Different urban settings affect multi-dimensional tourist-resident interactions

Abstract

City centre areas of many tourism destinations are increasingly overcrowding but two tourism development trends are to some extent counterbalancing the crowding phenomenon. Firstly, the decentralisation of tourists aiming to avoid large crowds and product standardisation and, therefore, visiting sites away from the city centre, including suburban areas. Secondly, the rise of ‘new urban tourism’ which means that tourists seek to experience authentic city life in local neighbourhoods. However, systematic examination and comparison of the frequency, intensity and quality of tourist-resident interactions in different urban settings is lacking. This paper does so for tourist-resident interactions in Central (as city centre area), Sha Tin (as suburban area) and Mong Kok (as new urban tourism area) in Hong Kong. The results from the tourist perspective are mostly in line with expectations, only co-presence appeared not to take place relatively less in Sha Tin and focussed interaction not relatively more in Mong Kok. From the perspective of residents, as expected, Sha Tin stands out as a place with less co-presence and less focussed interaction with tourists, and a higher quality of interaction. However, the expectation that Mong Kok as a new urban tourism area would involve both more co-presence and focussed interaction and a lower quality of interaction is not substantiated through our study. Overall, tourists experience a higher quality of interactions than residents, but the asymmetry is smaller in the suburban setting.

摘要

许多旅游目的地的市中心地区日益拥挤, 但有两种旅游发展趋势在一定程度上缓解了拥挤现象。首先, 游客分散, 以避免人群和产品标准化, 因此, 参观地点远离市中心, 包括郊区。其次, “新城市旅游”的兴起意味着游客寻求在当地社区体验真实的城市生活。然而, 缺乏对不同城市环境下游客与居民互动的频率、强度和质量的系统考察和比较。本文以香港中环(作为市中心)、沙田(作为郊区)和旺角(作为城市新旅游区)为研究对象。从游客的角度来看, 结果与预期基本一致, 只是在沙田出现的旅游者居民共处现象相对较少, 在旺角出现的集中性的互动现象相对较少。从居民的角度来看, 正如预期的那样, 沙田地区居民与游客的共处互动的较少, 较少集中性的互动, 但互动的质量较高。然而, 我们的研究并没有证实作为一个新的城市旅游区, 旺角将涉及更多的共同存在和集中的互动和较低的互动质量。总体而言, 游客体验到的互动质量比居民高, 但在郊区环境中, 这种不对称性较小。

Keywords:

• Tourist-resident interaction
• social contact
• urban setting
• city centre
• decentralisation
• new urban tourism
• Hong Kong

关键词:

• 旅游者与居民的互动；社会接触
• 城市环境
• 城市中心;去中心化；新城市旅游
• 香港

Introduction

The large influx of tourists from mainland China – with 51 million arrivals in 2018, constituting about 78 percent of the total number of tourist arrivals (HKTB Research, 2019), has brought tremendous change for Hong Kong. This market has stimulated the urban economy but, at the same time, many shopping malls, historical highlights as well as other facilities and services of tourist interest have become increasingly overcrowded – with detrimental effects on tourist-resident interactions. The city centre of Hong Kong with the most popular tourism attractions has the highest concentration of tourists, from mainland China in particular (Su et al., 2020). Generally speaking, it is the centre area of many tourism destinations across the globe that increasingly suffers from overcrowding due to excessive and concentrated growth in tourist arrivals.

Two tourism development trends are to some extent counterbalancing this (over)crowding of city centres in tourism destinations. One trend is the decentralisation of tourists aiming to avoid city centre area because of its large crowds and induced product standardisation. Tourist activity areas therefore expand beyond the city centre, including towards more peripheral sites – as indicated by Russo (2002). The second and related trend is the rise of so-called ‘new urban tourism’, meaning that tourists increasingly search for the true identity of cities and seek authentic experiences occurring in mundane places like cafes, markets and streets. These new urban tourism areas often include local, residential neighbourhoods located on the edge of or close to city centres (e.g. Füller & Michel, 2014; Maitland & Newman, 2004).

Both tourism development trends are supported and promoted through urban planning and marketing strategies in many tourism destinations across the globe. In general, less busy areas beyond the city centre and with potential tourist appeal are further developed and raised awareness of with the aim to redistribute and disperse tourists over the entire city and release some pressure from the city centre. In so doing, suburban and new urban tourism areas are often marketed as local tourist destinations not only offering different facilities and services but also different experiences and interactions with residents.

