2

Is the private sector more efficient? Big data analytics of construction waste
management sectoral efficiency

3

Jinying Xua, Weisheng Lub*, Meng Yec, Fan Xued, Xiaoling Zhange, and Billy Fook Pui Leef

1

4

This is the peer-reviewed post-print version of the paper:
Xu, J., Lu, W., Ye, M., Xue, F., Zhang, X. & Lee, B. F. P. (2020). Is the private sector
more efficient? Big data analytics of construction waste management sectoral
efficiency. Resources, Conservation and Recycling, 155, 104674. Doi:
10.1016/j.resconrec.2019.104674
The final version of this paper is available at: https://doi.org/10.1016/j.resconrec.2019.104674.
The use of this file must follow the Creative Commons Attribution Non-Commercial No
Derivatives License, as required by Elsevier’s policy.
5
6
7

Highlights

8
9

•

Big data can distinguish the public and private disparity with better certainty.

10

•

Overall, public and private sectors have no significant CWM efficiency difference.

11

•

The private sector outperformed their public counterpart in demolition projects.

12

•

The public sector performed superior in foundation and new building projects.

13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28

a

7/f, Department of Real Estate and Construction, The University of Hong Kong, Pokfulam,
Hong Kong, China. Email: jinyingxu@connect.hku.hk
b
7/f, Department of Real Estate and Construction, The University of Hong Kong, Pokfulam,
Hong Kong, China. Email: wilsonlu@hku.hk
c
7/f, Department of Real Estate and Construction, The University of Hong Kong, Pokfulam,
Hong Kong, China. Email: megan828@connect.hku.hk
d
5/f, Department of Real Estate and Construction, The University of Hong Kong, Pokfulam,
Hong Kong, China. Email: xuef@hku.hk
e
Yeung-B7309, Department of Public Policy, The City University of Hong Kong, Kowloon
Tong, Hong Kong, China. Email: xiaoling.zhang@cityu.edu.hk
f
Room 1202, Carnival Base Company Limited, Tsuen Wan, Hong Kong, China. Email:
fplee@carnivalbase.com.hk
*
Corresponding author.

29
30

1

31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50

Abstract
Efficiency disparity between the public and private sectors is a non-trivial issue that concerns
fundamental choices of socio-political-economic systems. Waste management academia and
industry also wrestle with issues relating to the choice between public and private sectors. To
examine the disparity exclusively caused by “sector”, in statistics language, one needs data that
is sufficiently big to control many other confounders, e.g., sites, project types, and construction
technologies. This paper attempts to ascertain the construction waste management (CWM)
efficiency disparity between the public and private sectors by using big data in Hong Kong.
The waste disposal records of 132 projects, including 70 public and 62 private projects, were
extracted and analysed. By comparing the waste generation flows (WGFs) and accumulative
WGFs, it is found that, by and large, there is no significant efficiency disparity in CWM
between the two sectors. However, a closer investigation discovered that the private sector
outperforms their public counterpart in demolition projects, while the latter performs better in
foundation and new building projects. Although there are private projects with higher CWM
performance, their divergence between the best and average projects are larger than public ones.
Such findings thus reject casual remarks that the private sector is more efficient in CWM. The
underlying reasons maybe the waste management index practice promoted in public projects
while the private sector is often incentivized to perform better CWM to save waste disposal
levies. Future research is recommended to delve into the causes of the efficiency disparity and
introduce CWM interventions accordingly.

51
52
53

Keywords: Public-private disparity; economic efficiency; construction waste management;
big data; Hong Kong

54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70

Introduction
Whether the public sector is more efficient than its private counterpart, or vice versa, is a nontrivial concern in many fundamental choices related to our political, social, and economic
systems (Sappington and Stiglitz, 1987; Meier and O'Toole, 2011; Mihaiu et al., 2010).
Majority of such research usually conducts comparative analyses, either qualitatively or
quantitatively, of the effects obtained in resources used. The comparisons even have an
implication to the choice of capitalism and socialism, privatization and nationalization in the
institutional economics. However, the debate on comparisons of private and public efficiency
has not been concluded. The comparison between the dynamism of private sector and the
inefficiency of public sector has been the principal political dogma of the past decades (Busch
and Gustafsson, 2002). The fanatic of private sector superiority acclaims that the compared to
public sector, private sector is more efficient and dynamic, while the public sector is slower
and more wasteful; that a higher proportion of private sector participation will make things
better (Simms and Reid, 2013). The neoliberalism also puts superiority in the private sector,
arguing it being efficient, dynamic, and modernize. While capitalism adherents are convinced
by the public sector with evidences that towering debts have been run up under privatization
2

71
72
73
74
75
76
77

and tremendous services are being brought back into public ownership. They think
privatization per se does not guarantee improved efficiency. There are also more and more
standpoints disputing that privatized services perform worse according to largest studies. The
reasons driving the efficiency disparity between the public and private sectors can be multiplex,
ranging from institutional, managerial, technical, and organizational, to information
accessibility aspects (Ring and Perry, 1985; Sappington and Stiglitz, 1987; Bretschneider, 1990;
Boyne, 2002; Bysted and Hansen, 2015).

78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96

Concerning the efficiency disparity between the public and private sectors, there are a rich vein
of works on different aspects. Some economists, from a theoretical perspective, have advanced
a strong argument that private firms are more efficient than public firms (Alchian, 1965; De
Alessi, 1974; Sheshinski and Lopez-Calva, 2003). They emphasized the importance of nontransferability of ownership and weakening of property rights in public sectors in support of
the efficiency disparity between public and private sectors. However, the priori theorized view
is weak in empirical confirmation with inconclusive results. Das (2012) confirms that in the
mining industry, the participation of the private sector can boost the overall productivity by
comparing the extraction efficiency of public and private mining firms. Bhattacharyya et al
(1994), on contrary, by analyzing public water utilities practice projects, offer evidence of
higher efficiency in public sector although they are more widely scattered between best and
worst. More empirical studies reveal that there is no convincing evidence of one form of
ownership systematically surpassing over another. For example, Byrnes et al (1986) failed to
find any evidence to support the superiority of privately owned utilities over publicly owned
ones by measuring efficiency directly in terms of a production function in the water utility
industry. The same result goes to Estache and Rossi (2002), who also choose samples of water
companies. Karas et al (2010) provide the results that the Russian public banks are not more
inefficient than private ones.

