probit analysis of planning statistics on case study zone author(s ... · to 2004). for the...
TRANSCRIPT
TitleProbit analysis of planning statistics on case study zoneseparation between other specified annotated business zonesand industrial zones in Hong Kong /
OtherContributor(s) University of Hong Kong
Author(s) Kwong, Wai-chun; 鄺為臻
Citation
Issued Date 2005
URL http://hdl.handle.net/10722/48872
Rights Creative Commons: Attribution 3.0 Hong Kong License
THE UNIVERSITY OF HONG KONG
PROBIT ANALYSIS OF PLANNING STATISTICS ON CASE STUDY:
ZONE SEPARATION BETWEEN OTHER SPECIFIED ANNOTATED
BUSINESS ZONES AND INDUSTRIAL ZONES IN HONG KONG
A DISSERTATION SUBMITTED TO
THE FACULTY OF ARCHITECTURE
IN CANDIDACY FOR THE DEGREE OF
BACHELOR OF SCIENCE IN SURVEYING
DEPARTMENT OF REAL ESTATE AND CONSTRUCTION
BY
KWONG WAI CHUN
HONG KONG
APRIL 2005
Declaration
I declare that this dissertation represents my own work,
except where due acknowledgment is made, and that it has
not been previously included in a thesis, dissertation or
report submitted to this University or to any other
institution for a degree, diploma or other qualification.
Signed : __________________________
Name : KWONG WAI CHUN
Date : April 2005
Witnessed by : ___________________________
(Professor Lawrence Wai-Chung Lai)
Acknowledgements
This dissertation was completed with the help and advice from many people.
Firstly, I would like to show my greatest gratitude to my supervisor,
Professor Lawrence Wai-Chung Lai, for his invaluable advice and guidance.
Without his continuous support, this dissertation would hardly be completed.
Secondly, I would also like to thank Miss Lin Yuet Yim, Veronica,
B.Sc.(Surveying), Ph.D. Candidate, for her valuable advice and assistance
on data analysis.
Thirdly, special thanks are given to the staff of the Planning Department
who handled my enquiries and my classmates who assisted me in many
ways.
I must also express my profound indebtedness to my family members for
their constant care and support.
- 1 -
Summary. This dissertation uses a Probit Model to evaluate non-aggregate
data to investigate zone separation between Other Specified annotated
Business (OU (Business)) Zones and Industrial (I) Zones. This dissertation
employs 1986 sets of town planning application data (with a span from 1975
to 2004). For the selected zones, it is found that getting planning permissions
in OU (Business) Zones from the Town Planning Board is easier than that in
the Industrial Zones. They also refute the supposed distinction between the
two selected zones in respect of commercial and office uses. The proposition
about rent-seeking activities in favour of large developers is refuted. The
findings prove that obtaining planning permissions in Sha Tin is easier. On
the contrary, it is difficult to obtain them in Aberdeen & Ap Lei Chau. The
proposition that a higher chance is associated for development after January
2001 by Town Planning Board is not refuted. The two common uses,
residential and office, are proved to stand higher chances of getting
permissions. It is also more difficult to obtain planning approvals on review
under s.17 after an application under s.16 is rejected in the first instance.
- 2 -
1. Introduction
Due to the changes in the global division of labour, Hong Kong industry
activities have shifted from manufacturing and production-oriented to
management and service-oriented activities1. Since 19 January 2001, the
Town Planning Board has introduced a new type of designated zone, the
“Other Specified annotated Business (OU (Business)” Zone, to allow greater
flexibility in new development and existing industrial or industrial-office
buildings2. Together with the existing Industrial “I” Zones, the town
planning intention of this new type of zones is to facilitate the continuous
growth in the economy and avoid the decay of existing industrial buildings.
Most past research was concerned with the development control of the
planning authority over the land market. Limited empirical studies were
based on planning application data using statistical methods. Only Dobry
(1975) and Brothern (1992a, 1992b) had conducted some empirical work
based on the aggregate application data.
1 Town Planning Board (2003). Town Planning Board Guidelines For Development
within “Other Specified Uses (Business)” Zone (TPB PG-NO. 22B of 31-5-2004),
Hong Kong: Printing Department. 2 ibid.
- 3 -
Other than Willis (1995), Tang and Tang (1999), Tang and Choy (2000),
Tang et. al. (2000), Lai and Ho (2001a, 2001c, 2001d, 2002a, 2002b, 2002c,
2003), Yung (2001, 2004), Kwan (2002), Ngai (2002), Chan (2003), Wan
(2004) who used probit/logit modelling. Most studies relied heavily on
population, economic, environmental figures to investigate the strategic
policies. Examples were there of Abercrombie (1945, 1948), Abercrombie et.
al. (1945) and Abercrombie and Plumstead (1949).
But Dobry’s (1975) work was an exception. He conducted qualitative
analysis of the impact of planning process and concluded that any delay in
planning process was costly. After the publication of the Dobry’s Report,
better and other statistical methods had been applied by different researchers.
Examples were: Brotherton (1984, 1992a, 1992b), Gilg and Kelly (1996),
McNamara and Healey (1984), Preece (1990), Sellgren (1990) and Willis
(1995). The trend was for abandoning aggregate in favour of non-aggregate
data in development controls analysis.
In Hong Kong, qualitative analysis began with Staley’s monograph (1994).
Staley (1994) investigated the planning mechanism of cost delay in the
Review of the Town Planning Ordinance 1992. His study discovered that the
transaction costs of the development would be added by the proposed
- 4 -
“planning certificates”. Developing the idea of Staley, Chau, Lai and
Hammer (1996) and Lai (1997) further reviewed the Town Planning
Ordinance 1996, and concluded that the planning mechanisms would lead to
higher transaction costs for the development and rent-seeking activities. For
the change of the use of non-aggregate data, (a) Tang and Tang (1999), Tang
and Choy (2000), Tang et. al. (2000) with the Hong Kong Polytechnic
University and (b) Lai and Ho (2001a, 2001c, 2001d, 2002a, 2002b, 2002c,
2003) and Chau and Lai (2004) with the University of Hong Kong had made
contribution.
Different researchers had their own preferences and characteristics in testing
their hypotheses. Gilg and Kelly (1996, p.205) classified the research on
development control statistics into four different categories.
1. simple statistical and cartographic analysis;
2. power struggle in decision making process analysis;
3. logical positivism analysis (technical exercises within decision making
processes); and
4. postmodernism analysis3
(random but sequence of event within
decision making processes).
3 The postmodernism approach involves modifying behaviour, hermeneutic, humanistic
or case study approaches in order to build a “tale” of how “things really are” rather
than a “meta-narrative’ based on modernist concepts of “order and pattern”.
- 5 -
Willis (1995a, 1995b) could be regarded as the first person who used
non-aggregate planning data in statistical study of planning control.
Non-aggregate data4 meant disaggregate data. Willis (1995a) used a logit
function to investigate for the likelihood of planning applications in respect
of waste disposal to be approved. In the same year, Willis (1995b)
developed two models for the analysis of development control. They were
the “Cognitive Continuum” and “Lens” models. Willis (1995b, p.1070) used
the “Cognitive Continuum” model to study the ability of people in
performing different kinds of work and the knowledge that was employed on
the work. For the “Lens” model (Willis 1995b, p.1071), “there is some
actual, underlying, hidden condition or state – planning permission – which
the planning officer (or planning committee) was trying to identify and
classify, from observable signs, cues or indicators it produces”. Willis (1995,
p.1070) mentioned that the development applications could be imagined as
producing some messages or signals which were different in such forms as
strength, frequency. Willis (1995) pointed out that the planning officers or
committees would try to follow or identify the hints or signals from the
proposed development projects in order to judge the applications.
4 The definition of Carols reads, “If each observation in our data set consists of an
attribute vector (representing an individual how has been interviewed), and an observed
choice, we say that we have disaggregated data. If, on the other hand, the data include
information on groups of people, we call it aggregate data.” (Carols (1979, p.6))
- 6 -
Many empirical studies had been conducted by researchers following in the
footsteps of Willis (1995b, p.1072). The functional forms of the hints in such
studies were: how the decision makers assigned the importance of the hints,
how to organize the data of the hints, what the relationship between the hints
and the decisions was, and whether there was consistency for the decisions
made.
