The effects of ‘publicness’ and quality of publicly accessible open space upon user satisfaction

Abstract

There is a worldwide trend for planning bodies to allow, require, or encourage developers to provide open spaces in their properties for public enjoyment to complement the open spaces on government lands. The two types of privately owned open space and government-owned open space, both accessible to public, can influence on people’s quality of life. Whether and how the land ownership of open spaces would influence their performance and, in turn, affect user satisfaction has not yet been studied and is worthy of exploration. This paper aims to examine the interactive relationships among three variables: ‘publicness’, quality/performance, and user satisfaction in both types of open space in Hong Kong. A questionnaire survey was administered to users of privately owned open space and government-owned open space and found that satisfaction with government-owned open space was marginally higher than with privately owned open space, with respect to activity facilities, amenities, and the overall utilization. Nevertheless, the differences in satisfaction were mainly the result of open space quality rather than land ownership per se. This study also examines the determinants of satisfaction, open space provision, and management and trends in spatial use. Findings from this study can deepen the understanding of utilizing both government-owned open space and privately owned open space from the user’s perspective, which will help improve the planning, design, and management of open spaces in cities.

Keywords

Open and green space property right public aspiration compact city structural equation modelling sustainable urban development

Introduction

Due to the fiscal constraints for providing and managing adequate, well-designed open spaces for the public, there is in the West a trend for government planning authorities to encourage or require developers and NGOs to participate in open space construction and/or management through public–private partnerships or private ownership (Gomes, 2019; More, 2005; Nemeth, 2009). Open space accessible to the public (OSAP) can be provided on land with different ownerships. Government-owned open space (GOS) is a traditional form of OSAP that is publicly owned and managed. Privately owned open spaces (POSs) are private spaces open to the public and are increasingly playing an important role in shaping urban environments. Around 123 POSs have been built in New York City since the 1960s to balance the distribution of all forms of OSAP throughout the city (Yoon and Srinivasan, 2015). Hong Kong (HK), with a highly commodified housing market, testifies to the global drive to privatize open space provision by relying on developers as the providers of urban open space. HK’s planning authority, the Town Planning Board, has encouraged POS within comprehensively planned developments (notably private housing estates or blocks in Comprehensive Development Area Zones) by way of planning conditions, interpreted as the actual private supply of planning by developers that satisfies both public interests and profit incentives (Lai et al., 2007; Tang and Wong, 2008).

This form of privatization implies a conceptual change in the publicness of open space. Publicness was defined by Nemeth and Schmidt (2011) as the extent to which a space is open to general public in terms of management, ownership, and users. While the three dimensions are accountable for the assessment of a space’s publicness, it allows for a bracketing of one or more axes according to empirical differences (Nemeth and Schmidt, 2011). Although publicness has been found to not always be directly related to public or private land ownership (Karacor, 2016), land tenure is the basic and determining factor for publicness and the (dis)advantages of POS over GOS have attracted researchers’ attention.

POSs have been criticized for their restrictions on openness, exclusiveness, inferior accessibility, and greater supervision requirements with increased control over use (Chow et al., 2016; Nemeth and Schmidt, 2011), leading to deleterious effects on citizenship and representation (Nemeth, 2009). On the other hand, management effectiveness and conviviality in space seem to have improved due to privatization (Gomes, 2019; More, 2005). Nevertheless, how alternative ownership regimes (private versus government) influence peoples’ perceptions and wellbeing has seldom been studied.

Satisfaction with the quality of OSAP appears to be a more important predictor of wellbeing than does its quantity (McEachan et al., 2018). Previous studies found the relationship between the performance attributes of OSAP and user satisfaction to vary considerably by city and case. For public spaces in South-eastern European cities, attributes associated with the facility, management, and misbehaviour of other users are more important for user satisfaction than issues related to physical attributes such as size and connectivity (Ostoic et al., 2017). Attributes of noise, conservation, shade features, and the relationships between users can explain 71.4% of the overall satisfaction on selected OSAPs in a Spanish city (Rey Gozalo et al., 2018).

In HK, users are in favour of more trees, dispersed seating designs, and larger parks (Lo and Jim, 2012), while their satisfaction is associated with the perceived quality of such spaces. This perceived quality is predictable through the factor, ‘features’, which refers to the design, facilities, and management of OSAP (Wan and Shen, 2015). Among the various performance attributes that affect user satisfaction, what attributes are related to the publicness of OSAP are questionable, as differences in the performance of POS and GOS were observed.

Variations in the quality and performance of GOS and POS exist in HK, where the two types of OSAP are planned, designed, and managed under two different mechanisms. GOS is provided by the Leisure and Cultural Services Department, while POS is provided by private developers following the Lands Department’s issue of a land lease. The latter is a new phenomenon associated with a global real estate revolution (Lai and Lorne, 2019). Does the publicness of OSAP influence their perceived performance and improve user satisfaction? If so, how? With these research questions in mind, the authors explore the interactive relationships among the three variables of public open space: ‘publicness’, quality/performance, and user satisfaction.

A theoretical framework is first established to illustrate the interactive relationships between the three variables. Then a questionnaire survey was conducted in selected GOS and POS in HK to investigate user satisfaction with the multi-dimensional attributes of the open spaces. Based on the responses, the relationships among publicness, quality, and satisfaction are analysed through structural equation modelling (SEM). Whereas earlier research focused on satisfaction with OSAP either collectively or on one speciﬁc attribute, the current research innovatively applies the SEM to several attributes of OSAP performance in four dimensions (e.g. activity facility, environment, amenity, and safety) to identify which dimension is (or is not) associated with publicness and has contributed to satisfaction. The results are discussed in relation to the determinants of satisfaction and differences in various kinds of OSAP, followed by the policy implications of this study.