97
98
99
100
101
102
103
104
105
106
107
108
109
110

When shifting to the area of waste management, the empirical studies on the public-private
argument are limited (Xu et al., 2018). Ichinose et al (2013) find that prefectures where private
sector participated in household solid waste collection with a higher proportion are more
efficient from the solid waste logistics practices in Japan. By conducting case studies of waste
management in Lebanon, Massoud et al (2003) suggest a mixture of private and public sectors
without distinguishing the efficiency disparity between the two. Simões et al. (2012) evaluated
the productivity and efficiency of the waste sector based on 228 waste collection and treatment
utilities and that private sector participation does improve the efficiency in waste collection but
benefits ephemerally in waste treatments. Massoud et al. (2003) suggested that, from the cost
perspective, private sector provided services were between than public sector provided ones
based on the comparison of municipal solid waste collection in two largest cities. Jacobsen et
al. (2013) also reached a similar conclusion by the analysis of multiple household waste
collection service cases. Lu et al. (2016) applied the Coase Invariant Theorem as a guiding
3

111
112
113
114

theory to examine potential waste management performance disparity among the public and
private construction clients. Therefore, the debate on the efficiency disparity between public
and private sectors is far from concluding. Especially, when narrowing down to construction
waste management, confident answer is due with the support of empirical research.

115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134

According to previous research, the answer to the debate of efficiency disparity between the
public and private sector in waste management with empirical supports is still missing. This
paper aims to fill this research gap by answering the question whether the private sector is more
efficient by using big empirical data reflecting waste management performance in Hong
Kong’s construction industry. Empirical big data analytics from real cases of housing
development will be adopted as the research method. There are two rationales behind this
research method. First, it is to make the classic inquiry more manageable by narrowing the
scope down to construction waste management (CWM), which is one of the most oft-examined
areas concerning efficiency disparity between sectors. In contrast to previous studies, the
research reported here seeks to understand the specific question of efficiency disparity of CWM
between the public and private sectors. Second, it is to make good use of a set of big data of
CWM in Hong Kong, and other useful information, such as the project location and features,
that will guide better business predictions and decision-making (Lu et al., 2015). It is
anticipated that big data can help uncover hidden patterns and unknown correlations, control
the numerous factoring confounding the efficiency of an economic system to allow the
exclusive contribution of “sector” to surface. To this end, this paper aims to contribute to the
general question of efficiency disparity between sectors by introducing big data analytics,
hoping that big data can open a new avenue, through which the classic question can be brought
to a conclusion.

135
136
137
138
139
140
141
142
143
144
145

To be specific, the authors plan to select representative housing construction projects from both
the public and private sector, extract their waste treatment data from the open data sets, and
conduct comparative analyses of the public and private sector. The detailed research methods
will be explained in the Data and methods section. The rest of the paper is organized as follows.
After the introductory section is a brief literature review focusing on how big data can be used
in applied statistics to examine the classic question of efficiency disparities between sectors.
Section 3 presents the big data and the data processing methods, which include standardization
of the project data for visualization and statistical analyses. Section 4 is to report the data
analyses, results, and findings, followed by an in-depth discussion in Section 5. Conclusions
are drawn in Section 6.

146
147
148
149
150

Big data and better certainty
A consensus on the definitions of “big data” is yet to be reached. Researchers are converged to
adopt Gartner’s three definitional characteristics of big data: volume, variety, and velocity
(McAfee et al., 2012). Volume indicates the large quantity of the data; velocity indicates that
4

151
152
153
154
155
156
157
158
159
160

the data is incoming in a high speed, which requires prompt processing to harness its value;
and variety indicates big data must be rich in semantics in the forms of structured, unstructured,
semi-structured, or a combination thereof (Russom, 2011; Zaslavsky et al., 2013). In view of
the fact that big data does not mean quality data, researchers are increasingly emphasizing
‘veracity’ - how accurate or truthful a data set may be - as the fourth ‘V” of big data. Big data
analytics have been developed to analyze big data in order to discover hidden patterns, nonlinear relationships and casual effects that will guide better informed decision-making (Lu et
al., 2015). Likewise, according to Agrawal (2016) and WEF (2012), big data can lead to some
potential knowledge or guidance information that can be utilized for further decision-makings.
Hence, value is advocated as the fifth ‘V’ of big data.

161
162
163
164
165
166
167
168
169
170
171
172

Lu et al. (2018) critiqued that too often big data analytics is mistakenly associated with ‘pattern
detection algorithms’, ‘unsupervised machine learning’, ‘deep learning’, ‘artificial
intelligence’, NoSQL database, Hadoop, and other fascinating data mining methods. Instead,
they echoed with Leek (2014) that applied statistics should not be left out when harnessing the
value of big data. In the history of probability theory, there is a “law of large numbers”, which
is a theorem that describes the average value of the results retrieved from a large number of
trials should be close to the expected value and will become closer when more trials are
conducted (Sen and Singer, 2017). If treating an economic sector as a complex system, it would
be legitimate to expect that big data can approach the entirety of the sector as it operates (i.e.,
tries naturally), and allow the inquiry of their efficiency disparity, if there is any, to be
ascertained.