In Hong Kong, the study on the so-called “Two-tier Plot Ratio” system by
Tang and Tang (1999) was the first application of the logistic regression
model to study the impact of new zoning on the redevelopment of private
housing. They utilized 627 sets of observation data (non-aggregate sampling)
from 1986 to 1997 to test their hypothesis. The hypothesis was whether the
time waiting for the approval of building plan and the “Consents to
Commence Work” from the local authority (Building Authority) will
become longer if the site area of the proposed redevelopment on the
combined site is smaller. Tang and Tang (1999, p.42) found that in urban
Kowloon, planning applications for the larger site had a greater chance of
success in obtaining permissions for uses in Residential (Group A) Zones
from Town Planning Board.
- 7 -
Tang and Choy (2000) made another attempt to investigate the criteria
involved in the decision making process and whether these criteria actually
followed some planning policies and guidelines. Tang and Choy (2000)
utilized 162 sets of observation data from 186 and 1997 and selected nine
independent variables (some of them from written planning policies or
guidelines) as the factors for evaluating the planning decisions of the Town
Planning Board. The results were similar to those of in Tang and Tang’s
(1999) that the larger the site and scale of a proposed development enabled
easier permission from the Town Planning Board.
Tang et. al (2000) further examined the development control regimes in
Hong Kong. They pointed out that actually Hong Kong’s planning control
system is a “Hybrid System” (Tang et. al (2000), p.2465). A “Hybrid
System” means the decision making is governed by the discretionary
permission process and the zoning plan. They developed sixteen variables as
factors in a logistic regression model and examined whether the “Hybrid
System” can provide flexibility for development control and give certainty
for development application on office use in Hong Kong Island. Seven out
of sixteen were found to be significant and they hence argued that there were
flexibility and certainty in the planning decision (Tang et. al (2000), p.2481).
- 8 -
2. Controls on planning development and the Hypotheses
The most readily observable statutory controls in Hong Kong are those
under the Buildings Ordinance and Town Planning Ordinance. Both are
regulated by a group of professionals when applications are submitted. Lai
(2001b) pointed out that the planning authority had the role of rationing the
development rights and such rationing was controlled by government town
planners. The Buildings Ordinance clearly states that the duty of Building
Authority is to maintain the safety and hygienic of the buildings according to
the designs, structural and safety requirements of the building plans
submitted for approval under Cap.123, s.16 of the Buildings Ordinance.
The Town Planning Ordinance authorizes the Town Planning Board to
regulate the redevelopment of land use in Hong Kong. So if any developer
intends to redevelop or change the use of land, they must check the relevant
statutory plan (i.e. OZP or DPA Plan) no matter the proposed development
uses are permitted by the government lease or not. Lai (1997a & 2002)
complained that such measures actually attenuated the property rights of a
land owner as they attenuated without compensation the agreed terms in the
lease.
- 9 -
Although there are ordinances, development guidelines and other
publications governing town planning control, a clear set of rules for
deciding planning application cannot be found. The planning legislation
does not state that decisions made by government town planners should be
based on “public interest”. However, this expression is often used in making
decisions by the Town Planning Board. But what the exactly meaning of
“public interest” is not clear. Moreover, planning guidelines and publications
are not statutory, and each planning application is said to be decided on a
“case-by-case basis” or its individual merits. All these suggest that planning
decisions can be heavily influenced by planners’ own preferences in their
professional judgment. Therefore, the local planning permission system can
be described as “self-governing” and insensitive to external influenced. Due
to the “secret” and “black-box” nature of the Town Planning Board, it would
be interesting for us to analyze or discover the actual factors or variables that
affected the outcomes of the planning application.
Hypothesis I investigates if the chance of success in obtaining planning
permissions on applications for developments in OU (Business) Zones is
easier to be approved by the Town Planning Board than those in Industrial
Zones. This is set in the background that the government is relaxing the
criteria for considering applications in order to restructure the industrial
- 10 -
sector of Hong Kong. The intention for introducing the OU (Business) Zone
is to allow more flexibility in industrial or industrial-office buildings. If the
hypothesis is not refuted, this would mean that Town Planning Board is
really closely following the Town Planning Board’s guidelines.
Hypothesis II examines if the planning applications for particular uses in OU
(Business) Zones are associated with a greater likelihood of being approved
by the Town Planning Board than that in Industrial Zones. It is conducted to
find out whether the two different zones are “separable” or not (Lai and Ho
2002b).
Hypothesis III evaluates if the chance of success in obtaining planning
permission on applications for development in larger sites (measured in
terms of proposed gross floor area (GFA) of the building(s) or use) is easier
to be approved by the Town Planning Board than those in smaller site. It is
based on the appraisal used by Lai and Ho (2001a, 2001b, 2001c, 2002a,
2003) about rent-seeking activities involved in the planning process. It is
assumed that for a larger site, more capital is required and this can act as a
proxy for the involvement of a larger developer. If the hypothesis is not
refuted, this will indicate that Town Planning Board is indeed favouring
larger developers and the decisions for cases are not really based on a
- 11 -
case-by-case, individual merits basis or following the guidelines.
Hypothesis IV is to find out if the planning applications for development in
particular regions are really associated with a greater likelihood of being
approved by the Town Planning Board than those of other regions in OU
(Business) and Industrial Zones. It is to investigate whether there is any
preference for any particular proposed uses in OU (Business) or Industrial
Zones by Town Planning Board when considering the planning applications.
It is proposed for investigating whether Town Planning Board will prefer
some particular regions so that easier permissions for case therein can be
obtained. If the hypothesis is not refuted, this will mean that all the regions
are considered as being the same by the Town Planning Board.
Hypothesis V investigates if the chance of success in obtaining planning
permissions for developments after January 2001 was easier than those
before January 2001. It evaluates whether the approval rates will be higher
or not after the introduction of the new type of zones. In theory, as the Town
Planning Board has relaxed the criteria for deciding applications since
January 2001; the likelihood of success in obtaining planning permission
should be higher. If the hypothesis is not refuted, this will mean that the
government allows more flexibility for developments in industrial or
- 12 -
industrial-office buildings.
Hypothesis VI examines if the planning applications for particular uses in
OU (Business) and Industrial Zones are associated with a greater likelihood
of being approved by the Town Planning Board than those of other uses in
OU (Business) and Industrial Zones. It investigates whether there is any
preference for proposed uses of Town Planning Board when considering the
planning applications.
Hypothesis VII evaluates if the chance of success in obtaining planning
permissions for developments for planning review under s.17(1) or planning
appeal under s.17(B) using the same proposal that fails in the s.16 process is
lower than that for planning application under s.16. The chance for obtaining
planning permission under s.17(1) or s.17(B) is very important for the
developers. If the probability of success on review or appeal is low,
developers may decide not to continue with the same application in order to
reduce further loss.
- 13 -
3. Probit Model
The Probit model was first utilized to study the effect of insecticides on the
insects in order to find out the critical dosage of the insecticides. Individual
insects were tested by applying an insecticide until the amount of insecticide
had reached a critical dosage and the insects all died after the critical limit.
Theil (1971) assumed that the proportion of insects killed was p, a probit
transformation was applied as y=F-1
(p), where y was the linear function to
the dosage of insecticides used. Lee and Trost (1978) followed the idea of
Probit Model and tried to use the model to investigate housing economics.
They criticized that many researchers or scholars had not fully utilized the
advantages of Probit Model. As the model could admit binary dependent
variables, many variables could be analyzed at the same time. Many earlier
works were only concerned with the expenditure of owners or renters. Lee
and Trost (1978) posed the hypothesis as to whether consumers would own
or rent their flats and the amount they would spend. This approach was for
the simultaneous determination of two factors. Since then, many research
projects based on simultaneous equations had been carried out. Examples
were the work of Goodman (1988) and Goodman and Kawai (1988), who
concluded that the choice of tenure (own/rent) and the housing decisions
were significant at the same time. They also mentioned that the income of
- 14 -
consumers was a significant variable for housing decision. One year later,
Potepan (1989) suggested that an increase in current interest rate could have
a positive effect on the improvement of housing. But an increase in income
would discourage the probability of housing improvement. Bourassa (1995)
used economic and demographic variables in a Probit Model to study their
effects on user costs and homeownership. The result was positive, which
meant that the variation in user cost could have a significant effect on the
probability of homeownership. Hsueh and Chen (1999) also applied the
Probit Model to examine the choice of tenure in Taiwan.