Theoretical framework and research methods

This section presents our research framework (‘Research framework’ subsection), the features of a social survey (‘Selection of OSAP’, ‘Satisfaction questionnaire design’, and ‘Onsite OSAP survey’ subsections), and the statistical analysis with SEM (‘Statistical method and SEM’ subsection).

Research framework

To figure out how satisfaction with OSAP forms, the ‘Expectation Confirmation Theory’ (ECT) is applied to establish the research framework of this paper. ECT is a cognitive theory that seeks to explain post-purchase or post-adoption satisfaction as a function of expectations, perceived performance/quality of product, and the disconfirmation of beliefs (Oliver, 1980). The theory has been widely applied to interpret consumers’ satisfaction with privately provided goods, and it was also been used to study public satisfaction on public goods such as state highway (Poister and Thomas, 2011). ECT posits that satisfaction is directly influenced by disconfirmed beliefs and perceived performances, while it is indirectly influenced by a mediational relationship that passes through a disconfirmation construct of these two variables. The original framework is illustrated in Figure 1(a). The adjusted framework for evaluating the effect of publicness on satisfaction is shown in Figure 1(b).

Figure 1.

The establishment of the theoretical research framework. (a) The original framework of the ECT and (b) the adjusted framework for evaluating the effect of publicness on satisfaction. Note: Square in solid line for objective element, and square in dotted line for subjective element. OSAP: open space accessible to the public.

In this study, user satisfaction (the extent in which a person is pleased with the OSAP) is regarded as the outcome of two processes: visitation to OSAP and disconfirmation of beliefs. First, people visit OSAPs and interact with the environments in which they are located to perceive their performances. Afterwards, by comparing the perceived performances to their expectations, the extent to which they are pleased with the OSAP is formulated.

Publicness is the input as one of the factors that contributes to the attributes of OSAP, while user satisfaction is the outcome that may be affected by the degree of publicness. Assuming that visitation and individuals’ expectations are homogeneous, this study examines the relationship among publicness, performance, and satisfaction through a questionnaire, statistical analysis, and SEM. To what extent the assumption of homogeneous expectation is untenable will be tested based on the data from the questionnaires.

Selection of OSAP

The distribution of OSAP varies from district to district over HK’s three main parts: Hong Kong Island, Kowloon, and the New Territories. Figure 2 shows the basic information of GOS and POS in districts.

Figure 2.

The number of GOS, the number of POS, and per capita area of GOS in Hong Kong’s 18 administrative districts. GOS: government-owned open space; POS: privately owned open spaces. Source: GOS data are from the Leisure and Cultural Services Department (LCSD) website excluding sports centres and sports grounds (https://www.lcsd.gov.hk/en/anti-smoking/index.html); verified with the LCSD by email. POS data are from the Lands Department (https://www.landsd.gov.hk/en/legco/gic.htm). Population data are from the Census and Statistics Department (https://www.bycensus2016.gov.hk/en/bc-mt.html). The area per capita of POS is not shown since POS was absent in some districts.

There are more than 1500 GOSs in HK known variously as parks, gardens, playgrounds, sitting-out areas, and rest gardens. North District in the New Territories contains the largest number at 136 followed by Central and Western District on Hong Kong Island with 114. The overall area/capital is only 1.4 m²/capita. Central and Western, Southern and Wan Chai Districts are the only districts that boast GOS resources of above 2.0 m²/capita. Seventy-seven POSs were built, with Central and Western District having the most.

As for private land in HK, it is different from land that is permanently and completely owned by the private sector or individuals in other First World countries. Under HK’s land tenure system (https://www.legco.gov.hk/research-publications/english/essentials-1617ise07-land-tenure-system-in-hong-kong.htm), all land, except that which comprises St John’s Cathedral, is leasehold in that the government can grant leases to users with terms of 75, 99, or, in limited cases, 999 years. Development controls for private land are based on statutory planning, the conditions of which are superimposed on contractual land control through government lease interests. The state sells these to individuals as commodities (Lai et al., 2007).

In this study, the authors selected the Central and Western District (the District) as the area for further study. It is a traditional part of HK’s core urban area not only because it contains the central business district, but it is also a place with some long-established residential areas. The number of POSs in the District is the highest among all of the districts, while it also has more GOSs and a higher location quotient of GOS to population than the other districts. The data on the 15 POSs in the District are presented in Table 1.

Table 1.

Basic information on the 15 POSs in Central and Western District.

Development name	Type of project	Area of POS (m²)	Development name	Type of project	Area of POS (m²)
Hong Kong Station Development	Commercial	9600	Hollywood Terrace	Residential	1400
Cheung Kong Centre^a	Commercial	5200	The Sail at Victoria	Residential	1377
The Belcher’s	Residential	3518	The Nova^a	Residential	1200
Pacific Place Phases I and II^a	Commercial	3000	Cosco Tower and Grand Millennium Plaza^a	Commercial	800
The University of Hong Kong	Institutional	2150	The Centrium	Commercial	700
Bank of China Tower	Commercial	1720	Island Crest^a	Residential	700
The Center^a	Commercial	1500	La Maison Du Nord	Residential	359
The Peak Galleria	Commercial	1500	Average		2315

POS: privately owned open spaces.