The three types of urban settings under scrutiny – i.e. city centre area, suburban area and new urban tourism area – may attract particular tourist segments and provide platforms for interactions with residents. Previous studies have analysed tourist-resident interactions in the urban, rural and urban-rural fringe setting (e.g. Loi & Pearce, 2015; Kastenholz et al., 2013; Zhang & Kwong, 2017) but, so far, little attention has been paid to systematically examining and comparing the frequency, intensity and quality of tourist-resident interactions in different settings within the urban area. Several studies suggest that new urban tourism areas involve more and better interactions for tourists, but not for residents (e.g. Maitland, 2010). The city centre area is generally understood as highly frequented by both tourists and residents, resulting in intensive but often also troublesome interactions (e.g. Kotus et al., 2015). Suburban areas often offer opportunities for tourists to encounter residents providing services for tourists, leading to relatively less interactions (e.g. Zhang et al., 2006). As such, tourist-resident interactions in cities are not homogenous in nature and the different urban settings should be taken into account to acknowledge their contextuality and, in so doing, develop a better understanding of the heterogeneity of the interactive processes taking place.

For the analysis of tourist-resident interaction, taking a bilateral approach is important because it enables the further investigation of asymmetry between tourists and residents in terms of the experienced quality of interaction and the types of interaction involved (Su et al., 2016). This asymmetry is well-known in tourism studies (e.g. Kwong & Li, 2020), but rare studies have systematically compared it and, so far, not yet by exploring whether and the extent to which this asymmetry differs between different urban settings.

This paper studies interactions between Mainland Chinese tourists and Hong Kong residents in a city centre area, a suburban area and a new urban tourism area. It contributes to the field of tourism studies in two ways. The first and main contribution is making a systematic examination and comparison of the frequency, intensity and quality of tourist-resident interactions in different urban settings. In so doing, the second contribution is applying a bilateral approach on interactions – by taking both the tourist and resident perspective into account as well as by investigating the asymmetry in the experienced quality of interaction as from the perspective of both tourists and residents in different urban settings.

Theoretical framework

For studies on intergroup relations and interactions in general, the contact theory proposed by Allport et al. (1954) is highly influential. According to Goffman (1967), social contact consists of two types, based on the intensity level of the interaction occurring among two or more individuals – i.e. co-presence and focussed interaction. Co-presence occurs when two or more individuals are simply aware of one another’s presence, involving ‘fleeting’ interaction. Focussed interaction occurs when people gather and cooperate to sustain a single focus of attention – e.g. in situations such as a conversation and a transaction (Goffman, 1967). Islam and Hewstone (1993) add the quality of the interaction as important dimension of social contact, referring to the conditions under which it occurs, and the type of relations involved. This quality is assessed based on equal or unequal status, involuntary or voluntary, cooperative or competitive, superficial or intimate, and pleasant or unpleasant. Thus, the main dimensions constituting social contact are co-presence, focussed interaction and (experienced) quality of interaction.

For an examination of the multi-dimensional tourist-resident interactions, several studies have applied social contact theory (e.g. Woosnam & Aleshinloye, 2013; Yu & Lee, 2014). Different types of interactions have been taken into account, making use of the distinction between co-presence and focussed interaction as well as by looking at the quality of the interaction. The types of tourist-resident interactions have mostly been analysed through measurements of frequencies of co-presence and focussed interaction, together with the specific activities involved, and an assessment of the quality of the interaction (e.g. Huang & Hsu, 2010). Reisinger and Dimanche (2009), for instance, found that co-presence and focussed interaction most frequently occur at tourist attractions and in transaction situations. With a focus on the quality of the interaction, Fan et al. (2017) indicate that an equal, cooperative, intense and friendly interaction creates a favourable relationship between tourists and residents. Moreover, Eusébio et al. (2012) pinpoint that the frequency and intensity of tourist-resident interaction depends on the kind of place or destination where the contact occurs.

Urban tourism destinations usually provide a variety of geographical settings with tourism appeal. Cohen (1972) demonstrated that an urban setting could set itself apart from others by the tourism attractions embedded. These attractions often have a distinctive spatial distribution in cities which is not to say that the areas involved are only visited by tourists and not by residents. Moreover, the same areas usually have a variety of additional functions that are of interest to and are being used by both groups – including shopping, housing and transportation functions. As such, tourists and residents increasingly share but also compete for the same facilities and areas in cities (Pearce, 2001). Overall, urban settings with tourism appeal provide many opportunities and meeting places (e.g. shopping malls, parks, streets and restaurants) for tourist-resident interactions and experiences (Wearing & Foley, 2017). Reisinger and Turner (2003) add that the differences between urban settings may affect the actual tourist-resident interaction taking place.