In Hong Kong, Lai and Ho (2001a, 2001c, 2001d, 2002a, 2002b, 2003), and
Chau & Lai (2004) evaluated different variables (such as proposed
development uses, site area, application time) using the Probit Model in
order to make predictions about the planning decisions. Their raw statistics
were obtained from the Town Planning Board files kept by Planning
Department. Lai and Ho (2001c) evaluated hypotheses about “zone
separation” in Hong Kong. In the work of Lai and Ho (2001d), 379 sets of
data from 1975 to 1998 were used to examine whether the likelihood of
obtaining planning permissions in Green Belt Zones by the Town Planning
Board in the proposed housing development project was greater or not.
Moreover, some evidence of rent seeking activities favouring larger
- 15 -
developers could also been found in the work of Lai and Ho (2001a, 2001b,
2001d & 2003), Kwan (2002), Ngai (2002), Chan (2003), Chau and Lai
(2004) and Wan (2004).
4. Modelling Planning Applications
Lai and Ho (2001c) mentioned that the Probit Model could be used to
identify the determinants of the choice between the two discrete alternatives
which was 1, 0 (it was convenient to identify “1” as approval and “0” as
rejection).
A univariate binary qualitative response model was expressed as Equation 1:
[Equation 1]
p(ai = 1) = f (xiββββ0) i ε {1, n}
where {ai} was a series of independent binary random variables in either 0
or 1; xi was the K-vector of known constants; and ββββ0 was the K-vector of
unknown parameters; f was a certain known function (Finney (1977)).
Ameniya (1986) suggested that it was more common to express the
probability in f(xi , ββββ0), but Equation 1 could be said as the most specification.
Moreover, specifying f as xiββββ0 was a more general approach in linear
regression model as xi could become the transformation of original
- 16 -
independent variables. f was just a distribution function and xi ββββ0 could also
approximated regarded as general transformation of non-linear function of
independent variables.
f could be used as function form in such statistic model as Probit Model,
Logit Model and Linear Probability Model. Equation 2 showed the Probit
Model for the planning applications.
[Equation 2]
p( xα1, xα2, …xαk) = f (β0 + β1xα1+β2xα2+…+ βk xαk)
Or equivalently be expressed as
[Equation 3]
f-1
[p( xα1, xα2, …,xαk)]= (β0 + β1xα1+β2xα2+…+ βkxαk)
The probability of successful planning application was modelled as a
function of particular development or use such as a resident use, commercial
use, office, hotel, bank, church, etc; the development size (measured in
=
- 17 -
terms of gross floor area); location as well as time. Let xα1, xα2,…, xαk be the
value taken by each of these k variables for αth planning application.
In order to estimate the parameters of β0, β1, β2,… and βk, the
maximum-likelihood model was used. The observation of the decisions was
defined in such a way that the first j’ applications were approved by the
Town Planning Board and the sequent j – j’ applications were rejected.
Therefore, the logarithmic form of the likelihood function could be
expressed as:
[Equation 4]
where p( xα1, xα2,…,xαk) was in form of Equation 2 and also the function of
β0, β1, β2,… and βk.
Thei (1971) and Tobin (1975) mentioned that by applying differentiation in
Equation 4 with respect to the above parameters and setting the derivatives
to zero, one could obtain a non-linear equation so that the estimates could be
derived numerically by an iterative procedure.
- 18 -
Amemyia (1981) pointed out that in a Log Likelihood Equation; there was
only 1 maximum, which meant that the equation was concave in nature. So
by applying an iterative procedure, our estimation could be converged into
one single maximum. And Long (1997) explained that the iterative
procedure should be started initially with a value and adding vectors of
adjustment in attempts afterwards could help improve the guess work. This
process would continue to run until convergence was found.
5. Data Description
From the planning statistics collected, there are 180 sets of planning
application data for the Other Specified annotated Business Zones from
2001 to 2004, while there are 1,951 sets of planning applications for
Industrial Zones from the period 1975 to 2004. Due to the missing of such
information as the proposed gross floor area, planning decision etc, some
sets of applications are excluded from the analysis. So, only 97 sets of
observations are used in Other Specified annotated Business Zones and
1,799 observations in Industrial Zones.
The dependent variable in the research is the “Decision” for the planning
applications from the Town Planning Board. The variable “Decision” equals
- 19 -
to 1 if the planning application is approved by the Town Planning Board. In
this dissertation, there are 25 of variables to be investigated compared with
the dependent variable. Each variable is the key factor postulated to have
affected the results of the application. The gross floor area variable, “GFA”,
is the proxy for the scale and size of the development. It is measured in
square meters. According to Lai and Ho (2001c), “GFA” was a much more
meaningful and reliable variable than “GSA” (used by Tang & Tang (1999),
Tang & Choi (2000)) in Model as the Accommodation Value (AV) of the
proposed development could be estimated on the basis of GFA. Moreover,
Lai and Ho (2001c) also pointed out that a larger value of “GFA” would
require greater capital investment, so this could act as a proxy for a larger
developer. This in turn throws light on the presence of rent-seeking
activities.
There are two zoning types or classes in this study: “Other Specified
annotated Business Zone” and “Industrial Zone”. So, a variable “OUB” for
the zoning area is assumed to equal 1 if the proposed development is located
in an OU (Business) Zone.
There are several common use variables which indicate the types of
proposed development applied for. They include: Residential (RES);
- 20 -
Commercial (COM); Office (OFF); Bank (BANK); Industrial-office (IO);
Showroom (SR); Hotel (HOTEL); Restaurant (REST) and other uses.
Moreover, in order to test the separation between the OU (Business) and
Industrial Zones, two new variables, “OUBCOM” and “OUBOFF”, are
defined. “OUBCOM” indicates the planning applications for commercial
uses in OU (Business) Zones. It is obtained by multiplying the two variables
of “OUB” and “COM”. “OUBOFF” represents planning applications for
office uses in OU (Business) Zones and is obtained by multiplying the
variables of “OUB” and “OFF”.
Geographic factors are also investigated under the Model. For broad regional
division, Hong Kong Island; Kowloon Peninsula (KLN) and the New
Territories (NT) are analyzed in the research. “KLN” equals 1 if the
proposed development is located in the Kowloon Peninsula and “NT” equals
1 if the proposed development is located in the New Territories.
For district division: Aberdeen & Ap Lei Chau (ABERDEEN); Chai Wan
(CW); Kwun Tong (KT); Ngau Tau Kok & Kowloon Bay (NTK); Cheung
Sha Wan (CSW); Tsz Wan Shan, Diamond Hill & San Po Kong (TWS); Sha
Tin (ST); Kwai Chung (KW); Tsuen Wan (TW) are differentiated.
- 21 -
The variable “S16” is used to identify the stage of an application. From the
collected data, there are 1741 observations on applications under s.16 and
155 observations under s.17(1) or s.17(B). “S16” equals 1 if no further
review or appeal cases arose after decisions were made by Town Planning
Board to reject approval under s.16.
Since 19 January 2001, OU (Business) Zones have been designated in order
to restructure industrial sector in Hong Kong. “AFTER01” equals 1 if the
planning decision is made after January 2001.
6. DISSUSSION
Table 5 in Appendix I presents the final linear form of Probit results of
independent variables. The variable “OUB” is significant at 0.90% level and
has a positive coefficient. This means that getting planning permissions in
OU (Business) Zones from Town Planning Board is easier than in the
Industrial Zones. Therefore, Hypothesis I is not refuted. It reveals that the
Town Planning Board has actually followed its own guidelines of relaxing
the development criteria for restructuring the industrial area.
“OUBCOM” and “OUBOFF” are two independent variables which are used
to evaluate Hypothesis II. Table 6 in Appendix I shows the Probit results of
- 22 -
the two independent variables, “OUBCOM” and “OUBOFF”, are 100% not
significant. This means that the commercial and office uses in OU (Business)
Zones are not associated with a higher chance to obtain planning
permissions. So, the two zoning classes (OU (Business) and Industrial Zones)
can be concluded as “inseparable” in respect of commercial and office uses.