Source: List of POS from the Lands Department (https://www.landsd.gov.hk/en/legco/gic.htm).

^aPOS selected for onsite questionnaire survey.

Six GOSs and six POSs were chosen for data collection and their geographic distribution is shown in Figure 3.

Figure 3.

The geographic distribution of the selected OSAPs in the Central and Western District. Note: The base map is sourced from Google Earth and edited by the authors. GOS: government-owned open space; POS: privately owned open spaces.

The selection of a POS to conduct the questionnaire was mainly based on the number of users. The authors visited all 15 POSs (Table 1) and stayed in each POS for around 20 minutes to observe user activity. Those POSs, with fewer than 10 visitors who either stayed for a period of time or passed by at the time of the observation, were excluded. In addition, in terms of scale and type of development, the authors excluded three extreme POSs including two with maximum and minimum areas and one located in an institutional (University) project site. That left six POSs as suitable for an onsite questionnaire survey.

The selected GOSs were typical of those with average objective performances in terms of area, green coverage rate, and facility for this GOS category. GOSs of sitting-out areas and rest gardens were too small to attract enough people to stay there for extended periods, so they were excluded from the site selection list. Each type of park, garden, and playground was ranked by area, green cover rate, and facility, while those in the middle clusters were selected. For example, nine parks were located in the District. Parks ranked fourth to sixth in terms of area and green coverage rate and boasting multiple facilities were selected for further study. The equal GOS and POS numbers guaranteed that the sampling sizes of these two types of OSAP were similar for comparison.

Satisfaction questionnaire design

To investigate the public’s satisfaction with OSAP, the authors designed a questionnaire to fathom how users liked the OSAP they visited according to four dimensions of performance: facility, environment, amenity/convenience, and safety. The design of the questionnaire was modelled after the ‘Quality of Public Open Space Tool’ proposed by the University of Western Australia (Lange et al., 2004). Indicators were developed to assess the dominant qualities of previous studies including those related to the attributes of space, nature, culture, history, noise, facilities, etc. (Chow et al., 2016; Van Herzele and Wiedemann, 2003).

Under each performance dimension, the authors asked questions about users’ perceived satisfaction of a few attributes. The attributes under the dimension of ‘Activity facility’ included diversity of active activities (e.g. sports, walking, etc.), diversity of passive activities (e.g. sitting, charting, etc.), frequency of visitation, and duration of stay. Attributes under the dimension of ‘Environment’ contained greenery, landscaping, air quality, quietness, and cleanliness. ‘Amenity’ included entry, seats, shading, and toilet. Most attributes are rated on a scale of 1–5. If the observed quality of attributes better matches the subjective aspiration, then the users were more satisfied with the OSAP.

Onsite OSAP survey

A questionnaire survey was conducted onsite in each selected OSAP. Only users at the selected places could respond to the questionnaire. Their participation was entirely voluntary and they could choose to stop or withdraw at any time during the survey. During the survey, the investigators attempted to ask as many visitors as possible to capture the general demographic distribution of users. The respondents covered a wide range of demographic groups regarding gender, age, education, and income. In total, 339 questionnaires were administered by field survey from February to March 2019, 155 of which were from GOSs and the rest from POSs. Except for some half-done responses, the effective return rate for the questionnaires was around 98%. Information on the questionnaire survey is listed in Table 2.

Table 2.

The information on the successfully administered questionnaires from selected OSAP.

Type of OSAP	Landowner	Sub-type of OSAP	No. of questionnaires	Venue (selected OSAP)
GOS	Government (1)	Park (1 – Park)	52	Belcher Bay Park
		Garden (1 – Garden)	43	Kwong Hon Terrace Garden
		Garden (1 – Garden)	43	Caine Lane Garden
		Playground (1 – Playground)	60	Sai On Lane Children’s Playground
				Fat Hing Street Children’s Playground
				Mui Fong Street Children’s Playground
			Subtotal	155
POS	Private sector (0)	Commercial large POS (0 – Commercial_Large)	71	Cheung Kong Center
		Commercial large POS (0 – Commercial_Large)	71	Pacific Place Phases I and II
		Commercial small POS (0 – Commercial_Small)	66	The Center
		Commercial small POS (0 – Commercial_Small)	66	Cosco Tower and Grand Millennium Plaza
		Residential small POS (0 – Residential_Small)	47	The Nova^a
		Residential small POS (0 – Residential_Small)	47	Island Crest^a
			Subtotal	184
			Total	339
Demographic distribution (%) of the respondents in total:
Gender: Male (45%); Female (55%)
Age: <15 years (3%); 15–24 (12%); 25–34 (26%); 35–44 (23%); 45–54 (15%); 55–64 (13%); >65 (8%)
Education: Primary (8%); Secondary (32%); Sub-degree (11%); Undergraduate (37%); Postgraduate (11%)
Income (HK$) : <10,000 (45%); 10,000–19,999 (25%); 20,000–29,999 (15%); >30,000 (15%)

GOS: government-owned open space; OSAP: open space accessible to the public; POS: privately owned open spaces.

^aRedevelopment project was jointly developed by the Urban Renewal Authority and a private developer.

Photos taken of the six sub-types of the visited venues are shown in Figure 4.

Figure 4.

