Bogdon, A. S., & Can, A. (1997). Indicators of local housing affordability: Comparative and
514 spatial approaches. Real Estate Economics, 25(1), 43-80.
515 Bouchlaghem, N. (2000). Optimising the design of building envelopes for thermal
516 performance. Automation in Construction, 10(1), 101-112.
517 Bubshait, A. A., & Almohawis, S. A. (1994). Evaluating the general conditions of a
518 construction contract. International Journal of Project Management, 12(3), 133-136.
519 Chan, A. P. C., Darko, A., Olanipekun, A. O., & Ameyaw, E. E. (2018). Critical barriers to
520 green building technologies adoption in developing countries: The case of
521 Ghana. Journal of cleaner production, 172, 1067-1079.
522 Chan, A. P., & Chan, A. P. (2004). Key performance indicators for measuring construction
523 success. Benchmarking: an international journal, 11(2), 203-221.
524 Chan, A. P., Scott, D., & Lam, E. W. (2002). Framework of success criteria for design/build
525 projects. Journal of management in engineering, 18(3), 120-128.
526 Chaplin, R., Martin, S., Yang, J. H., & Whitehead, C. (1994). Affordability: definitions,
527 measures and implications for lenders. University of Cambridge, Department of Land
528 Economy.
529 Chel, A., & Tiwari, G. N. (2009). Thermal performance and embodied energy analysis of a
530 passive house–case study of vault roof mud-house in India. Applied Energy, 86(10),
531 1956-1969.
532 Chen, J., Yang, Z., & Wang, Y. P. (2014). The new Chinese model of public housing: A step
533 forward or backward? Housing Studies, 29(4), 534-550.
27
534 Chiu, R. L. (2007). Planning, land and affordable housing in Hong Kong. Housing
535 Studies, 22(1), 63-81.
536 Chua, D. K. H., Kog, Y. C., & Loh, P. K. (1999). Critical success factors for different project
537 objectives. Journal of construction engineering and management, 125(3), 142-150.
538 Cox, W., et al. (2017). 13th annual demographia international housing affordability survey:
539 2017.
540 Dang, Y., Liu, Z., & Zhang, W. (2014). Land-based interests and the spatial distribution of
541 affordable housing development: The case of Beijing, China. Habitat International, 44,
542 137-145.
543 Darko, A., Chan, A. P. C., Yang, Y., Shan, M., He, B. J., & Gou, Z. (2018). Influences of
544 barriers, drivers, and promotion strategies on green building technologies adoption in
545 developing countries: The Ghanaian case. Journal of Cleaner Production, 200, 687-
546 703.
547 Desai, V. (2012). Urbanisation and housing the poor: overview.
548 Dezhi, L., Yanchao, C., Hongxia, C., Kai, G., Hui, E. C. M., & Yang, J. (2016). Assessing the
549 integrated sustainability of a public rental housing project from the perspective of
550 complex eco-system. Habitat International, 53, 546-555.
551 Elasfouri, A. S., Maraqa, R., & Tabbalat, R. (1991). Shading control by neighbouring
552 buildings: application to buildings in Amman, Jordan. International journal of
553 refrigeration, 14(2), 112-116.
554 Fan, Y., & Huang, A. (2011). How affordable is transportation? A context-sensitive
555 framework.
556 Ferguson, H., & Clayton, L. (1988). Quality in the constructed project: a guideline for owners,
557 designers and constructors. American Society of Civil Engineers.
558 Field, A. (2013). Discovering statistics using IBM SPSS statistics. sage.
559 Florides, G. A., Tassou, S. A., Kalogirou, S. A., & Wrobel, L. C. (2002). Measures used to
560 lower building energy consumption and their cost effectiveness. Applied Energy, 73(3-
561 4), 299-328.
562 Galster, G. C. (1985). Evaluating indicators for housing policy: Residential satisfaction vs
563 marginal improvement priorities. Social Indicators Research, 16(4), 415-448.
564 Gan, X., Zuo, J., Wu, P., Wang, J., Chang, R., & Wen, T. (2017). How affordable housing
565 becomes more sustainable? A stakeholder study. Journal of Cleaner Production, 162,
566 427-437.
567 Gilbert, A. (2016). Rental housing: The international experience. Habitat International, 54,
568 173-181
569 Golubchikov, O., & Badyina, A. (2012). Sustainable housing for sustainable cities: a policy
570 framework for developing countries.
571 Hamidi, S., Ewing, R., & Renne, J. (2016). How affordable is HUD affordable
572 housing? Housing Policy Debate, 26(3), 437-455.