Therefore, Hypothesis II is refuted.
Table 2 and Table 3 in Appendix I show that both the variables of “GFA”
and “1/GFA” are not significant. This result implies that gross floor area is
not a major factor governing the final decision of Town Planning Board.
This appears to contradict previous findings of Staley (1994), Hammer
(1996), Lai (1997), Lai and Ho (2001a), Kwan (2002), Chan (2003), Wan
(2004). Scale of proposed development is not significant in obtaining
planning permission for the types of used examined. Therefore, Hypothesis
III is refuted. There is no prima facie evidence is to support the argument
that development scale is affecting the decision of the Town Planning Board.
Therefore, no any rent-seeking activities have taken place to favour the large
developers in the types of uses examined.
For the investigation on districts, Kwai Chung is selected as the basis for
evaluating Hypothesis IV. From Table 5 in Appendix I, “ABERDEEN” and
- 23 -
“ST” are significant, respectively, at 3.26% and 0.63% levels. “ST”
possesses a positive coefficient, but the coefficient for “ABERDEEN” is
negative. This means that it is easier for developers to obtaining planning
permissions in Sha Tin. However, it is difficult for developers to getting
planning approvals in Aberdeen & Ap Lei Chau. Therefore, Hypothesis IV
is not refuted.
From Table 7 in Appendix I, the variable “AFTER01” is significant at
1.46% level and has a positive coefficient. This means that a higher chance
of success in planning application is associated with a proposed
development after January 2001. Therefore, Hypothesis V is not refuted.
Greater flexibility has apparently been granted for redeveloping the
industrial areas in order to suit the changing of manufacturing sector to
service-oriented sector in Hong Kong.
Besides “RES”, “OFF”, “BANK”, other uses’ variables (“COM”, “IO”,
“SR”, “HOTEL” and “REST”) are shown to be insignificant in the model.
From Table 6 in Appendix I, we can see that “RES” and “OFF” are
significant, respectively, at level of 3.49% and 5.78%, both with positive
coefficients. This means that it is easier to obtain planning permissions for
- 24 -
residential and office uses in OU (Business) and Industrial Zones than other
uses in the zones. Therefore, Hypothesis VI is not refuted.
Although the variable “BANK” is significant at level of 11.69%, it has a
negative coefficient. This means that getting planning approvals for bank
uses in OU (Business) and Industrial Zones is more difficult than other uses.
This result can provide some guidance for developers who are interest in
such uses in Industrial or OU (Business) Zones.
That “Office” uses stood a higher chance of success is logical because Hong
Kong’s resources are shifting from the industrial to serviced-oriented sector.
So the demand for office uses has increased. According to the guidelines, the
Town Planning Board should allow more flexibility in re-development and
facilitate the restructuring the industrial areas in Hong Kong. Therefore, it is
reasonable for Town Planning Board to have relaxed the criteria for
approving the office uses.
However, it is surprising to find that residential uses are associated with a
higher chance of success both in OU (Business) and Industrial Zones. Since
the OU (Business) and Industrial Zones are located in industrial areas, it is
not common to have residential uses in there. There are two possibilities for
this strange result. One is that the Town Planning Board does not closely
- 25 -
follow its own guidelines or instructions when making decision. The other is
that amount of applications is limited, thus reducing the significance of the
results.
The stage variable “S16”, as shown in Table 6 in Appendix I, is highly
significant and has a positive coefficient. This can reveal that if a planning
application made under s.16 is rejected, the success rate in a planning review
or an appeal will be very low consequently. Therefore, Hypothesis VII is
not refuted.
- 26 -
- 27 -
7. Conclusion
Zone separation is the focus of this dissertation. Even the label of a use in
Column 2 is identical in two zoning classes, the planning intention and thus
the decision criteria should be different. Otherwise, the two zones will
become inseparable, and the zoning regimes become meaningless in real
terms. In the hypotheses, commercial and offices uses are selected for
investigating the actual “separation” of the OU (Business) and Industrial
Zones. The result is surprising: there is no difference between the two
classes of zones. This means that make the introduction of newly designated
OU (Business) Zones is redundant as the planning intentions of the Town
Planning Board for them are actually the same regarding the proposed uses
in the two zoning classes.
No evidence is found supporting a large scale or small scale of proposed
development can affect the chance of success in planning application.
From the view of developers, the location of the proposed development will
directly affect the economic value. So, Sha Tin is the district that can have a
higher chance of success and may act as a reference for developers to invest.
And Aberdeen & Ap Lei Chau is a place in which it is difficult to apply for
planning approvals and this lower the incentive of developers.
- 28 -
Since January 2001, the planning applications have been more easily be
approved by the Town Planning Board due to the relaxation of the
development criteria. This may encourage developer to make more
proposals for development, so that the process of industrial area
restructuring can be speeded up.
Residential and office uses are the two popular uses from the view of Town
Planning Board as the approving rates are very high for both uses. But the
question is whether it is suitable to allow more residential uses in industrial
areas.
If an application fails under s.16, then the chance for success under s.17(1)
or s.17(B) would be much lower. This is normal since the Town Planning
Board is following the same principles when reviewing the case. Except
where the developers can provide new and strong evidence to prove that
their proposed developments are suitable to locate on the land, the chance of
rejection or dismissal is very high.
- 29 -
8. Limitations
The major limitation of the research is the restrictions in information
collected. Firstly, the OU (Business) Zones have only been introduced for 3
years. Therefore, the number of applications is quite small compared with
Industrial Zones. The great difference in the number of application
observations will definitely affect the results of the analysis.
Secondly, the missing of such information as “planning decision”, “gross
floor area” and “existing gross floor area” would make many data
inapplicable. The limited amount of the data reduces the significance and the
accuracy of the Probit Results.
Thirdly, there is the problem with the time horizons between OU (Business)
and Industrial Zones. The time period for OU (Business) Zones ranges from
2001 to 2004, but the time period for Industrial Zones from 1975 to 2004.
The difference in the time horizons may have affected the results. However,
if the same time horizon (from 2001 to 2004) is used, the available
observations would become too small for comparison, so it is worthwhile
compare the two zones according to different time horizons.
- 30 -
9. Further Research
Many interesting questions have been encountered during the research. The
guidelines published by the Town Planning Board have clearly stated that
the OU (Business) Zones can be perceived as a combination of commercial
and industrial zones. But why some of the industrial zones have been
changed to OU (Business) Zones, but no commercial zone changed to OU
(Business) Zones? And not all the industrial zones have been changed to OU
(Business) Zones. Just only some of them have been changed. Based on
what criteria or reasons for selecting the place or district to rezone? Further
analysis and research should be done in order to clarify the “actual” planning
criteria and intention of the Town Planning Board.
- 31 -
Bibliography
Abercrcombie, P. 1945. Greater London Plan 1944. London: HMSO.
_______. 1948. Hong Kong: Preliminary Planning Report. Hong Kong:
Government Printer.
_______. & Plumstead, D. 1949, A Civic Surveying and Plan for the City
and Royal Burgh of Edinburgh. Edinburgh: Oliver & Boyd.
_______., Owens. J & Mealand, A. 1945, A Plan for Bath. London: Pitman.
Aldrich, J.H. 1984. Linear Probability, Logit, and Probit Models. Beverly
Hills: Sage Publications.
Amemiya, T. 1981. Qualitative Response Models: a Survey. Journal of
Economic Literature 19(4): 1483-1536.
_______. 1986. Advanced econometrics. Oxford: Basil Blackwell.
Anderson, M.A. 1981. Planning Polices and Development Control in the
Sussex Downs AONB. Town Planning Review 52(1): 5-25.
Anderson, T.L. 1982. The New Resource Economics: Old Ideas and New
Applications. American Journal of Agricultural Economics 64(5):
928-934.
Benson, B.L. 1984. Rent Seeking from a Property Rights Perspective.
Southern Economic Journal 51(2): 388-401.
- 32 -
Blacksell, M. & Gilg, A.W. 1977. Planning Control in an Area of
Outstanding Natural Beauty. Social and Economic Administration
11(3): 206-215.
Booth, P. 1996. Controlling Development: Certainty, Discretion in Europe,
the USA and Hong Kong. London: UCL Press.