Photos of six sub-types of OSAP visited. (a) 1 – Park, (b) 1 – Garden, (c) 1 – Playground, (d) 0 – Commercial_Large, (e) 0 – Commercial_Small, and (f) 0 – Residential_Small. Source: Authors.

Statistical method and SEM

The statistical methods of analysis of variance (ANOVA) and correlation were applied to preliminary data analysis using SPSS 25. To compare if the satisfaction with GOSs was different from POSs, ANOVA analysis was conducted. Correlation analysis was then used to test how the perceived performance of OSAP related to user satisfaction. The statistical analysis helped identify the most influential factors to significantly affect user satisfaction either through ownership or the other attributes of OSAP.

The identified factors became the inputted variables for the SEM. SEM is a form of causal modelling that includes a collection of statistical techniques allowing a set of relations between independent variables and dependent variables to be examined, which provides integrative function to guide exploratory and confirmatory research (Bagozzi and Yi, 2012; Ullman, 2006). By involving a variable of land ownership (1 – public; 0 – private) in the causal model, SEM can figure out how the ownership of OSAP, through paths and weights, influence user satisfaction in using OSAP. Latent variables are designed to represent the perceived satisfaction of users in four dimensions of satisfaction: activity, environmental, amenity, and safety, which combine to form the overall satisfaction with OSAP. The model was adjusted by adding/excluding variables or inserting/removing paths until the model’s fitness became qualified.

Statistical results and findings

This section explains the ANOVA results (‘The variance of satisfaction by types of OSAP’ subsection), the correlation between attributes and satisfaction (‘Performance-based exploratory factors of user satisfaction’ subsection), and SEM of user satisfaction (‘SEM of user satisfaction’ subsection). Before analysing the effect of OSAP performance, whether individuals’ expectations to OSAP are homogeneous was tested. Focusing on the rated overall satisfaction on the OSAPs (one question in the questionnaire), no significant differences were observed among demographic groups with different age, gender, education, and income. Thus, expectation could to some extent be regarded as a constant variable in this case.

The variance of satisfaction by types of OSAP

Information on user satisfaction is collected by asking the following question: ‘Are you satisfied with this open space?’. Apart from overall satisfaction, user satisfaction with each dimension (activity facility/environment/amenity/safety) was also asked. Satisfaction degree is rated on a scale of 1 (very unsatisfied) to 5 (very satisfied). The results of ANOVA by ownership are shown in Table 3.

Table 3.

The results of the ANOVA of user satisfaction by land ownership.

Satisfaction	Land ownership	Mean	F	Sig.	Satisfaction	Land ownership	Mean	F	Sig.
Activity facility	1 – Public	3.81	7.566	.006*	Safety	1 – Public	4.42	.002	.969
Activity facility	0 – Private	3.52			Safety	0 – Private	4.42
Environment	1 – Public	3.98	.967	.326	Overall	1 – Public	4.07	3.650	.057
Environment	0 – Private	3.88			Overall	0 – Private	3.88
Amenity	1 – Public	3.91	12.261	.001*
Amenity	0 – Private	3.53

ANOVA: analysis of variance.

*P < 0.05.

As seen in Table 3, GOS users showed significantly higher satisfaction with the dimensions of activity facility and amenity than in POS. Better performance of GOS was also observed for environmental and overall satisfaction, but they were not statistically significant. Satisfaction with safety was almost the same in groups of GOS and POS and was excluded in later comparisons.

However, different types of OSAP under the same type of land ownership (private or government) might be perceived differently. For example, people may show different levels of satisfaction with government-owned parks and playgrounds. To see if all types of GOS were superior to POS in terms of user satisfaction, a comparative analysis was performed via an ANOVA plus post hoc test of Scheffe (Table 4).

Table 4.

The results of the ANOVA of user satisfaction by type of OSAP.

Satisfaction	Type of OSAP	Subset for α = 0.05			Satisfaction	Type of OSAP	Subset for α = 0.05
Satisfaction	Type of OSAP	1	2	3	Satisfaction	Type of OSAP	1	2	3
Activity facility	0 – Commercial_Small	3.38			Amenity	0 – Commercial_Small	3.41
	1 – Playground	3.40				1 – Playground	3.42
	0 – Residential_Small	3.55				0 – Residential_Small	3.57
	0 – Commercial_Large	3.62	3.62			0 – Commercial_Large	3.61
	1 – Garden	3.93	3.93			1 – Garden	4.02	4.02
	1 – Park		4.17			1 – Park		4.38
	Sig.	0.10	0.10			Sig.	0.05	0.57
Environment	1 – Playground	3.50			Overall	1 – Playground	3.63
	0 – Residential_Small	3.70	3.70			0 – Commercial_Small	3.79	3.79
	0 – Commercial_Small	3.83	3.83	3.83		0 – Residential_Small	3.85	3.85
	0 – Commercial_Large		4.06	4.06		0 – Commercial_Large	4.00	4.00	4.00
	1 – Garden			4.28		1 – Garden		4.19	4.19
	1 – Park			4.29		1 – Park			4.48
	Sig.	0.53	0.45	0.17		Sig.	0.42	0.32	0.13

ANOVA: analysis of variance; OSAP: open space accessible to the public.

Table 4 suggests that the homogeneous sets clustered by satisfaction were not consistent with land ownership. The least satisfied subsets of OSAP included both types of POS (e.g. 0 – Commercial_Small and 0 – Residential_Small) and GOS (e.g. 1 – Playground), in every dimension of satisfaction. Apart from publicness, how perceptions of the attributes of OSAP affect user satisfaction are then analysed.