173
174
175
176
177
178
179
180
181
182
183
184
185
186

The potential of big data to ascertain something in a complex system is further hyped in stories
of the psephological analysis of big data about voter behavior, such as Donald Trump’s election
campaign in 2016 and Brexit in the same year. These scenarios present a lot of uncertainties,
e.g., to vote or not, or to leave or to remain. It has been reported that Cambridge Analytica, a
data analytics company, can understand the scenarios with greater certainties with its
enormously voluminous and various big data. Actually, they were reported to manipulate the
public towards a more certain direction which was wanted by their clients. Inspired by the
stories, big data analytics may open a new avenue to examine the classic inquiry on efficiency
disparity between the public and private sectors. In statistical language, the data, if big enough,
can control the numerous confounders to examine the efficiency disparity that is solely
contributed by “sector”. This potential, however, has not been well explored in the literature.
It is under the potential that this paper tends to examine the CWM efficiency difference
between the public and private construction objects.

187

5

188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204

Data and methods
Data
The Hong Kong Environmental Protection Department (HKEPD) launched a Construction
Waste Disposal Charging Scheme (CWDCS) in 2006. It requires that all construction waste,
unless being reused or recycled, must be disposed of at designated government waste disposal
facilities such as landfills, public fills, or off-site sorting facilities. Prior to using the facilities,
a main contractor who is awarded a contract with more than HK$ 1 million is mandated to open
a billing account in the HKEPD for the contract solely, with basic information of the project
including the contract name, client (e.g., public or private), contract sum, site address, type of
construction work, etc. When the construction waste is disposed of at the facilities, HKEPD
records information on every load of construction and demolition waste, including vehicle
number, time, and weight when the vehicle enters and exits, and the billing account number
the vehicle uses. Unintentionally, the practice in Hong Kong generates a large data set which
makes probing into many aspects of CWM, such as the performance of public and private
projects. Figure 1 illustrates the structure of the data sets, which contains the following five
types of databases:
•

Project database, contains basic information (including site addresses, clients, project
types, and other details) of all projects with an account to access the above government
facilities for waste disposal. There are a total of 27,536 construction projects being
recorded in this database.

•

Facility database, contains basic information (address, capacity) of all government
CWM facilities.
Vehicle database, contains 9,863 waste hauling vehicles for construction waste
transport.
Waste Disposal database, records detailed information of every truckload of
construction waste disposed at the government waste management facilities. There are
a total of 7,866,085 disposal records being generated by all the construction projects
executed during the eight-year period from 2011 to 2018.

205
206
207
208
209
210

•

211
212

•

213
214
215
216
217
218
219

•

Moreover, the Hong Kong Buildings Department (HKBD) keeps the Building data of
all existing and new building projects, including the address of the site, number of
blocks and storeys, building type, domestic and non-domestic gross floor area (GFA).
The data can also be integrated with the waste disposal data above.

220

6

221
222

Figure 1. The big data set

223
224
225
226
227

It is argued as big data, according to the five ‘Vs' criteria as described above, despite the volume
is not as big as terra- or zetta-bytes. It is also argued that big data should not be mechanically
equated to big volume, but it covers a fuller picture of the subject matter that is not possible
provided by small data.

228
229
230
231
232

Data processing methods
The data processing methods, as illustrated in Figure 2, comprises of eight connected steps
ranging from project selection, data extraction, processing, visualization, to detailed
comparisons.

233

7

234

Figure 2. The procedures of data extraction and analysis

235
236
237
238
239
240
241
242
243
244
245
246
247

Project selection
The first step is to select projects for the analysis. A set of ‘qualified” projects is selected from
the big data set by meeting the following criteria:
(1) The project must be started after 2011 and finished before 2019 to be reflected
completely in the big data set;
(2) Project type is restricted to demolition, foundation, and new building, as these types of
projects are the major source of construction waste; and
(3) The project must be relatively sizeable, as these projects allow more regular patterns
than their smaller counterparts, which are often impacted by random factors.
By applying the three criteria, 132 projects, including 19 demolition, 59 foundation, and 54
new building projects were sourced. They can be further divided into two groups: 70 public
and 62 private projects. Basic information of the selected projects is shown in Table 1.

248
249

Table 1 Basic information of the selected projects

8

Project type

No.
of
projects

No. of raw
records

Contract Sum (HK$)

Public demolition
Private demolition
Public foundation
Private foundation
Public new building
Private new building
Total

8
11
31
28
31
23
132

2,734
10,289
238,039
200,075
232,393
49,684
733,214

[1016392, 19008080]
[15756600, 55166000]
[4404854, 923000000]
[94022505, 644652285]
[339388300, 4711780000]
[36890955, 3780000000]
[1016392, 3780000000]

Project
duration
(weeks)
[7,
30]
[28, 155]
[40, 173]
[41, 161]
[109, 441]
[52, 398]
[7, 441]

Average project
duration (weeks)
16.9
55.9
78.8
77.0
220.0
171.1
122.0

250
251
252
253
254

Data extraction
With the account number of the selected projects as the reference, the waste disposal records
of the project can be queried and extracted from our database. The query is done by SQL server.
Afterward, all the extracted data are sorted on the project basis.

255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271

Standardizing time and waste generation weight
Since projects are different from each other in terms of project duration, the first step to process
the data is to standardize the time to make the projects comparable. The time of a project is
counted from its first record of waste disposal. For example, if the first record of Project A is
the 27th week of 2015, then we count from week 27 of 2015 and calculate the following time
based on this baseline. Suppose Project A ends at 10th week of 2017, then the duration of Project
A is 87 weeks taken that one year is 52 weeks. Different projects may start from a different
time, which means their baselines are different. The time is then standardized by percentage,
see formula (2):
𝑡𝑡
(2)
𝑇𝑇𝑖𝑖 % = 𝑖𝑖�𝑇𝑇 ∗ 100%
Where ti is the time point of a project; T is the total time of the project
Take Project A as an example, T = 87 weeks, the standardized time of the second week is T2%
= 2/87*100% = 2.3%. The above standardization method is on a weekly basis. Actually, the
data available allows to do it on a daily basis, but it turned out to be “over-engineered” by
having so many days. It is also too sparse to examine waste generation on a monthly basis,
therefore, the weekly scale is used.