Many urban destinations try to spread tourists over different urban settings through dispersal strategies, and to do so may brand urban settings beyond the city centre as offering different types of interactions. The city centre is usually the most important area for tourism activities (Łapko, 2014), 2014). Two tourism development trends have surfaced in response to the high concentration of tourism activities in the city centre: the rise of tourist areas outside the centre as an alternative to the centre (e.g. Russo, 2002), and the rise of new urban tourism in which tourists leave the city centre to enjoy the authentic life of a local neighbourhood (e.g. Füller & Michel, 2014). These trends give rise to different urban settings for interactions between tourists and residents and therefore to differences in the intensity and quality of the interactions.

In popular tourist cities, the city centre area is often a well-developed area with a concentration of tourist attractions, combined with many supporting facilities and infrastructures of a high quality. It attracts and concentrates a large number of tourists to visit, experience and perform activities on the traditional ‘beaten track’ (Kotus et al., 2015). However, with a large and increasing number of tourists flocking in, the city centre may become congested and the goods and services standardised – hampering social interactions and experiences. As a consequence, tourist activities may spatially expand beyond the city centre and into more peripheral and suburban areas in the city – or even further, into surrounding villages and towns, as Popp (2012) indicated. Despite the fact that the suburban setting often provides less traditional tourism attractions, the area can still be of tourist interest.

Tourists visit the suburban area for several reasons, including easy access, less travel time as well as less overall costs. Moreover, they usually spend less time in this area compared to tourists visiting the city centre area (Zhang et al., 2006). Taking into account that the longer tourists stay in an area, the more likely interactions with residents will occur (Prentice et al., 1994), tourists in suburban areas may have relatively less co-presence and focussed interaction with residents than tourists in the city centre area. When visiting the suburban area, tourists usually do not have high social or cultural expectations with respect to interactions with residents (Su & Wall, 2010). For that reason, they may be more easily satisfied in terms of the experienced quality of the interaction – potentially resulting in tourists in the suburban area giving a higher quality assessment than tourists in the city centre area.

Due to lacking studies on tourist-resident interactions in the specific suburban setting, we will build our assumptions here by making use of available studies on interactions in the village setting. From the perspective of residents, most residents who had intensive interactions with tourists in villages were involved in the tourism industry (Kastenholz et al., 2013; Su et al., 2016). These focussed interactions tend to be rare or short as well as formal and business oriented. However, most tourism-related industries such as retail, nightclubs and restaurants are concentrated in central areas (Lau & McKercher, 2006). A relatively larger amount and variety of focussed interaction may therefore take place in the city centre. As such, and despite building on studies from a village setting, residents in suburban areas are assumed to have less focussed interaction than residents in the city centre. Taking into account the higher density of tourists in the city centre (Simpson, 1999), residents in suburban are also assumed to have less co-presence with tourists. Regarding the experienced quality of interaction, the assessment is assumed to be higher in the suburban area than in the central area. The reason for this is that the central area is popular among mass tourists and when their presence and activities lead to the experience of overcrowding, it may annoy and antagonise residents (Zhang et al., 2006). However, suburban areas usually have lower use levels and are less crowded, which can be assumed to result in a higher quality of interactions for residents.

Another urban setting for tourist-resident interactions is often located on the edge of or close to the historic city centre (Maitland & Newman, 2004). These so-called ‘new urban tourism’ areas usually are connected to former working-class and post-industrial transitional neighbourhoods (Füller & Michel, 2014). They provide tourists with traditional and intricate street patterns, various types and styles of buildings, a wide range of small retail and gastronomy, and access to local people living their everyday life (Maitland, 2010). Being part of mundane and authentic city life is the most important visiting motive of new urban tourists, seeking interactions with residents in their daily life spaces – as a substitute for designated tourist spaces (Dirksmeier & Helbrecht, 2015).

As opposed to mass tourists in the city centre who tend to stay in their ‘tourist bubble’, new urban tourists show more interest in interacting with residents in local, residential neighbourhoods (Edensor, 2001). Their more active quest for local experiences in new urban tourism areas can be assumed to result in more interactions with residents (Maitland & Newman, 2014). This may not only be the case for focussed interaction – as Luo et al. (2015) argued – but also for co-presence – since being ‘in the observation’ mode can be an important part of being a new urban tourist (Wildish & Spierings, 2019). Moreover, both types of interaction are likely to occur more when tourists stay in an area longer, as Prentice et al. (1994) argued, and many new urban tourists do tend to spend a significant amount time in local neighbourhoods by making use of Airbnb in seeking access to local life. Regarding the experienced quality of interaction, more focussed interaction with residents is positively related with tourists experiencing a higher trip satisfaction in general (Pizam et al., 2000). For new urban tourists in particular, having both relatively more focussed interactions with residents and attaching more importance to these interactions for their trip satisfaction, we assume that they give higher assessments for the quality of interaction with residents than tourists in city centre areas.