Performance-based exploratory factors of user satisfaction

Under each dimension, the perceived performance of OSAP was represented by a few observable attributes. The relationship between satisfaction and the observed attributes was analysed (see Table 5).

Table 5.

Correlations between perceived performance and user satisfaction.

Satisfaction Perceived performance	Activity facility	Environment	Amenity	Safety	Overall
Activity facility performance
Active_activity	0.15*	0.18*	0.29*	0.03	0.25*
Passive_activity	0.19*	0.18*	0.15*	0.10	0.15*
Frequency	0.09	0.04	0.13	0.09	0.06
Duration	0.24*	0.26*	0.31*	0.13	0.29*
Environmental performance
Greenery	0.41*	0.50*	0.51*	0.22*	0.50*
Landscaping	0.39*	0.53*	0.47*	0.27*	0.54*
Air_quality	0.37*	0.54*	0.43*	0.27*	0.52*
Quietness	0.28*	0.52*	0.36*	0.28*	0.49*
Cleanliness	0.28*	0.59*	0.40*	0.41*	0.44*
Amenity performance
Entry	0.32*	0.47*	0.41*	0.26*	0.47*
Seats	0.19*	0.30*	0.44*	0.22*	0.37*
Shading	0.29*	0.38*	0.52*	0.17*	0.43*
Toilet	0.19*	0.34*	0.45*	0.13	0.37*

*P < 0.05.

Note: Figures in bold font indicate the observable attributes related to the corresponding dimension of perceived performance in the questionnaire. No question of attribute is attached to the dimension of ‘Safety’.

User satisfaction with each dimension was found positively (>0) correlated to the perceived performance of OSAP. Overall (the last column in Table 5), the most influential attributes to satisfaction were landscaping (0.54), greenery (0.50), and air quality (0.52).

Different correlations were found between each attribute observed and the four dimensions of satisfaction. Except for the ‘Activity facility’ dimension, in which the designated attributes were not highly correlated (<0.25), the attributes in ‘Environment’ and ‘Amenity’ were representative of their corresponding dimensions (0.41–0.59). The inconsistency between ‘Activity facility’ and related attributes indicated that perceived satisfaction of activity facility was less represented by human behaviour in using OSAP (i.e. what they do in such spaces), but more by perceptions of the surrounding environment (i.e. what it is like when one uses the OSAP). These provide some insights on how to adjust the measurement model in SEM.

The results suggested that the correlations with statistical significance were in the range of 0.15–0.59, which indicated varied effects of performance attributes, as well as the complexity of explaining the causal factors of satisfaction. In addition, whether the difference in performance was related more to publicness or other factors was questionable. Consequently, SEM was applied.

SEM of user satisfaction

The latent variables of the sub-dimensional satisfaction with OSAP are measured by the observed variables of perceptions to OSAP attributes (listed in Table 5). For example, the satisfaction with ‘Environment’ was reflected by the latent variable of ‘perceived_environment’, measured by peoples’ ratings to the attributes of greenery, landscaping, etc. Same principle is applied to the dimension of ‘Amenity’. Since ‘Activity facility’ is not effectively represented by the assigned variables, satisfaction associated with ‘perceived facility’ was regarded as a variable related to ‘perceived environment’ instead of an independent latent variable. Similarly, the lack of measurement for ‘Safety’ generates the observed variable of ‘perceived safety’.

The paths are analysed from ‘publicness’ to the sub-dimension of satisfaction, and from the sub-dimension to overall satisfaction. After the discrepancy approach of the maximum likelihood estimates was run, the Chi-square of the model was found to be 374.1 with degrees of freedom at 60 and at probability level at 0.00. The goodness-of-fit index equalled 0.86, which indicated a high fitness of the model.

The critical ratios for most of the regression weights were significant at the 0.05 level (P < 0.05) except for the ‘land ownership → Perceived safety satisfaction’ and ‘Land Ownership → Perceived activity satisfaction’ paths. Table 6 lists the resulting estimates. Figure 5 illustrates the diagram of the SEM (with SPSS Amos 25).

Table 6.

Regression weights and the significance of path coefficients in SEM.

Dependent variable	Independent variable	Estimate	CR	P	Standardized estimate (E)
Perceived environment	Land ownership	−0.01	−0.11	0.92	−0.01
Perceived amenity	Land ownership	0.64	5.01	***	0.32
Perceived safety	Land ownership	0.00	−0.04	0.97	0.00
Perceived facility	Land ownership	0.30	3.20	**	0.15
Perceived facility	Perceived environment	0.59	8.91	***	0.53
Air quality	Perceived environment	1.00			0.75
Landscaping	Perceived environment	0.92	12.39	***	0.75
Greenery	Perceived environment	0.68	10.89	***	0.65
Cleanliness	Perceived environment	0.71	9.36	***	0.56
Quietness	Perceived environment	0.68	8.91	***	0.53
Shading	Perceived amenity	1.00			0.75
Seats	Perceived amenity	0.73	8.25	***	0.61
Toilets	Perceived amenity	0.77	7.52	***	0.53
Entry	Perceived amenity	0.49	6.81	***	0.47
Satisfaction overall	Perceived environment	0.47	8.28	***	0.52
Satisfaction overall	Perceived amenity	0.25	5.87	***	0.31
Satisfaction overall	Perceived facility	0.15	3.67	***	0.19
Satisfaction overall	Perceived safety	0.16	4.22	***	0.17

CR: critical ratio; SEM: structural equation modelling.