272
273
274
275
276
277
278
279
280

Meanwhile, waste generation weight is also standardized, as the weight could range radically
from one project to another depending on their size. The approach is to treat the total weight
of waste generation of a project as 100%, and the weekly generation as a certain percentage of
the total waste generation of that specific project. Adding together the construction waste
weight of every vehicle using the same account number at one day, the total waste disposed of
by the corresponding project at that day can be calculated. The construction waste of every
project is further calculated on a weekly basis for easier analysis. The weight of weekly
construction waste was further accumulated to calculate the waste generated until a time point.
9

281
282
283
284
285

The ratio of the disposed waste at the corresponding week in the total waste generation of the
project is calculated using formula (3):
𝑤𝑤
𝑟𝑟𝑖𝑖 % = 𝑖𝑖�𝑊𝑊 ∗ 100%
(3)
where ri is the percentage of total waste generation at week i, wi is the waste generated at week
i, while W is the total waste generation.

286
287
288
289
290

By using the 𝑇𝑇𝑖𝑖 % as the x-axis and the 𝑟𝑟𝑖𝑖 % as the y-axis, the waste generation flow (WGF),
which portrays the changing of construction waste generation ratio with the changing of time,
can be plotted. The WGF curves can provide a clear and general visualization of the waste
generation performance of projects at different stages.

291
292
293
294
295

The accumulative percentage of disposed waste till week j is calculated using formula (4):
𝑗𝑗

𝐴𝐴𝐴𝐴𝑖𝑖 % = ∑1 𝑟𝑟𝑖𝑖 ∗ 100%

(4)

where APi is the accumulative percentage of waste generated till week i, pi is the percentage of
total waste generation at week i.

296
297
298
299
300
301

Similar to WGF, applying the 𝑇𝑇𝑖𝑖 % as the x-axis and the 𝐴𝐴𝐴𝐴𝑖𝑖 % as the y-axis, the accumulative
waste generation flow (accumulative WGF) can be drawn. Different from WGF, an
accumulative WGF curve brings a straightforward illustration of how waste is accumulated
during the project developing process, serving as another important indicator of construction
waste management efficiency.

302
303
304
305
306
307

Plotting weekly and accumulative WGF curves
With the standardized time (Ti%) as the x-axis, percentage of total waste generation (ri%) as
the y-axis, the weekly WGF curves can be drawn. Replacing the percentage of total waste
generation (ri%) with accumulative percentage of total waste generation (APi%), the
accumulative WGF curves can be portrayed.

308
309
310
311
312
313
314
315

Scaling up
The same approaches of data extraction, time and waste generation weight standardization and
WGF curve plotting are applied to all other projects. The data processing, analysis and scaling
up take place in Microsoft Excel 2016. Projects of the same type (there are a total of six types
of projects, e.g., public demolition, private demolition, public foundation, private foundation,
pubic new building, private new building) are arranged together for the representative WGF
curves for comparisons and analyses.

316
317
318
319
320

Averaging waste generation proportion with interpolation
To compare between the public and private clients, a representative curve for one type of
projects is desired. Since different projects have different time scale, all the project should be
averaged on a unified time scale. A 0.1% of scale is used to standardize the timeline to keep
10

321
322
323
324
325
326
327
328
329
330
331
332

the resolution of the data better. For the waste generation percentage at every time point, the
percentage covered a period is averaged under the assumption that the waste is evenly
generated for that period and then be disposed of once. For example, project B disposes of 15%
of waste at 25% of time after its last disposal at 15% of time, then, for every 0.1% of time
between 15% and 25% of the completion time, the averaged waste generation proportion is
15%/(25%-15%)*0.1% = 0.1%. By this averaging calculation, every project has its averaged
waste generation percentage at a 0.1% time point. Averaging the averaged waste generation
percentage of same type projects at every time points, an average waste generation percentage
value as their representative can be generated. For the accumulative WGF, linear interpolation
is used to fill the vacant value for the 0.1% of time scale representation. After acquiring all the
points on a 0.1% of time basis, the representative WGF and accumulative WGF curves can be
plotted.

333
334
335
336
337
338
339

Calculating the difference between the public and private sectors
After the interpolation, every type of project has a representative set of WGF and accumulative
WGF data set on the 0.1% time scale. Therefore, their difference can be calculated directly by
doing minus between the two sectors. Since this paper aims at investigating if private sector
outlaws, metrics of the private sector is used to minus that of the public sector and keep their
difference as the indicator of their waste management efficiency disparity.

340
341
342
343
344
345
346
347

Comparison
After processing the data, the results of public projects and private projects are compared
separately. Firstly, the overall waste generation rate (WGR) of all projects will be calculated
and displayed to investigate whether their WGRs have a significant difference. Afterward, the
WGFs and AWGFs of the public and private sector of demolition projects, foundation projects,
and new building projects will be compared one by one to examine their waste management
performance disparity.

348
349
350
351
352
353
354

Microsoft Excel 2016 is used to process and analyze all the data and further draw the figures.
It is capable of dealing with big data when the data is structured and can be converted in
different worksheets even for a 1-million-row dataset. Our 12,828-row data is structured and
can be converted between different worksheets and calculated easily using simple formulas.
By drawing the WGFs and accumulative S-curves, how the waste is generated during the whole
process of the project can be delineated.

355
356

Analyses, results, and findings

357
358
359
360

Demolition projects
Figure 3 displays the contrast between the WGFs in the 8 public and 11 private demolition
projects. The x-axis is the standardized time, and y-axis the standardized waste generation
11

361
362
363
364
365
366
367
368
369
370
371
372

percentages. Both types of projects have their own patterns as shown in Figures 3 (a) and (b):
private projects have evenly distributed proportions except three outliers; public projects have
more randomly distributed ratios with the change of time. By contrast, the weekly waste
generation percentages in public demolition projects are generally larger than private ones and
fluctuate more obviously. It is largely due to the longer duration and larger waste disposal times
of most private projects than the public projects. Figure 3 (c) is the comparison of the average
percentage of total waste generation between public and private demolition projects on a 0.1%time basis. On this meticulous time scale, the averaged WGFs of both types of projects fluctuate
dramatically. At the first 15% of the time, public demolition projects generate a much larger
proportion of waste than the private. Afterward, they change alternatively until around 80% of
the time, when private projects produce wastes at a far higher speed. These findings indicate
the better control of waste generation in private projects at early and middle stages.