Bourassa, S.C. 1995. A Model of Housing Tenure Choice in Australia.
Journal of Urban Economics 37(2): 161-175.
Bramley, G., Bartlett, W. & Lambert, C. 1995. Planning, the Market and
Private Housebuilding. London: UCL Press.
Brotherton, D.I. 1982. Development Pressures and Control in the National
Parks 1966-1981. Town Planning Review 53(4): 439-459.
_______. 1984. A Response to McNamara and Healey. Town Planning
Review 55(1): 97-101.
_______. 1992a. On the Quantity and Quality of Planning Applications.
Environment and Planning B: Planning and Design 19(3): 337-357.
_______. 1992b. On the Control of Development by Planning Authorities.
Environment and Planning B: Planning and Design 19(4): 465-478.
_______. 1993. The Interpretation of Refusal and Appeal Rates. The
Planner 79(1): 24-25.
Buller, H. & Hoggart, K. 1986. Nondecision-making and Community Power:
Residential Development Control in Rural Areas. Progress in
Planning 25(3): 135-203.
33
Bunnell, G. 1995. Planning Gain in Theory and Practice Negotiation of
Agreement in Cambridgeshire. Progress in Planning 44(1): 1-113.
Carlos, D. 1979. Multinomial Probit, the Theory and its Application to
Demand Forecasting. New York: Academic Press.
Chan, T.S.F. 1999. Residential Construction and Credit Market Imperfection.
Journal of Real Estate Finance and Economics 18(1): 125-139.
Chan, K.W. 2003. Development Control in Agriculture Zone: a Probit
Analysis of Hong Kong Planning Statistics. Unpublished
B.Sc.(Surveying) dissertation, The University of Hong Kong.
Chau, K.W., Lai, L.W.C. & Hammer, A.M. 1996. An Economic Analysis of
the 1996 Town Planning Bill. Unpublished research monograph,
Centre of Real Estate and Urban Economics, Department of Real
Estate and Construction, University of Hong Kong.
Chau, K.W. & Lai, L.W.C. 2004. Planned Conversion of Rural Land: a Case
Study of Planning Applications for Housing and Open Storage Uses in
Agriculture Zones. Environment and Planning B: Planning and Design
2004, 31(6): 863- 878.
Chiu S.T.S. 2002. An Economic Inquiry on Planning: a Case Study of
Recreation Zoning in Hong Kong. Unpublished B.Sc.(Surveying)
dissertation, The University of Hong Kong.
Curry, N. 1992. Controlling Development in the National Parks of England
and Wales. Town Planning Review 63(2): 107-121.
Diamond, D. 1992. Evaluation the Effectiveness of Land Use Planning: a
Study for the Department of Environment. London: HMSO.
34
Dobry, G. 1975. Review of Development Control System: Final Report.
London: HMSO.
Finney, D.J. 1977. Probit Analysis. Cambridge: Cambridge University Press.
Gifford, A.Jr. 1987. Rent Seeking and Non-price Competition. Quarterly
Review of Economics and Business 27(2): 63-70.
Gilg, A. 1984. Literature Review: Editor’s Review. Countryside planning
yearbook 1984: 158-173. Norwich: Geo Books.
Gilg, A., & M. Kelly. 1996. The Analysis of Development Control
Decisions: A Position Statement and Some New Insights from Recent
Research in South-west English. Town Planning Review 67(2):
203-228.
Goodman, A.C. 1988. An Econometric Model of Housing Price, Permanent
Income, Tenure Choice, and Housing Demand. Journal of Urban
Economics 23(3): 327-353.
_______. & Kawai, M. 1988. Permanent Income, Hedonic Prices and
Demand for Housing: New Evidence. Journal of Urban Economics
12(2): 214-237.
Harrison, M.L. 1972. Development Control: The Influence of Political,
Legal and Ideological Factors. Town Planning Review 43(3): 254-274.
Healey, P., P. McNamara, M. Elson, & A. Doak. 1988. Land Use Planning
and the Mediation of Urban Change: The British Planning System in
Practice. Cambridge: Cambridge University Press.
35
Home, R. 1987. Measuring Trends in Town Planning Control through
Decision Statistics. Local Government Studies 13 (5): 51 – 62.
Hong Kong Government Lands Department, Lands Administrative Office.
2001. Practice Note: “Other Specified Use (Business)” Zones. Hong
Kong: Printing Department.
Hong Kong Government Planning Department. 1995. Town Planning in
Hong Kong: A Quick Reference. Hong Kong: Government Printer.
Hong Kong Government Planning Department. 2003. Hong Kong Planning
Standards and Guidelines Chapter 5: Industry. Hong Kong:
Government Printer.
Hsueh L.M. & Chen H.L. 1999. An Analysis of Home-ownership Rate
Changes in Taiwan in the 1980s. Asian Economic Journal 13(4):
367-388.
Hsueh, T.T. 1976. Growth Pattern and Structure Change of the Hong Kong
Economy. United College Journal 12-13(17): 293-326.
Keyes, J. 1986. Controlling Residential Development in the Green Belt: a
Case-study. The Planner 72(11): 18-20.
Kwan, K.Y. 2002. Modelling Planning Application Statistics in Hong Kong:
a Probit Analysis of Zone Separation of Unspecified Use and
Industrial (Group D) Zones. Unpublished B.Sc.(Surveying)
dissertation, The University of Hong Kong.
36
Lai, L.W.C. 1991. Some Economic Aspects of Agriculture in the New
Territories. Planning and Development 7(1): 42-44.
_______. 1994. The Economics of Land-use Zoning: a Literature Review
and Analysis of the work of Coase. Town Planning Review 65(1):
77-98.
_______. 1996. Zoning and Property Rights: A Hong Kong Case Study.
Hong Kong: Hong Kong University Press.
_______. 1997a. Town Planning in Hong Kong: A Critical Review. Hong
Kong: City University of Hong Kong Press.
_______. 1997b. Property Rights Justifications for Planning and a Theory of
Zoning. Progress in Planning 48(3): 161-246.
_______. 1998. The Leasehold System as a Means of Planning by Contract:
The case of Hong Kong. Town Planning Review 69(3): 249-275.
_______. 1999. Town Planning in Hong Kong: A Review of Planning
Appeal Decisions. Hong Kong: Hong Kong University Press.
Lai, L.W.C. & Fong, K. 2000. Town Planning Practice: Context, Procedure
and Statistics for Hong Kong. Hong Kong: Hong Kong University
Press.
Lai, L.W.C. & Ho, Daniel C.W. 2003. Some Design and Location Factors
for Cultural Ponds, Pools and Cages in Hong Kong. Aquaculture
Economics and Management 6(4): 249-275.
37
Lai, L.W.C., Ho, Daniel C.W. & Leung, H.F. 2004. Change in Use of Land:
a Practical Guide to Development in Hong Kong. Hong Kong: Hong
Kong University Press.
Lai, L.W.C. & Ho, W.K.O. 2001a. Low-rise Residential Developments in
Green Belts: a Hong Kong Empirical Study of Planning Applications.
Planning Practice & Research 16(3/4): 321-335.
_______. 2001b. Small is Beautiful: a Probit Analysis of Planning
Applications for Small Houses in Hong Kong. Environment and
Planning B: Planning and Design 28(4): 611-622.
_______. 2001c. A Probit Analysis of Development Control: a Hong Kong
Case Study of Residential Zones. Urban Studies 38(13): 2425-2437.
_______. 2001d. Zone Separation: a Probit Analysis of Hong Kong Planning
Application Statistics. Environment and Planning B: Planning and
Design 28(6): 923-932.
_______. 2002a. Planning for Open Storage of Containers in a Major
International Container Trade Centre: an Analysis of Hong Kong
Development Control Statistics using Probit Modelling. Environment
and Planning B: Planning and Design 29(4): 571-587.
_______. 2002b. An Econometric Study of the Decisions of a Town
Planning Authority: Complementary & Substitute Uses in Industrial
Activities in Hong Kong. Managerial and Decision Economics 23(3):
127-135.
_______. 2002c. Using Probit Models in Planning Theory: an Illustration.
Planning Theory 1(2): 146-162.