**P < 0.01.

***P < 0.001.

Figure 5.

The results of the SEM. Note: The observed variables are signified with rectangles and latent variables are with ellipses. The number over path is the factor loading of the independent variable, which was calculated as the Standardised Estimate. e1–e14 in circles are the residues over the corresponding variables, measurement or equations, showing the unexplained components of the relationship between independent and dependent variables.

Overall satisfaction was related mostly to perceived environment (E = 0.51) followed by perceived amenity (E = 0.31) and perceived activity facility (E = 0.19). Air quality and landscaping were the most influential variables for the perceived environment with coefficients of 0.75. The perceived amenity was affected by shaded areas (E = 0.75) and seats (E = 0.65).

The effects of land ownership on overall satisfaction were significant through the paths to amenity (E =0.32) and activity facility (E = 0.15) performances. The accumulative effect of land ownership on satisfaction was calculated by adding up the structural regression weights in the two paths, which resulted in 0.13 (calculation: 0.32 × 0.31 + 0.15 × 0.19).

Overall, the effect of land ownership was marginal. It was not strong enough to overcome the effect of other attributes such as air quality, landscaping, provision of shaded area, etc. People would most likely be satisfied with the amenities and facilities of GOS over those of POS, which suggests there is room to improve POS performance.

Discussion

Determinants of user satisfaction with OSAP

OSAP ‘Environment’ (landscaping, greenery, air quality, cleanliness, and quietness) was of vital importance to user satisfaction, as it mostly affects how much people are pleased with a place. Unlike previous studies, which emphasized the active use of OSAP (Giles-Corti et al., 2005; King et al., 2015), this study found that OSAP is not a place where many people expect to perform physical exercises except on areas serving children ages 5–12 and the elderly.

‘Amenity’, with the items of shading, seats, and toilet facilities, influences the extent people perceive the convenience and overall satisfaction with OSAP. Sitting and relaxing are the most popular activities for people who utilize OSAP, which indicates the importance of ‘Environment’ in performance evaluations.

This phenomenon is understandable in the context of HK, where people live densely and under pressure. OSAP provides space for people to slow down and relax. The findings are consistent with another comparative study between the ‘concrete jungle’ of HK and the ‘garden city’ of Singapore, which realized the public’s desire for the healing and therapeutic functions of OSAP in HK were much higher (Xue et al., 2017).

The optimization of microclimate conditions in OSAP in terms of air quality, landscaping, shading, ventilation, thermal comfort, etc. tends to be an effective strategy for improving user satisfaction with OSAP, especially in compact cities.

Pros and cons of POS versus GOS

The effect of publicness on user satisfaction mainly affects ‘Amenity’ followed by ‘Activity facility’, where GOS surpasses POS. As the results suggested, facilities for active activities were absent, as seats and shaded areas were insufficient and toilet facilities were inconvenient for POS users. The function of POS was found highly monotonous rather than diverse like GOS.

Another finding was the limited effect of POS in improving the urban climate due to its smaller area and confined ecological value. This was in line with the argument that privatizing open space may cause excessive development density and more congested urban areas (Tang and Wong, 2008). Regarding the restricted amenities and activities, only those who lived or worked in specified geographical catchment areas of POS were more likely to use and benefit from them. This has been considered a typology, in which privatized spaces are not intended for the public, but for select groups and sometimes with time limits (Mantey and Kepkowicz, 2018).

Nevertheless, POS does achieve the purpose of balancing urban green space distribution, facilitating capital accumulation, saving money, and adapting to changing political, economic, and social demands. The combination of GOS and POS reduces the overall average distance to the nearest public spaces for both workers and residents throughout HK (Yoon and Srinivasan, 2015).

From a fiscal perspective, the opportunity costs of providing OSAP in a core urban area such as Central were extremely high. In addition, people who worked or lived nearby were under more pressure, hence accessible POS helped relieve their stress and improve their productivity. These factors were in line with the research findings, which suggested that POSs were beneficial for developers, residents, and municipalities in terms of cost and notions of exclusivity (Rigolon and Nemeth, 2018). Under HK planning system, rising private responsibility does not mean the replacement of public governance in terms of public expectation, but it implies a variety of ‘publicness’.

Similar to HK, the privatization of OSAP reshaped the notion of ‘publicness’ in Europe and other parts of Asia. The commonly recognized binary of ‘private’ and ‘public’ was redefined through the creation of a hybrid landscape (Athanassiou, 2017; De Magalhaes and Trigo, 2017a), which collaboratively formed urban spaces in both the public sphere and arena of capital accumulation (Nachmany and Hananel, 2019). The new form of OSAP required a different kind of public management and policy, in which judiciously designed accountability mechanisms and clear decisions are reached by all key stakeholders (De Magalhaes and Trigo, 2017b). Better understanding of space use and approaches for improving and integrating various kinds of OSAP with differentiated property regimes, typologies, functions, etc. remains to be further discovered.

Implications for planning policy

Contemporary OSAPs comprise a set of heterogeneous spaces – each with its distinct degree of publicness, conditions of use, tolerance, and services (Boydell and Searle, 2014; Chiodelli and Moroni, 2014). In HK, GOSs such as parks and gardens serve multiple functions that include, but are not limited to, physical activity, recreation, image-building, environmental improvement, social interaction, community cohesion, etc. GOSs of playgrounds mainly serve children and are disadvantaged in terms of ‘Environment’ and other activities such as sports and jogging. The inadequacies of key design elements such as seats, pavilions, and greenery were identified in HK’s GOS (Lo and Jim, 2012).