373
374

375
376

Figure 3. Waste generation flow of public and private demolition projects

377
378
379
380
381
382

Figure 3 (d) further compare the private and public projects by showing their difference with
the changing of time. The y-axis is the difference between private and public average waste
generation percentage. The difference also waves greatly. Generally, it increases rapidly at 25%
of progress, then fluctuates gently to 80% of the time, and increases at high speed again in the
final period. Such results indicate that private projects generate wastes in a slower way at the
12

383
384
385
386
387
388
389

beginning but faster at the late stage than the public projects. However, the difference is
actually small within the range of 0.3% of total waste. A t-test (two samples assuming equal
variances) on the average percentage of total waste generation proves that the two sectors
perform equally (mean average percentage of total waste generation for public projects is 0.001
and private projects 0.000999) with no significant difference (P=0.971). From this aspect, the
performance of the public and private sector in demolition waste management projects have no
substantial dissimilarity.

390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410

The accumulative waste generation curves of public and private demolition projects are shown
in Figure 4. Different from Figure 3, the y-axis of Figure 4 is the accumulative waste generation
percentage. It is easy to recognize their difference from Figures 4 (a) and (b): the shapes of
private curves are more concentrated while public projects vary greatly from each other. To
compare their differences, the average accumulative percentage on a 0.1% of time basis of all
public demolition projects and all private demolition projects are calculated separately and then
plotted in the same figure, see Figure 4 (c). The dots are the actual averaged accumulative waste
generation ratios of public and private projects at different time points. Combining the four
sections in Figure 3, it can be interpreted that for private projects, the accumulative waste
generation is less than 10% in the first 20% of the time for most projects. Afterward, the
difference in waste generation speed among projects varies. For public projects, the situation
diverges more fiercely. The two curves in Figure 4 (c) is sketched by the averaged accumulative
percentage of waste generated after linear interpolation on a 0.1%-time basis. It represents the
difference between the public and private projects: the averaged accumulative waste generation
of public projects is higher than that of private ones. This difference is rooted in the early stages
of the first 20% time and keeps until 80% of the time when the difference is narrowed. Figure
4 (d) illustrates the difference in a more straight-forward way. Public projects cumulate an
average of nearly 30% more waste than private projects in the first 20% of the time. Afterward,
their difference narrowed in ups and downs. These features also echo the findings of WGF
comparisons.

411

13

412
413
414

Figure 4. Comparison of accumulative waste generation flow of public and private demolition
projects

415
416
417
418
419
420
421
422
423
424
425
426
427

Interpreting the finding from the comparisons of WGF and accumulative WGF between
demolition projects, the answer to the question “is the private sector more effective” is that the
private and public sector has no significant difference in waste management performance of
demolition projects. Yet, the investigations into details indicate that private sector control of
WGF during the demolishing process is more effective than the public. The waste management
performance here is considered by the control of waste generation. Moderate waste generation
is better for the capacity planning of waste disposal facility and the transportation arrangement
of contractors. Speedy waste generation in a compact society like Hong Kong will cause
pressure to public transportation, environment, as well as the waste disposal facilities. Besides,
the variance among private projects is much smaller than public projects. It implies the
consistency in waste management performance, which is critical for waste management
prediction and planning.

428
429
430
431
432
433

Foundation projects
The changes in the weekly percentage of total waste generation over time in foundation projects
are displayed in Figure 5. Figures 5 (a) and (b) show that there are some similarities and
differences in the WGFs of private and public projects. For most of the recorded points, the
percentage of total waste generation is less than 5%. More private projects have a high waste
14

434
435
436
437
438
439
440
441
442
443

generation proportion at the middle stage than public projects. Also, the peaks of the public
projects come later, one at 26% of the time and the others around 60-80% to completion. Figure
5 (c) further depicts their difference, which changes moderately with some big waves. A more
clear-cut result of the contrast is presented in Figure 5 (d). By and large, public/private projects
exceed the other alternatively within the range of 0.1%. Private projects outpace public ones
on the waste generation speed for a larger proportion of time at early, middle, and later stages.
Even though, a t-test on the average percentage of total waste generation by assuming a null
hypothesis that the two means are equal shows that the two sectors perform the same in
foundation projects on the aspect of waste management with the p-value being 0.97. The mean
percentage of total waste generation is 0.099 for public projects and 0.099 for private projects.

444

445
446

Figure 5. Waste generation flow of public and private foundation projects

447
448
449
450
451
452
453

Figure 6 illustrates the accumulative WGFs of public and private foundation projects.
Obviously, the distribution of the S-shape curves of public projects are more focused while the
private ones are sparser, as shown in Figures 6 (a) and (b). This major difference implies that
public projects are more standardized in waste management while the performance of waste
management in private projects staggers with each other. Figure 6 (c) further compare the two
sectors by contrasting their averaged accumulative percentage of total waste generation. The
15

454
455
456
457
458
459
460
461
462
463

distribution of dots demonstrates that private projects have more scarce distribution in
accumulative WGF, which is also evidenced in Figures 6 (a) and (b). The interpolated average
S-shape curves in Figure 6 (c) advance the comparison in a more pruned way. In general, the
gap between the public and private sector in foundation projects is slight when compared with
demolition projects. The accumulative percentage of total waste in private projects is larger
than that of public projects for most of the time. Figure 6 (d) shows the difference in average
accumulative waste generation percentage in a simple and direct way. At the peak of the curve,
private projects cumulate an average of 6% more waste at the middle stage. Public projects
reverse the trend at 75% of the time and reach the biggest difference of 3% accumulative waste
at 80% of the time.