From the perspective of residents, a relatively lower density of tourists in the new urban tourism area not necessarily implies less interactions with tourists. Residents could actually have more interactions with tourists in these settings than in the city centre area because new urban tourists more actively search for focussed interaction and also more actively observe locals, when in co-presence. To do so, while simultaneously trying to ‘blend in’ and behave as a local (McCabe, 2005), these tourists appropriate many public spaces when strolling neighbourhood streets and visiting local markets, bars, shops and restaurants (Pappalepore et al., 2010). It is through this use of residents’ mundane and daily life spaces in residential neighbourhoods that the latter may perceive and experience more co-presence and focussed interaction with new urban tourists. Regarding the experienced quality of interaction, residents in general have better attitudes towards alternative tourists than towards mass tourists (Gursoy et al., 2010) but when it comes to their own local neighbourhood, new urban tourists may still be experienced as intrusive (Gu & Ryan, 2008). New urban tourism may, and increasingly does, raise concerns with respect to neighbourhood commercialisation, increasing living costs, tourist crowds and tensions between residential and tourist activities (Dirksmeier & Helbrecht, 2015). Especially because the interactions with tourists take place in mundane and everyday spaces, we will test the assumption that residents experience a lower quality of interaction with tourists in a new urban tourism area than in city centre area.

When tourists visit cities, they tend to be more explorative and aware of experiences in the city and interactions with residents whereas residents’ experiences and interactions are more often based on repetition and routines (Sutton, 1967). More specifically, tourists’ interactions with people and places are an important constituent of the urban experience. Combined with their holiday mood it is likely, although inconclusive, that tourists are more positive about the interactions than residents (Dirksmeier & Helbrecht, 2015). Residents often seem to differ in their experiences of and responses to interactions with tourists whereas many studies show that responses and attitudes usually shift from positive to negative due to tourist-resident interactions (e.g. Ap & Crompton, 1993; Quinn, 2007). Despite the different findings in the above, we follow Kwong and Li (2020), for instance, in assuming that the quality assessment of the interaction by tourists and residents is asymmetrical. We will further investigate this asymmetry by testing it in the different urban settings under scrutiny.

Drawing on the discussions regarding the quality of interaction in the above, both tourists and residents in the suburban area are assumed to experience a higher quality of the interaction, compared to the city centre area. Taking into account the lower tourist density in the suburban area, which usually makes tourism less intrusive, we expect that the asymmetry in quality assessment between tourists and residents is smaller than in the city centre area. For the new urban tourism area, and compared to the city centre area, tourists are assumed to experience a higher quality of interaction whereas residents are assumed to experience a lower quality. Together, this results in the assumption that the asymmetry in quality assessment between tourists and residents in the new urban tourism area is larger than in the city centre area.

Altogether, the following hypotheses will be tested in this paper: (1) Tourists in the suburban area have less co-presence and focussed interaction but experience a higher quality of interaction than in the city centre area; (2) Residents in the suburban area have less co-presence and focussed interaction but experience a higher quality of interaction than in the city centre; (3) Tourists in the new urban tourism area have more co-presence and focussed interaction and also experience a higher quality of interaction than in the city centre area; (4) Residents in the new urban tourism area have more co-presence and focussed interaction but experience a lower quality of interaction than in the city centre area; (5) The asymmetry between tourists and residents in terms of the quality of interaction is smaller in the suburban area and larger in the new urban tourism area than in the city centre area.

Research design

Urban settings

Hong Kong consists of Hong Kong Island, the Kowloon Peninsula, the New Territories (which are connected to mainland China) and over 200 outlying islands, officially divided into 18 districts. Among them, Kowloon and Hong Kong Island are main urban areas within Hong Kong, while the New Territories encompasses the suburban and rural areas (He, 2020). The results presented in this article are based on a case study in three different urban areas in Hong Kong – i.e. Central on Hong Kong Island, Sha Tin in New Territories and Mong Kok in Kowloon. These were selected to represent the typical urban settings of ‘city centre area’, ‘suburban area’ and ‘new urban tourism area’ respectively (Figures 1 and 2).

Figure 1. Location of the three urban settings studied in Hong Kong. Source: Authors.

Figure 2. Urban settings of tourist-resident interactions: Central, Sha Tin and Mong Kok. Source: Authors.