For POS, both the function of activity and the scope of service have been weakened, as indicated by the inferior performance of ‘Activity facility’ and ‘Amenity’. The integration of GOS and POS to serve the diverse groups of people throughout the city is worthy of attention. POSs are more like market-oriented spaces, excluding children and the elderly while allowing highly stressed workers in core urban areas to relax.

Changing lifestyles, political and economic environments, and demands for improving the quality of life tend to be the forces behind the modification of space utilization. Youths go to gyms for physical exercise and to parks to play the Pokémon Go video game (Graells-Garrido et al., 2017). Through information technology, virtual communications through the app of Meetup and digital platforms of social and health clubs help people find opportunities to join others in outdoor activities in OSAP (Huang and Franck, 2018). GOS provided by traditional public sector for civic recreation and leisure is no longer sufficient to meet the diverse needs, as the sociology of OSAP is constantly changing.

Updating the planning and management guidelines for OSAP will improve the performance of existing spaces and raise user satisfaction. In HK, the policies mainly refer to the Planning Standards and Guidelines, published by the Planning Department to advise on GOS provision, and the Public Open Space in Private Developments Design and Management Guidelines, published by the Development Bureau to advise on POS arrangements in private developments. To promote the public welfare and social wellbeing, the government should comprehensively improve the quantity, quality, and distribution of both GOSs and POSs around the city.

Limitations and further study

The findings in this study were derived based on user surveys of selected OSAPs in HK. The representativeness of the survey data, even though it was justified in terms of site selection and the investigative process, should be interpreted with caution. As mentioned in the ‘Selection of OSAP’ section, POSs with few visitors during the preliminary observation were excluded, which means the overall POS accessibility in the city was problematic and the attractiveness of the selected POSs was above average. A deeper understanding of the overall attractiveness of POSs in HK would be helpful for elaborating on the results. Differences in satisfaction with the selected POSs and GOSs were discovered, which may be related to the implementations of the different policies for POS and GOS. Future studies should try to explain the connections between performance and the two different mechanisms of planning, design, and management, for HK and other cities with increasing POS.

Conclusion

Based on a questionnaire survey conducted in HK and the results of SEM, this paper demonstrates the causal relationships between the publicness of OSAP, their perceived performance, and user satisfaction. Even though POS and GOS are generally open to the public 24 hours a day, satisfaction with POS was lower than that with GOS. Private land ownership negatively affected the ‘Activity facility’ and ‘Amenity’ of a place. Nevertheless, the effects of publicness were not as significant as those associated with OSAP quality and perceived performance. The attributes of air quality, landscaping, provision of shaded areas, etc. helped determine peoples’ perceptions under any form of land ownership. Both POSs and GOSs have their unique (dis)advantages in shaping urban ecosystem and improving peoples’ quality of life, while the quality attributes are the key to user satisfaction. In urban development, the privatization of public space is an emerging trend in many countries, which results in adjusted policies for the planning, design, and management of OSAP. To improve user satisfaction with urban space, one should pay more attention to its hierarchical functions, the structured distribution to differentiated users, changing behaviours in space use, the rights of underprivileged groups, and the comprehensive integration of POS and GOS to meet the diverse public demands.

Footnotes

Acknowledgements

The authors thank the student helpers and Research Assistant Mr Ronald Yu in the data collection process and the respondents for participating in the survey. Appreciations are extended to Dr Jenny Cheung and Mr Chohong Choi for proofreading drafts of this paper.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was funded and supported by grants from the Research Grants Council of the Hong Kong SAR, China (Ref. No. UGC/FDS25/E02/16 and UGC/IDS25/16).

Daniel CW Ho is Professor and Associate Dean of the Faculty of Design and Environment, Technological and Higher Education Institute of Hong Kong (THEi), and Honorary Professor in Department of Real Estate and Construction, and Department of Urban Planning and Design, Faculty of Architecture, The University of Hong Kong. His research interests are in the fields of planning and development, facilities management, and building performance assessment. His recent publications focused on assessment of performance of property and FM agencies, management effects on building performance, and FM service outsourcing.

Lawrence WC Lai is a Professor in Planning, Law, and Economics with the Department of Real Estate and Construction, as well as the Ronald Coase Centre for Property Rights, The University of Hong Kong. He has a social science degree and is a Registered Professional Planner. His research interests focus on property rights, neo-institutional economics, and planning theory. He extends his research and publications on military heritage site conservation to illustrate various planning, property rights, as well as economics principles.

Anqi Wang is working as a senior research assistant in the Technological and Higher Education Institute of Hong Kong (THEi). She graduated with MSc in Sustainable Urban Development and PhD in Town Planning, both from The Hong Kong Polytechnic University. Her current research interests are in land use planning, urban green space, and urban modelling, mainly in the context of Hong Kong. She is also interested in conducting comparative study of land use planning between Hong Kong and other international cities.

ORCID iD

Anqi Wang

References

Athanassiou

(2017) The hybrid landscape of public space in Thessaloniki in the context of crisis. Landscape Research 42: 782–794.

Bagozzi

(2012) Specification, evaluation, and interpretation of structural equation models. Journal of the Academy of Marketing Science 40: 8–34.