38
_______. 2003. Modelling Development Control of Residential
Development: a Probit Analysis of Rent Seeking and Policy
Autonomy in Town Planning in Hong Kong, in Columbus, F. (ed.),
Policy and Economics of Asia, Vol. VII, Ch. 4, Nova Science
Publishers, Huntington, NY, pp. 155-176.
Larkham, P.J. 1988. Changing Conservation Areas in the English Midlands:
Evidence from Local Planning Records. Urban Geography 9(5):
445-465.
_______. 1990a. Development Control Information and Planning Research.
Local Government Studies 16(2): 1-17.
_______. 1990b. The Use and Measurement of Development Pressure. Town
Planning Review 61(2): 171-183.
Lee, L.F. & R.P. Trost. 1978. Estimation of Some Limited Dependent
Variable Models with Application to Housing Demand. Journal of
Econometrics 8(3): 357-382.
Liu, H.L. 2003. Zone Separation: A Probit Analysis of Hong Kong Planning
Application Statistics Relating to Open Storage Use. Unpublished
B.Sc.(Surveying) dissertation, The University of Hong Kong.
Long, J.S. 1997. Regression Models for Categorical Limited Dependent
Variables, Thousand Oaks: Sage Publications.
McNamara, P. & Healey, P. 1984. The Limitations of Development Control
Data in Planning Research: A Comment on Ian Brotherton's recent
study. Town Planning Review 55(1): 91-97.
39
Mills, D.E. 1989. Is Zoning a Negative Sum Game? Land Economics. 65(1):
1-12.
Milroy, B. M. 1991. Into Post-modern Weightlessness, Journal of Planning
Education and Research 10(2): 181-188.
Ngai, T.H. 2002. An Analysis of the Statutory Planning Control Mechanism
in Hong Kong: A Probit Study of Agriculture Zones. Unpublished
B.Sc.(Surveying) dissertation, The University of Hong Kong.
Planning, Environment and Lands Branch. 1991. Comprehensive Review of
the Town Planning Ordinance – Consultative Document. Hong Kong:
Hong Kong Government Printer.
Planning, Environment and Lands Branch. 1996. Consultation Paper on
Town Planning Bill. Hong Kong: Hong Kong Government Printer.
Potepan, M. 1989. Interest Rates, Incomes, and Home Improvement
Decisions. Journal of Urban Economics 25(3): 282-294.
Pountney, M.T. & Kingsbury, P.W. 1983. Aspects of Development Control:
Part 1: the Relationship with Local Plans. Town Planning Review
54(2): 139-154.
Preece, R.A. 1981. Patterns of Development Control in the Cotswolds Areas
of Outstanding Natural Beauty. Oxford: University of Oxford.
_______. 1990. Development Control Studies: Scientific Method and Policy
Analysis. Town Planning Review 61(1): 59-74.
40
Sellgren, J. 1990. Development-control Data for Planning Research: The
Use of Aggregated Development-control Records. Environment and
Planning B: Planning Design 17(1): 23-37.
Staley, S. 1994. Planning Rules and Urban Economic Performance. Hong
Kong: Chinese University Press.
_______. 2001. Ballot-box Zoning Transaction Costs, and Urban Growth.
Journal of the American Planning Association 67(1): 25-37.
Tang, B.S. & Choy, L.H.T. 2000. Modelling Planning Control Decisions: a
Logistic Regression Analysis on Office Development Applications in
urban Kowloon, Hong Kong. Cities 17(3): 219-225.
Tang, B.S., Choy, L.H.T. & Wat, J.K.F. 2000. Certainty and Discretion in
Planning Control: a Case Study of Office Development in Hong Kong.
Urban Studies 37(13): 2465-2483.
Tang, B.S. & Tang, R.M.H. 1999. Development Control, Planning Incentive
and Urban Redevelopment: Evaluation of a Two-tier Plot Ratio
System in Hong Kong. Land Use Policy 16(1): 33-43.
Tewdwr-Jones, M. 1995. Development Control and the Legitimacy of
Planning Decisions. Town Planning Review 66(2): 163-181.
Theil, H. 1971. Principle of Econometrics. New York: John Wiley & Sons.
Tobin, J. 1975. Essays in Economics, Vol. 2: Consumption and
Econometrics. Amsterdam: North Holland Publishing Company.
41
Town Planning Board. 2003. Definitions of Terms Used in Statutory Plans
(updated on 12-11-2004), Hong Kong: Printing Department.
Town Planning Board. 2003. Town Planning Board Guidelines for
Development within “Other Specified Uses (Business)” Zone (TPB
PG-NO. 22B of 31-5-2004), Hong Kong: Printing Department.
Town Planning Board. 2003. Town Planning Board Guidelines for Use /
Development within “Industrial” Zone (TPB PG-NO. 25B of
30-9-2003), Hong Kong: Printing Department.
Tullock, G. 1993. Rent Seeking. London: Edward Elgar.
Underwood, J. 1981. Development Control: a Review of Research and
Current Issues. Progress in Planning 16(3): 175-242.
Watkins, C., Wale, C., Haines-Young, R. & Murdock, A. 2001.
Contextualizing Development Pressure: The Use of GIS to Analyze
Planning Applications in the Sussex Downs Area of Outstanding
Natural Beauty. Town Planning Review 72(4): 373-391.
Whitehand, J.W.R. 1989. Development Pressure, Development Control and
Suburban Townscape Change: Case Studies in South-east England.
Town Planning Review 60(4): 403-420.
Wan, C.H. 2004. An Analysis of the Statutory Planning Control Mechanism
in Hong Kong: a Probit Study of Other Specified Annotated Business
Zones. Unpublished B.Sc.(Surveying) dissertation, The University of
Hong Kong.
Willis, K.G. 1995. Judging Development Control Decisions. Urban Studies
32(7): 1065-1079.
42
Wong, Oi Yee 1999. The Evolution of Industrial Land Use Planning in
Hong Kong. Unpublished B.Sc.(Surveying) dissertation, The
University of Hong Kong.
Yung, P. 2001. Decisive Criteria in Development Control Decisions: a
Probit Inquiry. Unpublished B.Sc.(Surveying) dissertation, The
University of Hong Kong.
_______. 2004. Analysis of Development Control Statistics in Hong Kong:
Evaluating Factors of Success and Zone Separation. Unpublished
PhD. thesis, Faculty of Architecture, The University of Hong Kong.
Legislation
Dangerous Goods Ordinance. Chapter 295, Law of Hong Kong.
Personal Data (Privacy) Ordinance. Chapter 486, Law of Hong Kong.
Town Planning (Amendment) Ordinance. Chapter 131, Laws of Hong Kong.
i
Appendix I
Table 2: First Linear form of Probit Results of independent variables
Dependent Variable: DECISION Method: ML - Binary Probit Date: 12/27/04 Time: 23:15 Sample(adjusted): 5 1896 Included observations: 1886 Excluded observations: 6 after adjusting endpoints Convergence achieved after 9 iterations Covariance matrix computed using second derivatives
Variable Coefficient Std. Error z-Statistic Prob.
C 0.164809 0.157332 1.047524 0.2949
GFA -1.23E-07 1.30E-06 -0.094916 0.9244 OUB 0.198681 0.295064 0.673348 0.5007 RES 0.564583 0.294169 1.919246 0.0550 COM 0.040897 0.092628 0.441518 0.6588 OFF 0.122335 0.084764 1.443230 0.1490
BANK -0.191394 0.125480 -1.525293 0.1272 IO -0.052168 0.104579 -0.498837 0.6179 SR -0.051346 0.111217 -0.461670 0.6443
HOTEL 0.444570 0.369069 1.204571 0.2284 REST 0.084101 0.131637 0.638885 0.5229 S16 0.537224 0.111688 4.810042 0.0000
AFTER01 0.128933 0.192783 0.668799 0.5036 KLN 0.055124 0.103342 0.533408 0.5938 NT 0.050894 0.105672 0.481617 0.6301
ABERDEEN -0.299571 0.143646 -2.085481 0.0370 KT 0.042544 0.100060 0.425186 0.6707
CSW -0.078886 0.100259 -0.786824 0.4314 ST 0.265174 0.114105 2.323952 0.0201
TWS -0.000825 0.158823 -0.005194 0.9959 NTK 0.056670 0.361326 0.156839 0.8754 TW 0.238020 0.166217 1.431981 0.1521 CW 0.177554 0.214087 0.829351 0.4069
Mean dependent var 0.780488 S.D. dependent var 0.414026 S.E. of regression 0.407981 Akaike info criterion 1.039807 Sum squared resid 310.0935 Schwarz criterion 1.107395 Log likelihood -957.5377 Hannan-Quinn criter. 1.064697 Restr. log likelihood -992.5827 Avg. log likelihood -0.507708 LR statistic (22 df) 70.08992 McFadden R-squared 0.035307 Probability(LR stat) 6.40E-07
Obs with Dep=0 414 Total obs 1886 Obs with Dep=1 1472
ii
The McFadden R-squared value (the percentage that the independent
variables can explain the dependent variable) is 0.035307. And it is
surprising that the “GFA” variable is highly insignificant at 92.44%. This
shows that the probability of linear relationship between “GFA” and the
Decision of the planning application is low. In order to reduce the problem
of heteroscedasticity, the variable “GFA” is reciprocated into “1/GFA” in
Table 3.