573 Huang, Y., Niu, J. L., & Chung, T. M. (2013). Study on performance of energy-efficient
574 retrofitting measures on commercial building external walls in cooling-dominant
575 cities. Applied energy, 103, 97-108.
576 Isalou, A. A., Litman, T., & Shahmoradi, B. (2014). Testing the housing and transportation
577 affordability index in a developing world context: A sustainability comparison of
578 central and suburban districts in Qom, Iran. Transport policy, 33, 33-39.
579 Jusan, B. M. M. (2007). Identification of user's expectations in mass housing using means-end
580 chain research model. Journal Alam Bina, 9(4), 1-19.
581 Ke, Y., Wang, S., Chan, A. P., & Cheung, E. (2011). Understanding the risks in China's PPP
582 projects: ranking of their probability and consequence. Engineering, Construction and
583 Architectural Management, 18(5), 481-496.
28
584 Kramer, A. (2018). The unaffordable city: Housing and transit in North American cities. Cities.
585 Larose, D. T. (2006). Data mining methods & models. John Wiley & Sons.
586 Lee, W. L., & Yik, F. W. H. (2002). Regulatory and voluntary approaches for enhancing energy
587 efficiencies of buildings in Hong Kong. Applied Energy, 71(4), 251-274.
588 Lim, C. S., & Mohamed, M. Z. (1999). Criteria of project success: an exploratory re-
589 examination. International journal of project management, 17(4), 243-248.
590 Lin, S. (2012). The study of the low occupy rate of public rental housing based on Shanghai,
591 Nanjing, Wuhan, Zhengzhou. Price Theories Pract., 7, 21-22.
592 Mandelli, S., Barbieri, J., Mereu, R., & Colombo, E. (2016). Off-grid systems for rural
593 electrification in developing countries: Definitions, classification and a comprehensive
594 literature review. Renewable and Sustainable Energy Reviews, 58, 1621-1646.
595 Mattioli, G., Lucas, K., & Marsden, G. (2018). Reprint of Transport poverty and fuel poverty
596 in the UK: From analogy to comparison. Transport Policy, 65, 114-125
597 Mingfang, T. (2002). Solar control for buildings. Building and environment, 37(7), 659-664.
598 Mohit, M. A., Ibrahim, M., & Rashid, Y. R. (2010). Assessment of residential satisfaction in
599 newly designed public low-cost housing in Kuala Lumpur, Malaysia. Habitat
600 international, 34(1), 18-27.
601 Morrissey, J., Moore, T., & Horne, R. E. (2011). Affordable passive solar design in a temperate
602 climate: An experiment in residential building orientation. Renewable Energy, 36(2),
603 568-577.
604 Müller, R., & Turner, R. (2007). The influence of project managers on project success criteria
605 and project success by type of project. European management journal, 25(4), 298-309.
606 Mulliner, E., Smallbone, K., & Maliene, V. (2013). An assessment of sustainable housing
607 affordability using a multiple criteria decision making method. Omega, 41(2), 270-279.
608 National Summit on Housing Affordable, 2006 –Achieving a National Affordable Housing
609 Agreement: Background paper 2: Key Terminology and Indicators, National Summit
610 on Housing Affordability.
611 Nguyen, M. T., Basolo, V., & Tiwari, A. (2013). Opposition to affordable housing in the USA:
612 Debate framing and the responses of local actors. Housing, Theory and Society, 30(2),
613 107-130.
614 Nikolaidis, Y., Pilavachi, P. A., & Chletsis, A. (2009). Economic evaluation of energy saving
615 measures in a common type of Greek building. Applied Energy, 86(12), 2550-2559.
616 Opoku, R. A., & Abdul-Muhmin, A. G. (2010). Housing preferences and attribute importance
617 among low-income consumers in Saudi Arabia. Habitat international, 34(2), 219-227.
618 Osei-Kyei, R., & Chan, A. P. (2017). Perceptions of stakeholders on the critical success factors
619 for operational management of public-private partnership projects. Facilities, 35(1/2),
620 21-38.
621 Pacheco, R., Ordóñez, J., & Martínez, G. (2012). Energy efficient design of building: A
622 review. Renewable and Sustainable Energy Reviews, 16(6), 3559-3573.
623 Pallant, J. (2013). SPSS survival manual. McGraw-Hill Education (UK).
624 Pérez-Lombard, L., Ortiz, J., Coronel, J. F., & Maestre, I. R. (2011). A review of HVAC
625 systems requirements in building energy regulations. Energy and Buildings, 43(2-3),
626 255-268.