464

465
466

Figure 6. Accumulative waste generation flow of public and private foundation projects

467
468
469
470
471
472
473
474

Closer probes into the comparison between the projects of the public and private sectors do
disclose some differences. The averaged WGFs and their difference curve illustrate the
divergences in waste generation efficiency between the two sectors: the public sector has a
smaller extent of variation than the private sector. The accumulative WGFs corroborate this
inference. Though averagely, the accumulative WGFs has a minor difference, the variance
among private projects is greater. On this aspect, the public sector has more consistent waste
management performance to a small degree.

475

16

476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496

New building projects
The WGFs of public and private new building projects are illustrated in Figure 7. Figures 7
(a) and (b) demonstrate that the waste generated in one week is less than 5% of the
corresponding total waste amount for most projects, be they public or private. Four private
projects produce more than 15% of total waste in one week, but their peaks happen at different
time. No public projects generate more than 10% of waste in one week, and their peaks also
occur at different time ranges. The difference is that all peaks happen before 72% of public
projects progress while randomly in private ones. The reason behind can be the standardization
of public projects which will result with controlled waste generation at early and middle stages
while the divergent standards among private projects themselves lead to the waste management
efficiency disparity. To better detect the difference between public and private projects, the
averaged WGFs are plotted on a 0.1% basis, as shown in Figure 7 (c). At the very early and
late stages, the difference is obvious: private projects generate waste quicker than their public
counterparts. But at middle stages, public projects generate more waste than private ones.
Therefore, Figure 7 (d) gives a clear and direct way to expand the comparison. Although for
most of the time, their difference fluctuates alternatively, it keeps within 0.1% for most of the
time, which is very slight. It is hard to compare the overall performance of the two-sector; thus,
a t-test on the average percentage of total waste generation is conducted. It indicates that the
two sectors have no significant difference in the average waste generation speed with a twotailed p-value of 0.892. Their means of average waste generation percentages are 0. 998 and
0.996, for public and private sector respectively.

497

17

498
499

Figure 7. Waste generation flow of public and private new building projects

500
501
502
503
504
505
506
507
508
509
510
511
512
513

The accumulative waste generation curves of public and private new building projects are
demonstrated in Figure 8. Similar to foundation projects, the distribution of the S-shape curves
of public projects are more concentrated while the private ones are more diverse, as shown in
Figures 8 (a) and (b). There are eight outliers in the private projects, among which five
generates waste very quickly at early stages (more than 65% of waste at 40% of the time) and
the other three very slowly at early stages (more than 65% of the time to produce 40% waste).
Figure 8 (c) presents a more direct comparison between the public and private sectors. It is
observable that private projects have more outlier points while public ones are more compact.
This difference might be rooted in the promotion of waste management index (WMI) practice
in public new building projects. The curves show that for the first half of the time, the
accumulative waste proportion of private projects is slightly larger than public projects at the
same progress. The difference between the two sectors keeps at an average of around 5%, as
shown in Figure 8 (d), which is moderate.

514

18

515
516

Figure 8. Accumulative waste generation flow of public and private new building projects

517
518
519
520
521
522
523
524
525
526
527
528
529
530
531

Interpreted from Figures 7 and 8, it is safe to draw a conclusion that the overall efficiency
difference between public and private sector is not significant although there are indeed some
slight differences during the process. For private projects, the performance between individual
projects varies more largely than public ones. However, the findings from WGFs and
accumulative WGFs is different from the results of the WGR comparison. This can be
interpreted that although public projects generate waste more consistently between different
projects, their overall waste management efficiency is still lagging compared to their private
counterparts. In other word, private projects generate less waste on the same contract sum.
Although the performance between one project and another may have a bigger difference, there
are some private new building projects that are highly efficient in construction waste
management. Therefore, it is necessary to enhance the regulation of private projects from the
waste management perspective to enhance the overall efficiency. The successful waste
management practices of public projects and some outstanding private projects should be
promoted.

532
533
534
535
536

Discussions
Both the private and public sectors sponsor construction project development. In Hong Kong,
a city evaluated as the world’s freest economy for 25 consecutive years in a row (Sum, 2019),
building construction clients use the same pool of contractors and sub-contractors, which are
19

537
538
539
540
541
542
543
544
545
546
547

all private companies freely competing in the market. Under the same set of environmental
regulations and construction codes, they should perform the same without an apparent disparity
between their CWM efficiency measured by WGRs. However, the big data analytics in this
study discovered that CWM efficiency varies with sectors and different types of construction
projects they undertake. According to the comparisons of CWM efficiency of three categories
of projects, the public and private sectors have no statistically significant difference in
demolition and foundation projects. Surprisingly, in the sampled new building projects, the
average WGR of the public sector is significantly larger than the private sector. In summary,
the private sector is not consistently more efficient than public, or vice versa, in managing
construction waste. Such results reject the significant efficiency disparity between the public
and private sectors concerning construction waste management.

548
549
550
551
552
553
554
555
556
557
558
559
560
561
562

Although under the same set of environmental regulations and construction codes, it is widely
considered that the public sector clients are subject to more stringent social scrutiny, therefore
tend to care more about CWM efficiency. The public sector clients are promoting waste
management index (WMI) practice in Hong Kong, which might be the reason that public
foundation and new building projects have much similar WGF and accumulative WGF
between each other while the divergence is large in private ones. On the private side, there are
often casual remarks in Hong Kong that the private sector outlaws in managing their
construction projects. However, the private sector cares more about the savings of material
consumption and disposal fees; therefore, better managing construction waste. All these joints
forces may explain the finding that no sector is consistently better than another. Nevertheless,
CWM performance during the whole process differs observably from one project to another.
Although the companies are under the same market conditions, CWM efficiency is determined
not only by the client-contractor dyadic relationship, but also by the companies’ environmental
awareness, corporate social responsibility, and management discretion.