Boydell

Searle

(2014) Understanding property rights in the contemporary urban commons. Urban Policy and Research 32: 323–340.

Chiodelli

Moroni

(2014) Typology of spaces and topology of toleration: City, pluralism, ownership. Journal of Urban Affairs 36: 167–181.

Chow

McKenzie

Sit

(2016) Public parks in Hong Kong: Characteristics of physical activity areas and their users. International Journal of Environmental Research and Public Health 13(7): 639.

De Magalhaes

Trigo

(2017a) ‘ Clubification’ of urban public spaces? The withdrawal or the re-definition of the role of local government in the management of public spaces. Journal of Urban Design 22: 738–756.

De Magalhaes

Trigo

(2017b) Contracting out publicness: The private management of the urban public realm and its implications. Progress in Planning 115: 1–28.

Giles-Corti

Broomhall

Knuiman

, et al. (2005) Increasing walking – How important is distance to, attractiveness, and size of public open space? American Journal of Preventive Medicine 28: 169–176.

Gomes

(2019) The birth of public space privatization: How entrepreneurialism, convivial urbanism and stakeholder interactions made the Martim Moniz square, in Lisbon, ‘privatization-ready’. European Urban and Regional Studies 27(1): 86–100.

10.

Graells-Garrido

Ferres

Caro

, et al. (2017) The effect of Pokemon go on the pulse of the city: A natural experiment. EPJ Data Science 6: 19.

11.

Huang

Franck

(2018) Let’s meet at Citicorp: Can privately owned public spaces be inclusive? Journal of Urban Design 23: 499–517.

12.

Karacor

(2016) Public vs. private: The evaluation of different space types in terms of publicness dimension. European Journal of Sustainable Development 5: 51–58.

13.

King

Litt

Hale

, et al. (2015) ‘The park a tree built’: Evaluating how a park development project impacted where people play. Urban Forestry & Urban Greening 14: 293–299.

14.

Lai

LWC

Lorne

(2019) Sustainable urban renewal and built heritage conservation in a global real estate revolution. Sustainability 11: 850.

15.

Lai

LWC

Yung

RYM

, et al. (2007) The private supply of and public demand for planning: Compliance with planning conditions in the absence of direct statutory enforcement measures. Planning Practice and Research 22: 535–557.

16.

Lange

Giles-Corti

Broomhall

(2004) Quality of Public Open Space Tool (POST): Observers’ Manual. Perth: School of Population Health, the University of Western Australia.

17.

AYH

Jim

(2012) Citizen attitude and expectation towards greenspace provision in compact urban milieu. Land Use Policy 29: 577–586.

18.

McEachan

RRC

Yang

Roberts

, et al. (2018) Availability, use of, and satisfaction with green space, and children’s mental wellbeing at age 4 years in a multicultural, deprived, urban area: Results from the born in Bradford cohort study. Lancet Planetary Health 2: e244–e254.

19.

Mantey

Kepkowicz

(2018) Types of public spaces: The polish contribution to the discussion of suburban public space. The Professional Geographer 70: 633–654.

20.

(2005) From public to private: Five concepts of park management and their consequences. The George Wright Forum 22: 12–20.

21.

Nachmany

Hananel

(2019) A tale of two neighborhoods: Toward a new typology of land rights. Land Use Policy 80: 233–245.

22.

Nemeth

(2009) Defining a public: The management of privately owned public space. Urban Studies 46: 2463–2490.

23.

Nemeth

Schmidt

(2011) The privatization of public space: Modeling and measuring publicness. Environment and Planning B: Planning and Design 38: 5–23.

24.

Oliver

(1980) A cognitive model of the antecedents and consequences of satisfaction decisions. Journal of Marketing Research 17: 460.

25.

Ostoic

van den Bosch

CCK

Vuletic

, et al. (2017) Citizens’ perception of and satisfaction with urban forests and green space: Results from selected southeast European cities. Urban Forestry & Urban Greening 23: 93–103.

26.

Poister

Thomas

(2011) The effect of expectations and expectancy confirmation/disconfirmation on motorists’ satisfaction with state highways. Journal of Public Administration Research and Theory 21: 601–617.

27.

Rey Gozalo

Barrigon Morillas

Montes Gonzalez

, et al. (2018) Relationships among satisfaction, noise perception, and use of urban green spaces. The Science of the Total Environment 624: 438–450.

28.

Rigolon

Nemeth

(2018) Privately owned parks in new urbanist communities: A study of environmental privilege, equity, and inclusion. Journal of Urban Affairs 40: 543–559.

29.

Tang

Wong

(2008) A longitudinal study of open space zoning and development in Hong Kong. Landscape and Urban Planning 87: 258–268.

30.

Ullman

(2006) Structural equation modeling: Reviewing the basics and moving forward. Journal of Personality Assessment 87: 35–50.

31.

Van Herzele

Wiedemann

(2003) A monitoring tool for the provision of accessible and attractive urban green spaces. Landscape and Urban Planning 63: 109–126.

32.

Wan

Shen

(2015) Salient attributes of urban green spaces in high density cities: The case of Hong Kong. Habitat International 49: 92–99.

33.

Xue

Gou

Lau

(2017) The green open space development model and associated use behaviors in dense urban settings: Lessons from Hong Kong and Singapore. Urban Design International 22: 287–302.

34.

Yoon

Srinivasan

(2015) Are they well situated? Spatial analysis of privately owned public space, Manhattan, New York City. Urban Affairs Review 51: 358–380.