Central contains several clusters of commercial buildings and office buildings together with a variety of tourism-related attractions and facilities, including the Peak Tram, the Mid-Level Escalator and the Central Ferry with connections to the outlying islands and, for instance, Macau. Central is also the home to a considerable number of residents and many people commute there for work on a daily basis. Sha Tin is one of the new towns in Hong Kong, comprising residential quarters, shopping centres, and a variety of educational, cultural, recreational and sports facilities. It is well-connected to the Shenzhen border and has become a hotspot for visitors from the mainland, including for cross-border parallel trading. Mong Kok is well-known and popular for its markets, including the Flower market and the Goldfish market, streets with many small stores, street food, historic heritage and popular culture.

Sampling and data collection

For this study, two survey questionnaires were developed to collect data – i.e. one for mainland Chinese tourists and another for local residents in Hong Kong. Through a stratified sampling approach, these tourists and residents were divided into three groups. This was done based on whether they visited or live in the three urban settings respectively. Survey respondents were selected randomly for each group.

An online survey was conducted with mainland Chinese tourists having visited Central, Mong Kok and/or Sha Tin before. The first question of the tourist questionnaire asked tourists to select one of these sites only for further research and each respondent could fill out one questionnaire only. The online survey was conducted in Mandarin by the wjx.cn survey company. This company has a database of 2.6 million people. Our survey was randomly sent in June 2017 to mainland Chinese living in a variety of mainland cities. The reliability of the online questionnaires received was confirmed by the IP-address of the respondents and the research sample was diverse in terms of gender, age, job status and origin. The acceptance rate of the online survey was about 70%. Altogether, 416 valid questionnaires were collected – i.e. 130 in Central, 121 in Sha Tin and 165 in Mong Kok.

An on-site survey was conducted (in Cantonese) with residents living in Central, Mong Kong or Sha Tin. Residents were approached for participation in the research at these particular sites through a random intercept approach. This resulted in an acceptance rate of 1 out of 7 on average. The fieldwork was undertaken in June 2017 on both weekdays and during the weekends. Altogether, 315 valid questionnaires were collected – i.e. 107 in Central, 96 in Sha Tin and 112 in Mong Kok.

Survey questionnaires

In addition to questions about residents’ and tourists’ demographic characteristics, the meeting places for tourist-resident interactions in the different urban settings, the survey questionnaires focussed on three related dimensions of tourist-resident interaction – i.e. co-presence, focussed interaction and quality of the interaction. For studying co-presence and focussed interaction, a measurement was developed that simultaneously considers the activities involved and their frequencies. The items measuring aspects of co-presence and focussed interaction were adopted from the study by Fan et al. (2017) and fine-tuned based on on-site investigations in June 2017. The items measuring quality aspects of interaction were adopted from studies by Huang and Hsu (2010) and Islam and Hewstone (1993). Tourist-resident interactions were analysed by using 7-point scales for 23 items. Together, they contained two subscales: 16 items for co-presence and focussed interaction and 7 items for the quality of the interaction.

For the frequency of co-presence and focussed interaction between tourists and residents, the questionnaire provided answer categories ranging from 1 – representing ‘never’ – to 7 – representing ‘very frequently’ – for activities involved. Thus, a higher score denotes a higher frequency of interaction. For the activities themselves a list of 16 items with typical interactional behaviours between tourists and residents (e.g. sitting next to each other, dining in the same restaurant, chatting casually and making friends) was devised. The quality of the interaction was assessed by asking tourists and residents about their subjective experiences when having interactions with each other (i.e. harmonious, friendly, interesting, equal, cooperative, close and profound). To do so, answer categories ranging from 1 – representing ‘strongly disagree’ – to 7 – representing ‘strongly agree’ – were provided. Thus, a higher score denotes a higher assessment of the quality of interaction.

Data analysis

Before analysing the data, missing values were computed through single imputation. Next, factor analysis was applied in order to explore the dimensions of tourist-resident interactions. The factor structure was examined through varimax-rotated factor analysis. Items with low loadings and cross-loading issues were eliminated. More specifically, the items with loadings lower than .4 and with loadings higher than .4 on more than one factor were deleted (Choo & Petrick, 2014). All items associated with tourist-resident interactions were extracted into three factors, accounting for 65.933% of the variance in the data. These three factors were in line with multi-dimensional tourist-resident interactions, i.e. co-presence, focussed interaction and quality of the interaction. Co-presence (factor 1) consisted of six items (e.g. sitting around), focussed interaction (factor 2) consisted of six items (e.g. chatting casually) and experienced quality of the interaction (factor 3) consisted of seven items (e.g. harmonious). Moreover, both the values of the Kaiser-Meyer-Olkin measure (.902) and of the Bartlett’s test of sphericity value (11394.164, p < .001) confirm that our data is suitable for factor analysis. In addition, the value of the Cronbach’s alpha (.829), as a measure of scale reliability, suggest that the factors extracted have relatively high internal consistency.