In the model, only 3 variables: “RES” (5.50%), “ABERDEEN” (3.70%) &
“ST” (2.01%) are significant in the first linear test. All other variables are
shown insignificant in the result.
iii
Table 3: Second Linear form of Probit Results of independent variables
excluding “SR”, “AFTER01”, “KT”, “TWS”, “NTK”
Dependent Variable: DECISION
Method: ML - Binary Probit
Date: 12/27/04 Time: 23:18
Sample(adjusted): 5 1896
Included observations: 1887
Excluded observations: 5 after adjusting endpoints
Convergence achieved after 4 iterations
Covariance matrix computed using second derivatives
Variable Coefficient Std. Error z-Statistic Prob.
C 0.158148 0.149031 1.061175 0.2886
1/GFA 0.161176 0.603195 0.267204 0.7893
OUB 0.336421 0.228931 1.469529 0.1417
RES 0.563647 0.290954 1.937236 0.0527
COM 0.042319 0.090564 0.467285 0.6403
OFF 0.138081 0.081116 1.702270 0.0887
BANK -0.181282 0.121689 -1.489719 0.1363
IO -0.029324 0.096815 -0.302883 0.7620
HOTEL 0.459435 0.364650 1.259935 0.2077
REST 0.102220 0.129431 0.789765 0.4297
S16 0.533835 0.111353 4.794088 0.0000
KLN 0.058608 0.101103 0.579688 0.5621
NT 0.059787 0.104252 0.573491 0.5663
ABERDEEN -0.315850 0.138558 -2.279552 0.0226
CSW -0.101254 0.091600 -1.105400 0.2690
ST 0.256517 0.108526 2.363653 0.0181
TW 0.230376 0.161524 1.426270 0.1538
CW 0.113444 0.205327 0.552501 0.5806
Mean dependent var 0.780074 S.D. dependent var 0.414306
S.E. of regression 0.407872 Akaike info criterion 1.036163
Sum squared resid 310.9263 Schwarz criterion 1.089035
Log likelihood -959.6202 Hannan-Quinn criter. 1.055634
Restr. log likelihood -994.0981 Avg. log likelihood -0.508543
LR statistic (17 df) 68.95581 McFadden R-squared 0.034683
Probability(LR stat) 3.26E-08
Obs with Dep=0 415 Total obs 1887
Obs with Dep=1 1472
iv
The McFadden R-squared value is changed from 0.035307 to 0.034683; this
indicates that the effect of exclusion is little. After the reciprocal of the
variable of “GFA”, the value is still insignificant at 78.93%; this can prove
that the size of the proposed development does not affect much of the
decision for the planning permission of the Town Planning Board.
v
Table 4: Third Linear form of Probit Results of independent variables
excluding “1/GFA”, “COM”, “IO”, “KLN”, “NT”, “CW”
Dependent Variable: DECISION
Method: ML - Binary Probit
Date: 12/27/04 Time: 23:20
Sample(adjusted): 1 1896
Included observations: 1894
Excluded observations: 2 after adjusting endpoints
Convergence achieved after 3 iterations
Covariance matrix computed using second derivatives
Variable Coefficient Std. Error z-Statistic Prob.
C 0.222815 0.117465 1.896853 0.0578
OUB 0.352840 0.224417 1.572249 0.1159
RES 0.583800 0.283461 2.059539 0.0394
OFF 0.138498 0.073060 1.895675 0.0580
BANK -0.187993 0.116746 -1.610274 0.1073
HOTEL 0.210204 0.324907 0.646967 0.5177
REST 0.107872 0.126817 0.850614 0.3950
S16 0.526542 0.110415 4.768740 0.0000
ABERDEEN -0.314072 0.137006 -2.292392 0.0219
CSW -0.102906 0.090385 -1.138525 0.2549
ST 0.267453 0.105485 2.535469 0.0112
TW 0.223629 0.160056 1.397197 0.1624
Mean dependent var 0.779831 S.D. dependent var 0.414470
S.E. of regression 0.407911 Akaike info criterion 1.032898
Sum squared resid 313.1486 Schwarz criterion 1.068039
Log likelihood -966.1542 Hannan-Quinn criter. 1.045836
Restr. log likelihood -998.3685 Avg. log likelihood -0.510113
LR statistic (11 df) 64.42869 McFadden R-squared 0.032267
Probability(LR stat) 1.38E-09
Obs with Dep=0 417 Total obs 1894
Obs with Dep=1 1477
vi
Table 5: Final Linear form of Probit Results of independent variables
excluding “HOTEL”, “REST”, “CSW”
In the final results, nearly all the variables are significant. “OUB”, “RES”,
“OFF”, “BANK”, “ABEDEEN” and “ST” are significant at 0.90%, 3.49%,
5.78%, 11.69%, 3.26% and 0.63% levels respectively.
Dependent Variable: DECISION
Method: ML - Binary Probit
Date: 12/27/04 Time: 23:21
Sample(adjusted): 1 1896
Included observations: 1895
Excluded observations: 1 after adjusting endpoints
Convergence achieved after 3 iterations
Covariance matrix computed using second derivatives
Variable Coefficient Std. Error z-Statistic Prob.
C 0.203654 0.112141 1.816047 0.0694
OUB 0.444329 0.170039 2.613109 0.0090
RES 0.596750 0.282895 2.109439 0.0349
OFF 0.136440 0.071920 1.897112 0.0578
BANK -0.182854 0.116625 -1.567886 0.1169
S16 0.531834 0.109911 4.838781 0.0000
ABERDEEN -0.288833 0.135189 -2.136516 0.0326
ST 0.280650 0.102767 2.730925 0.0063
TW 0.243893 0.158585 1.537936 0.1241
Mean dependent var 0.779420 S.D. dependent var 0.414748
S.E. of regression 0.408172 Akaike info criterion 1.032477
Sum squared resid 314.2164 Schwarz criterion 1.058821
Log likelihood -969.2717 Hannan-Quinn criter. 1.042176
Restr. log likelihood -999.8810 Avg. log likelihood -0.511489
LR statistic (8 df) 61.21854 McFadden R-squared 0.030613 Probability(LR stat) 2.69E-10
Obs with Dep=0 418 Total obs 1895
Obs with Dep=1 1477
vii
Table 6: Probit Results of independent variables to investigate whether
OU (Business) Zones and Industrial Zones can be distinguished
as separate
Dependent Variable: DECISION
Method: ML - Binary Probit
Date: 12/27/04 Time: 22:58
Sample(adjusted): 5 1896
Included observations: 1890
Excluded observations: 2 after adjusting endpoints
Convergence achieved after 19 iterations
Covariance matrix computed using second derivatives
Variable Coefficient Std. Error z-Statistic Prob.