627 Pinto, M. B., & Pinto, J. K. (1991, June). Determinants of cross-functional cooperation in the
628 project implementation process. Project Management Institute.
629 Pocock, J. B., Hyun, C. T., Liu, L. Y., & Kim, M. K. (1996). Relationship between project
630 interaction and performance indicators. Journal of construction engineering and
631 management, 122(2), 165-176.

29
632 Pullen, S., Arman, M., Zillante, G., Zuo, J., Chileshe, N., & Wilson, L. (2010). Developing an
633 assessment framework for affordable and sustainable housing. Australasian Journal of
634 Construction Economics and Building, The, 10(1/2), 60.
635 Rankin, J., Fayek, A. R., Meade, G., Haas, C., & Manseau, A. (2008). Initial metrics and pilot
636 program results for measuring the performance of the Canadian construction
637 industry. Canadian Journal of Civil Engineering, 35(9), 894-907.
638 Rashvand, P., & Zaimi Abd Majid, M. (2013). Critical criteria on client and customer
639 satisfaction for the issue of performance measurement. Journal of management in
640 engineering, 30(1), 10-18.
641 Ren, H., & Folmer, H. (2017). Determinants of residential satisfaction in urban China: A multi-
642 group structural equation analysis. Urban Studies, 54(6), 1407-1425.
643 Riazi, M., & Emami, A. (2018). Residential satisfaction in affordable housing: A mixed
644 method study. Cities.
645 Ruparathna, R., Hewage, K., & Sadiq, R. (2016). Improving the energy efficiency of the
646 existing building stock: A critical review of commercial and institutional
647 buildings. Renewable and sustainable energy reviews, 53, 1032-1045.
648 Saberi, M., Wu, H., Amoh-Gyimah, R., Smith, J., & Arunachalam, D. (2017). Measuring
649 housing and transportation affordability: A case study of Melbourne, Australia. Journal
650 of transport geography, 65, 134-146.
651 Sambasivan, M., & Soon, Y. W. (2007). Causes and effects of delays in Malaysian construction
652 industry. International Journal of project management, 25(5), 517-526.
653 Stasiowski, F. A., & Burstein, D. (1994). Total quality project management for the design firm:
654 How to improve quality, increase sales, and reduce costs. John Wiley & Sons.
655 Stevens, J. D. (1996). Blueprint for measuring project quality. Journal of management in
656 engineering, 12(2), 34-39.
657 Stone, M. E. (2006). What is housing affordability? The case for the residual income
658 approach. Housing policy debate, 17(1), 151-184.
659 Susilawati, C., & Armitage, L. (2005). Barriers to building partnerships between major
660 stakeholders in affordable housing investment in Queensland. Pacific Rim Property
661 Research Journal, 11(3), 232-252.
662 Teck-Hong, T. (2012). Housing satisfaction in medium-and high-cost housing: The case of
663 Greater Kuala Lumpur, Malaysia. Habitat International, 36(1), 108-116.
664 Tighe, J. R. (2010). Public opinion and affordable housing: A review of the literature. Journal
665 of Planning Literature, 25(1), 3-17.
666 Torbica, Ž. M., & Stroh, R. C. (2001). Customer satisfaction in home building. Journal of
667 Construction Engineering and Management, 127(1), 82-86.
668 Wai, S. H., Yusof, A. M., & Ismail, S. (2012). Exploring success criteria from the developers’
669 perspective in Malaysia. International Journal of Engineering Business
670 Management, 4(Godište 2012), 4-33.
671 Yan, H., Elzarka, H., Gao, C., Zhang, F., & Tang, W. (2018). Critical Success Criteria for
672 Programs in China: Construction Companies’ Perspectives. Journal of Management in
673 Engineering, 35(1), 04018048.
674 Yik, F. W. H., Burnett, J., & Prescott, I. (2001). Predicting air-conditioning energy
675 consumption of a group of buildings using different heat rejection methods. Energy and
676 Buildings, 33(2), 151-166.
677 Yuan, J., Li, W., Zheng, X., & Skibniewski, M. J. (2018). Improving Operation Performance
678 of Public Rental Housing Delivery by PPPs in China. Journal of Management in
679 Engineering, 34(4).
680 Zhang, C., Jia, S., & Yang, R. (2016). Housing affordability and housing vacancy in China:
681 The role of income inequality. Journal of Housing Economics, 33, 4-14.
30
682 Zhang, X. (2004). Concessionaire selection: methods and criteria. Journal of construction
683 engineering and management, 130(2), 235-244.

31