Demographic characteristics of tourists’ and residents’ sample were analysed in terms of age, gender, level of education and amount of monthly income. The meeting places for tourist-resident interactions per urban setting in Hong Kong, as indicated by the respondents through answering open questions, were presented in bars charts with frequencies. After that, differences between the city centre, suburban area, new urban tourism area with respect to tourist-resident interactions were examined by using the Kruskal-Wallis H test. This is a rank-based nonparametric test that can be used to determine whether there are any statistically significant differences between two or more independent groups. Afterwards, a post hoc test was run on all possible pairs of mean ranks to tell which specific groups were significantly different.

Possible differences between tourist-resident interactions in the three urban settings under scrutiny were analysed and compared in three steps. The first step consisted of analysing and comparing the interactions mainland Chinese tourists have with Hong Kong residents and the second step did the same for the interactions Hong Kong residents have with mainland Chinese tourists. For the third step, the asymmetry of tourist-resident interaction was tested for the different urban settings. Based on our hypotheses, formulated at the end of the theoretical framework, the Sha Tin as suburban area and Mong Kok as new urban tourism area were compared with Central as city centre area. As such, differences for the comparison between Sha Tin and Central as well as between Mong Kok and Central are presented but not for between Sha Tin and Mong Kok.

Interactions between mainland Chinese tourists and Hong Kong residents in Central, Sha Tin and Mong Kok

Demographic characteristics of sample

The demographic characteristics of the tourists’ and residents’ sample have been presented in Table 1. For the tourists, the gender division is largely similar for all three urban settings (39% males/61% females) whereas for the residents it does show some variation – with the largest difference between Sha Tin (45% males) and Mong Kok (61% males). Also, the average age of tourists is largely similar (roughly 32 years) for the three settings whereas for the residents it shows some variation – with the largest difference between Sha Tin (27 years) and Central (39 years). Most tourists (74% and above) are highly educated, with a bachelor or master degree. The two highest percentages can be found in Mong Kok (90%) and Sha Tin (93%). Most residents in Central and Sha Tin are highly educated (51% and 78% respectively) but mostly have a low to medium education in Mong Kok (56%). Most tourists have a medium monthly income, ranging between about 47% in Central to about 69% in Sha Tin. Most residents also have a medium monthly income but the range – between 34% in Central and 46% in Mong Kok – is below that of tourists. However, the resident sample shows higher percentages for high income in each of the urban settings – with the highest percentages in Central (29%) and Mong Kok (35%).

Table 1. Demographic characteristics of tourists and residents per urban setting.

Demographics		Central (% of respondents)		Sha Tin (% of respondents)		Mong Kok (% of respondents)
		Tourist	Resident	Tourist	Resident	Tourist	Resident
Gender	Male	39.2	49.5	38.8	44.8	39.4	60.7
	Female	60.8	50.5	61.2	55.2	60.6	39.3
Age	Average	32.27	38.9	32.12	27.34	30.92	34.87
Education	College or below	26.2	49.5	6.6	21.9	10.3	56.3
	Bachelor	38.5	45.8	62.8	63.5	60.0	37.5
	Master or above	35.4	4.7	30.6	14.6	29.7	6.3
Monthly income*	Low	37.7	37.4	18.2	47.9	35.8	17.9
	Medium	46.9	33.6	69.4	35.4	54.5	46.4
	High	15.4	29.0	12.4	16.7	9.7	35.7
Total		130	107	121	96	165	112

Note: Monthly income*: Tourist (Less than 8.000 RMB/8001-16000 RMB/More than 16.000 RMB), Resident (Below 10.000 HKD/10.000-20.000 HKD/More than 20.000 HKD).

Source: Authors.

Meeting places for tourist-resident interactions

The (semi-)public spaces where tourists and residents meet and interact are presented in Figure 3. From the tourist perspective, the main meeting places (>10%) are similar for all three urban settings – i.e. the shopping mall, the park, the street and the restaurant. To some extent, the shopping mall stands out in Sha Tin (48%), followed by Mong Kok (39%) and Central (33%). The highest scores for the park, street and restaurant can be found in Central (18%), in Mong Kok (16%) and Sha Tin (16%) but the differences between the urban settings are limited. Reflecting the geographical context of each setting, some specific meeting places can also be noticed – i.e. the harbour in Central, the station in Sha Tin and the market in Mong Kok.