C 0.685146 0.043865 15.61944 0.0000
1/GFA 0.422657 0.643613 0.656694 0.5114
OUB 0.389106 0.187398 2.076355 0.0379
COM 0.062083 0.085617 0.725121 0.4684
OUBCOM 6.287131 143156.7 4.39E-05 1.0000
OFF 0.144096 0.069545 2.071970 0.0383
OUBOFF -0.132368 608531.7 -2.18E-07 1.0000
Mean dependent var 0.779894 S.D. dependent var 0.414427
S.E. of regression 0.413509 Akaike info criterion 1.051865
Sum squared resid 321.9741 Schwarz criterion 1.072400
Log likelihood -987.0124 Hannan-Quinn criter. 1.059426
Restr. log likelihood -996.1091 Avg. log likelihood -0.522229
LR statistic (6 df) 18.19340 McFadden R-squared 0.009132 Probability(LR stat) 0.005767
Obs with Dep=0 416 Total obs 1890
Obs with Dep=1 1474
From the results, the two variables “OUBCOM” and “OUBOFF” are 100%
not significant. This can indicate that the application uses of commercial and
office uses in OU (Business) Zones do not have a higher chance to be
approved by the Town Planning Board. Therefore, the two zoning classes
(OU (Business) and Industrial Zones) can be regarded as “inseparable” on
the commercial and office uses.
viii
Table 7: Probit Results of independent variable, “AFTER01”
Dependent Variable: DECISION
Method: ML - Binary Probit
Date: 12/27/04 Time: 23:22
Sample(adjusted): 1 1896
Included observations: 1896 after adjusting endpoints
Convergence achieved after 3 iterations
Covariance matrix computed using second derivatives
Variable Coefficient Std. Error z-Statistic Prob.
C 0.747163 0.033379 22.38455 0.0000
AFTER01 0.305092 0.124887 2.442932 0.0146
Mean dependent var 0.779536 S.D. dependent var 0.414669
S.E. of regression 0.414150 Akaike info criterion 1.053800
Sum squared resid 324.8596 Schwarz criterion 1.059652
Log likelihood -997.0022 Hannan-Quinn criter. 1.055954
Restr. log likelihood -1000.130 Avg. log likelihood -0.525845
LR statistic (1 df) 6.255817 McFadden R-squared 0.003128
Probability(LR stat) 0.012379
Obs with Dep=0 418 Total obs 1896
Obs with Dep=1 1478
The variable of “AFTER01” is significant at 1.46%. But the McFadden
R-squared is quite low which is only 0.003128. This may be resulted of the
limited amount of planning application data after January 2001.
ix
Appendix II
Fig. 1 (Photograph taken on 20-3-2005)
Fig. 2 (Photograph taken on 20-3-2005)
Figs. 1 & 2 show the proposed fast food development in Unit A8, G/F,
Koon Wah Mirror Group Building, 2 Yuen Shun Circuit, Siu
Lek Yuen, Sha Tin (File Reference: A/ST/603)
x
Fig. 3 (Photograph taken on 19-3-2005)
Fig. 4 (Photograph taken on 19-3-2005)
Figs. 3 & 4 show the proposed office development in Unit 1-7, 20/F, Chai
Wan Industrial City Phase I, Chai Wan (File Reference:
A/H20/129)
xi
Fig. 5 (Photograph taken on 20-3-2005)
Fig. 6 (Photograph taken on 20-3-2005)
Figs. 5 & 6 show the proposed development of temporary showroom
(display of autoparts) for a period of 3 years in Unit 1(Part),
G/F, Topsail Plaza, 11 On Sum Street, Sha Tin (File reference:
A/ST/585)
xii
Fig. 7 (Photograph taken on 20-3-2005)
Fig. 8 (Photograph taken on 20-3-2005)
Figs. 7 & 8 show the proposed development of temporary office (estate
agency) for a period of 3 years in Unit 2(Part), G/F, Topsail
Plaza, 11 On Sum Street, Sha Tin (File reference: A/ST/587)
xiii
Fig. 9 (Photograph taken on 20-3-2005)
Fig. 10 (Photograph taken on 20-3-2005)
Figs. 9 & 10 show the proposed development of temporary retail shop (toy
shop) for a period of 3 years in Unit 3(Part), G/F, Topsail Plaza,
11 On Sum Street, Sha Tin (File reference: A/ST/586)
xiv
Fig. 11 (Photograph taken on 19-3-2005)
Fig. 12 (Photograph taken on 19-3-2005)
Figs. 11 & 12 show the proposed hotel development in 24 Lee Chung Street,
Chai Wan (File reference: A/H20/126. The application was
rejected under s.16)
xv
Fig. 13 (Photograph taken on 20-3-2005)
Fig. 14 (Photograph taken on 20-3-2005)
Figs. 13 & 14 show the wholesale (stationery) development in Unit 9B,
Level 1, Wah Yiu Industrial Centre, 30-32 Au Pui Wan
Street, Sha Tin (File reference: A/ST/588)
xvi
Index of Names
Abercrombie……………………………… 3, 31
Abercrombie and Plumstead……………... 3, 31
Abercrombie et. al............................... …... 3, 31
Ameniya...................................................... 15, 31
Bourassa……………………………... …... 14, 32
Brothern……………………………........... 2, 32
Brotherton………………………….... …... 3, 32, 38
Chan…………………………………. …... 3, 15, 22, 33
Chau, Hammer and Lai…………………... 4, 33
Dobry………………………………........... 2, 3, 34
Finney…………………………………….. 15, 34
Gilg and Kelly…………………………..... 3, 4, 34
Goodman………………………………….13, 34
Goodman and Kawai………………........... 13, 34
Hsueh and Chen …………………………. 14, 35
Kwan……………………………………... 3, 15, 22, 35
Lai………………………………………… 3, 4, 8, 14, 22, 36
Lai and Ho………………………………... 3, 4, 10, 14, 15, 19, 22, 37
Lee and Trost……………………………... 13, 38
Long……………………………………… 18, 38
McNamara and Healey…………………… 3, 32
Ngai…………………………………......... 3, 15, 39
Potepan…………………………………… 14, 39
Preece…………………………………….. 3, 39
Sellgren………………………………........ 3, 40
Staley…………………………………....... 3, 4, 22, 40
Tang and Choy…………………………… 3, 4, 7, 19, 40
Tang and Tang……………………………. 3, 4, 6, 7, 19, 40
Tang et. al.………………………………... 3, 4, 7, 40
Theil……………………………………… 13, 40
Tobin…………………………………....... 17, 41
Wan………………………………………. 3, 15, 22, 41
Willis…………………………………....... 3, 5, 6, 42
Yung……………………………………… 3, 42
xvii
Index of Terms
Aggregate…………………………………………. 2, 3, 40
Case-by-case……………………………………… 9, 11
Column 2………………………………………….. 27
Development control(s)…………………………… 2, 3, 4, 5, 7, 31, 32, 33,
34, 37, 38, 39, 40, 41, 42
District(s)…………………………………………..20, 22, 27, 30
GFA………………………………………………..10, 19, 22
Hypothesis(es)……………………………………..6, 9, 10, 11, 12, 13, 21,
22, 23, 25
Industrial Zone(s)…………………………………. 1, 9, 10, 11, 12, 18, 19,
20, 21, 22, 23, 24, 27, 29,
30
Non-aggregate……………………………………..1, 3, 4, 5, 6
Other Specified Uses annotated Business Zone(s)... 1, 2, 18, 19, 41
OU (Business) Zone(s)……………………………. 1, 2, 9, 10, 11, 12, 19, 20,
21, 22, 23, 24, 27, 29, 30
Planning application(s)……………………………. 1, 2, 5, 6, 9, 10, 11, 12,
15, 16, 17, 18, 19, 20, 23,
25, 27, 28, 32, 33, 35, 37,
38, 41
Planning permission(s)…………………………….1, 5, 9, 10, 11, 12, 14, 21,
22, 23
Probit Model………………………………………. 1, 13, 14, 15, 16, 31, 37
Public interest……………………………………...9
Region(s)…………………………………………..11
Rent-seeking……………………………………….1, 4, 10, 19, 22
Statutory plan(s)……………………………….......8, 41
s.16 ………………………………………………... 1, 8, 12, 21, 25, 28
s.17(1)……………………………………………... 12, 21, 28
s.17(B)…………………………………………….. 12, 21, 28
Town Planning Board……………………………...1, 2, 6, 7, 8, 9, 10, 11, 12,
14, 17, 18, 19, 21, 22, 24,
27, 28, 30, 41
Zone Separation……………………………………1, 14, 27, 35, 37, 38, 42
Zoning……………………………………………...6, 19, 22, 27, 36, 39, 40