563
564
565
566
567
568
569
570
571
572
573
574
575

The big data sourced in this study did provide a fuller picture to examine CWM performance,
and their detailed WGFs. The big data analytics helped to gain more statistical confidence,
although the analyses did not tell which sector is more efficient than another. Big data opens a
window to address the moot question of whether the private sector is more efficient than the
public sector, but such data can never be too big to examine a complex system such as CWM.
In this study, the data of 132 selected projects, including 70 public projects and 62 private
projects, of three categories, i.e., demolition, foundation, and new building are sourced from a
big data set. It is bigger than any other data sets one can ever see in the literature. Yet, it appears
thin in ascertaining the efficiency disparity between the public and private sectors. Researchers
are encouraged to exploit passive, unintentionally left-over big data to examine the classic
inquiry of efficiency disparity between the public and private sectors. Even in further future,
meta-analyses based on the big-data enabled analyses can bring the inquiry into a conclusion.

576

20

577
578
579
580
581
582
583
584
585
586
587

The methods and findings of this research will also supplement the body of knowledge by
adding a new method and perspective to the examination of public and private sectoral
efficiency disparity, as well as an answer to the investigation. It is hard to decide which one is
better, the better way is to analyze which sector is doing greater in which specific aspect and
improve the other sector using the experience and knowledge learnt from the empirical practice
comparison. For the field of practice, the answers obtained from empirical case analyses
indicate that the private sector needs more improvement than its public counterpart in the aspect
of CWM. Private companies should pay more attention to their social environmental
responsibilities and learn from the public counterparts. Although the public sector is doing
better at a general scale, some private companies are actually doing better, whose experiences
are worthy learning by the public sector and other private companies.

588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608

Conclusion
Efficiency disparity between the public and private sectors is a fundamental scientific inquiry
that is closely related to grand challenges such as capitalism vs. socialism, privatization vs.
nationalization, or public-private partnership. Over the past decades, it has attracted numerous
studies, but the line of inquiry is largely inconclusive. This research aimed to ascertain the
difference of construction waste management efficiency between the public and private
projects, with a view to understanding the inquiry in a smaller setting and considering big data
analytics as a promising solution to a moot question. In this study, the data of 132 selected
projects, including 70 public projects and 62 private projects distributed in three categories of
projects, including demolition, foundation, and new buildings are extracted and processed for
the statistical and visual comparisons. The investigations looking at the waste generation flow
(WGF) and accumulative WGF found that there is no statistical significance found in all three
types of projects, i.e., demolition, foundation, and new building projects. A comfortable
conclusion is that no sector is consistently more efficient than another in managing construction
waste, although closer inspections detect the better performance in private demolition projects,
public foundation and new building projects. Additionally, CWM performance ranges radically
from one project to another, especially in private projects, which suggests the standardization
of waste management practice in private projects as a feasible short-term strategy for CWM
improvement. Further, improving CWM performance perhaps can be pursued through
contractors’ individual practices rather than through the project ownership only.

609
610
611
612
613
614
615
616

Big data demonstrated its power to paint a fuller picture of CWM in different projects. It allows
us to obtain more statistical confidence, and to probe into detailed WGFs in individual projects.
However, the data is still not big enough to allow the examination of many relevant aspects,
e.g., contractors’ internal CWM policies and onsite practices. By having the data with sufficient
variety, it might be possible to ultimately answer the question of efficiency disparity between
the public and private sectors. This big, but still not various enough data forms the major
limitation of this research.
21

617
618
619
620
621
622

Acknowledgment
This research is supported by the Public Policy Research (PPR) (Project No.: 2018.A8.078.18D)
and Strategic PPR (Project Number: S2018.A8.010) Funding Schemes, both from the Policy
Innovation and Co-ordination Office of the Government of the Hong Kong Special
Administrative Region.

623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656

References
Agrawal, A., Cockburn, I., & McHale, J. (2006). Gone but not forgotten: knowledge flows,
labor mobility, and enduring social relationships. Journal of Economic Geography, 6(5),
571-591.
Alchian, A. A. (1965). Some Economics of Property Rights. II Politico 30 (4):816-29
Bhattacharyya, A., Parker, E., & Raffiee, K. (1994). An examination of the effect of ownership
on the relative efficiency of public and private water utilities. Land Economics, 197-209.
Busch, T., & Gustafsson, O. (2002). Slack in the Public Sector: A comparative analysis of a
private and a public enterprise for refuse collection. Public Management Review, 4(2),
167-186.
Allison, G. T. (1980). Public and private management: are they fundamentally alike in all
unimportant respects? (pp. 283-298). Cambridge, MA: John F. Kennedy School of
Government, Harvard University.
Boyne, G. A. (2002). Public and private management: what’s the difference?. Journal of
Management Studies, 39(1), 97-122.
Bretschneider, S. (1990). Management information systems in public and private organizations:
An empirical test. Public Administration Review, 536-545.
Byrnes, P., Grosskopf, S., & Hayes, K. (1986). Efficiency and ownership: further evidence.
The Review of Economics and Statistics, 337-341.
Bysted, R., & Hansen, J. R. (2015). Comparing Public and Private Sector Employees’
Innovative Behaviour: Understanding the role of job and organizational characteristics,
job types, and subsectors. Public Management Review, 17(5), 698-717.Das, A. (2012).
Who extracts minerals more efficiently—Public or private firms? A study of Indian mining
industry. Journal of Policy Modeling, 34(5), 755-766.
Daw, J. (2015). Explaining the persistence of health disparities: social stratification and the
efficiency-equity trade-off in the kidney transplantation system. American Journal of
Sociology, 120(6), 1595-1640.
De Alessi, L. (1974). An Economic Analysis of Government Ownership and Regulation:
Theory and the Evidence from the Electric Power Industry. Public Choice 19 (Fall): 1-42.
Estache, A., & Rossi, M. A. (2002). How different is the efficiency of public and private water
companies in Asia?. The World Bank Economic Review, 16(1), 139-148.
Galinienė, B., & Dzemydaitė, G. (2012). Spatial data envelopment analysis method for the
evaluation of regional infrastructure disparities. Social Technologies, 2(2), 390-403.
22