Figure 3. Meeting places for tourist-resident interactions in Central, Sha Tin and Mong Kok. Source: Authors.

From the perspective of residents, there is a striking difference in terms of the diversity of meeting places in the urban settings. The largest differences can be found when comparing Mong Kok with Sha Tin, with Central in an intermediate position. In Mong Kok, the main meeting places (>10%) are the shopping mall (29%), the market (17%), the street (14%) and everywhere (17%). In Central, the main meeting places are the shopping mall (50%) and the street (14%). For Sha Tin, the shopping mall clearly stands out as the most important meeting place with a score of 77%. Once again, some specific meeting places reflecting the geographical context of each setting can be noticed – i.e. the pier in Central, the station but also the cultural centre, museum and university in Sha Tin, and the market in Mong Kok.

Altogether, the shopping mall is indicated as the main meeting place from both the tourist and resident perspective in each urban setting. At the same time, interesting differences can be found when comparing the two perspectives – e.g. the park being an important meeting place in Central according to tourists but not to residents and the same for the restaurant in Sha Tin and the market in Mong Kok.

Tourists’ interactions with residents

The results for mainland Chinese tourists’ interactions with Hong Kong residents in Central, Sha Tin and Mong Kok can be found in Table 2. Looking at the median scores (M) reveals that, overall, tourists had more co-presence (≥ 4) than focussed interaction with residents (≤ 4). The most frequent activities of co-presence include walking on roads and taking a bus or subway (M = 6). The most frequent activities of focussed interaction include chatting casually (M = 4) and the least frequent activities include being invited to the home by residents (M = 2). All median scores for quality are 5, indicating that overall the quality of interaction was assessed as good.

Table 2. Kruskal-Wallis test statistics for tourists’ interactions with residents in three urban settings (C-Central, ST-Sha Tin, MK-Mong Kok).

Interactions	Median	Mean Rank			K-W test statistics	Pairwise comparisons
		Central	Sha Tin	Mong Kok
Co-presence^a		147.38	250.36	225.95	51.734*	C-ST/C-MK
Sitting around	4	180.35	237.69	209.27	14.773*	C-ST
Dining in restaurants	5	173.43	232.12	218.81	17.542*	C-ST/C-MK
Walking on roads	6	153.91	235.40	231.78	41.519*	C-ST/C-MK
Wandering in the area	6	150.83	233.78	235.40	45.474*	C-ST/C-MK
Taking a bus or subway	6	157.39	243.54	223.07	38.141*	C-ST/C-MK
Shopping	5	149.37	251.72	223.40	51.342*	C-ST/C-MK
Focussed interaction^a		227.18	225.73	181.15	14.164*	C-MK
Chatting casually	4	205.44	230.13	195.05	6.237
Having photos taken by residents	4	200.74	236.53	194.06	9.728
Taking photos for residents	2	234.07	209.53	187.60	11.738*	C-MK
Bargaining	4	211.89	219.22	197.97	2.391
Making friends	2	233.64	217.71	181.94	15.186*	C-MK
Inviting to home	2	235.90	211.25	184.89	14.329*	C-MK
Quality^b		191.09	228.79	207.34	6.185*	C-ST
Harmonious	5	198.27	224.55	204.79	3.430
Friendly	5	197.89	217.33	210.38	1.788
Interesting	5	199.29	228.67	200.96	5.048
Equal	5	194.70	222.43	209.15	3.494
Cooperative	5	201.20	224.95	202.19	3.335
Close	5	194.66	226.05	206.53	4.522
Profound	5	197.49	224.49	205.45	3.495

Note:

^a Each item was asked on a 7-point Likert scale where 1 = ‘Never’ and 7 = ‘Daily’.

^b Each item was asked on a 7-point Likert scale where 1 = ‘Strongly disagree’ and 7 = ‘Strongly agree’.

* Asymptotic significances (2-sided tests) are displayed. The difference is significant at the 0.05 level. Significance values have been adjusted by the Bonferroni correction for multiple tests.

Source: Authors.

When comparing the different urban settings, the Kruskal-Wallis test statistics reveal significant differences for co-presence for both pairs of settings, with a mean rank score of 147.38 for Central, 250.36 for Sha Tin and 225.95 for Mong Kok. These results are not consistent with the hypothesis that tourists in the suburban area have less co-presence than in the city centre area. In that respect, shopping stands out for Sha Tin – with a significant difference compared to Central – which may be performed relatively more by tourists in, for instance, the shopping mall (Figure 3b). The results are consistent with the hypothesis that tourists in new urban tourism area have more co-presence than in the city centre area. In that respect, wandering in the area stands out for Mong Kok – with a significant difference compared to Central – which may be performed relatively more by tourists when observing local life in, for instance, the shopping mall, park or streets (Figure 3c).