657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694

Hampson, K., Kraatz, J. A., & Sanchez, A. X. (2014). The global construction industry and
R&D. In R&D Investment and Impact in the Global Construction Industry (pp. 42-61).
Routledge.
HKHA (Hong Kong Housing Authority) (2008). Stepping forward in sustainable housing:
Sustainability Report 2007-2008. Accessed on 26 August 2019 at: https://bit.ly/2lVe0zy.
Ichinose, D., Yamamoto, M., & Yoshida, Y. (2013). Productive efficiency of public and private
solid waste logistics and its implications for waste management policy. IATSS Research,
36(2), 98-105.
Jacobsen, R., Buysse, J., & Gellynck, X. (2013). Cost comparison between private and public
collection of residual household waste: multiple case studies in the Flemish region of
Belgium. Waste Management, 33(1), 3-11.
Karas, A., Schoors, K., & Weill, L. (2010). Are private banks more efficient than public banks?
Evidence from Russia. Economics of Transition, 18(1), 209-244.
Leek, J. T. (2014). Svaseq: removing batch effects and other unwanted noise from sequencing
data. Nucleic Acids Research, 42(21), e161-e161.
Lu, W., Chen, X., Ho, D. C., & Wang, H. (2016). Analysis of the construction waste
management performance in Hong Kong: the public and private sectors compared using
big data. Journal of Cleaner Production, 112, 521-531.
Lu, W., Chen, X., Peng, Y., & Shen, L. (2015). Benchmarking construction waste management
performance using big data. Resources, Conservation and Recycling, 105, 49-58.
Lu, W., Webster, C., Peng, Y., Chen, X., & Chen, K. (2018). Big Data in Construction Waste
Management: Prospects and Challenges. Detritus, 4, 129-139.
Lu, W., Yuan, H., Li, J., Hao, J. J., Mi, X., & Ding, Z. (2011). An empirical investigation of
construction and demolition waste generation rates in Shenzhen city, South China. Waste
Management, 31(4), 680-687.
Lu, W. M., & Lo, S. F. (2007). A closer look at the economic-environmental disparities for
regional development in China. European Journal of Operational Research, 183(2), 882894.
Massoud, M. A., El-Fadel, M., & Malak, A. A. (2003). Assessment of public vs private MSW
management: a case study. Journal of Environmental Management, 69(1), 15-24.
McAfee, A., Brynjolfsson, E., Davenport, T. H., Patil, D. J., & Barton, D. (2012). Big data: the
management revolution. Harvard business review, 90(10), 60-68.
Meier, K. J., & O'Toole Jr, L. J. (2011). Comparing public and private management:
Theoretical expectations. Journal of Public Administration Research and Theory, 21,
i283-i299.
Mihaiu, D. M., Opreana, A., & Cristescu, M. P. (2010). Efficiency, effectiveness and
performance of the public sector. Romanian Journal of Economic Forecasting, 4(1), 132147.

23

695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727

Petersen, O. H., Hjelmar, U., Vrangbæk, K., & la Cour, L. (2012). Effects of contracting out
public sector tasks: A research-based review of Danish and international studies from
2000-2011.
PSIRU (Public Services Privatisation Research Unit). Public and private sector efficiency. Last
accessed on 03 September 2019at: https://bit.ly/2kJjcqd.
Ring, P. S., & Perry, J. L. (1985). Strategic management in public and private organizations:
Implications of distinctive contexts and constraints. Academy of Management
Review, 10(2), 276-286.
Russom, P. (2011). Big data analytics. TDWI best practices report, fourth quarter, 19(4), 1-34.
Sappington, D. E., & Stiglitz, J. E. (1987). Privatization, information and incentives. Journal
of Policy Analysis and Management, 6(4), 567-585.
Sen, P. K., & Singer, J. M. (2017). Large Sample Methods in Statistics (1994): An Introduction
with Applications. CRC Press, Boca Raton, USA.
Sheshinski, E., & Lopez-Calva, L. F. (2003). Privatization and its benefits: Theory, evidence
and challenges. In K. Basu, B. N. Pulin, & R. Ranjan (Eds.), Markets and Governments.
New Delhi: Oxford University Press.
Simms, A., & Reid, S. (2013). The private sector is superior'-time to move on from this old
dogma. The Guardian,(online) 25th April.
Simões, P., Cruz, N. F., & Marques, R. C. (2012). The performance of private partners in the
waste sector. Journal of Cleaner Production, 29, 214-221.
Sum, L.K. (2019). Hong Kong ranked world’s freest economy for 25th successive year, beating
Singapore at 2nd place and mainland China in 100th. SCMP, 25 Jan 2019. Last accessed
on 12 September 2019 at: https://bit.ly/2GeB84j.
WEF. (2012). Big Data, Big Impact: New Possibilities for International Development. World
Economic Forum. Accessed at 1 July 2019 at: https://bit.ly/1d4OZCf.
Xie, H., Chen, Q., Lu, F., Wang, W., Yao, G., & Yu, J. (2019). Spatial-temporal disparities and
influencing factors of total-factor green use efficiency of industrial land in China. Journal
of Cleaner Production, 207, 1047-1058.
Xu, J., Lu, W., Xue, F., Chen, K., Ye, M., Wang, J., & Chen, X. (2018). Cross-boundary
collaboration in waste management research: A network analysis. Environmental Impact
Assessment Review, 73, 128-141.
Zaslavsky, A., Perera, C., & Georgakopoulos, D. (2013). Sensing as a service and big
data. arXiv preprint arXiv:1301.0159.

24