For focussed interaction, the Kruskal-Wallis test statistics reveal significant differences between Central and Mong Kok, with a mean rank score of 227.18 for Central, 225.73 for Sha Tin and 181.15 for Mong Kok. While the difference is not significant, it is consistent with the hypothesis that tourists in the suburban area have less focussed interaction than in the city centre area. In that respect, taking photos for residents stands out for Sha Tin – without a significant difference compared to Central – which is performed relatively less in, for instance, the shopping mall, park or street (Figure 3b). The results are not consistent with the hypothesis that tourists in the new urban tourism area have more focussed interaction than in the city centre area. In that respect, making friends stands out for Mong Kok – with a significant difference compared to Central – which may be performed relatively less when visiting, for instance, the shopping mall, park or street (Figure 3c).

For experienced quality, significant differences show but only between Central and Sha Tin, with a mean rank score of 191.09 of Central, 228.79 for Sha Tin and 207.34 for Mong Kok. This is consistent with the hypothesis that tourists in the suburban area experience a higher quality of the interaction than in the city centre area. In that respect, close stands out for Sha Tin – without a significant difference compared to Central – as a relatively higher quality of interactions with residents taking place in, for instance, the shopping mall, park or restaurant (Figure 3b). While the difference is not significant, the results are also consistent with the hypothesis that tourists in the new urban tourism area experience a relatively higher quality of the interaction than in the city centre. In that respect, friendly stands out for Mong Kok – without a significant difference compared to Central – as a relatively higher assessment of interactions taking place with residents in, for instance, the shopping mall, park or restaurant (Figure 3c).

Residents’ interactions with tourists

The results for Hong Kong residents’ interactions with mainland Chinese tourists’ interactions in Central, Sha Tin and Mong Kok can be found in Table 3. Looking at the median scores (M) reveals that, overall, residents had more co-presence (≥ 5) than focussed interaction (≤ 3) with tourists. The most frequent activities of co-presence include walking on roads and shopping (M = 6). The most frequent activities of focussed interaction include chatting casually (M = 3) and the least frequent activities include inviting tourists to the home (M = 1). All median scores quality is 4, indicating that overall the quality of interaction was assessed as sufficient.

Table 3. Kruskal-Wallis test statistics for residents’ interactions with tourists in three urban settings (C-Central, ST-Sha Tin, MK-Mong Kok).

Interactions	Median	Mean Rank			K-W test statistics	Pairwise comparisons
		Central	Sha Tin	Mong Kok
Co-presence^a		169.86	136.82	164.82	7.633*	C-ST
Sitting around	5	180.93	149.39	143.48	11.001*	C-MK/C-ST
Dining in restaurants	5	158.83	140.81	171.94	6.393
Walking on roads	6	165.59	141.88	164.57	4.640
Wandering in the area	6	172.50	131.29	167.04	12.790*	C-ST
Taking a bus or subway	5	150.15	146.52	175.33	6.637
Shopping	6	169.77	145.42	157.54	3.793
Focussed interaction^a		163.92	162.51	148.48	1.910
Chatting casually	3	164.07	152.71	156.73	.850
Taking photos for tourists	3	170.77	159.24	144.74	4.650
Having photos taken by tourists	2	154.72	172.91	148.35	4.263
Bargaining	1	159.63	161.54	153.41	.548
Making friends	1	136.06	186.95	154.14	18.965*	C-ST
Inviting to home	1	139.82	174.57	161.17	10.949*	C-ST
Quality^b		140.50	176.44	158.92	7.898*	C-ST
Harmonious	4	143.98	175.58	156.33	6.508*	C-ST
Friendly	4	145.36	171.72	158.32	4.514
Interesting	4	143.58	174.38	157.74	6.080*	C-ST
Equal	4	134.80	183.46	158.34	15.168*	C-ST
Cooperative	4	143.23	192.79	142.29	21.016*	C-ST
Close	4	145.90	170.86	158.54	3.982
Profound	4	136.26	175.96	163.38	10.895*	C-ST

Note:

^a Each item was asked on a 7-point Likert scale where 1 = ‘Never’ and 7 = ‘Daily’.

^b Each item was asked on a 7-point Likert scale where 1 = ‘Strongly disagree’ and 7 = ‘Strongly agree’.

* Asymptotic significances (2-sided tests) are displayed. The difference is significant at the 0.05 level. Significance values have been adjusted by the Bonferroni correction for multiple tests.

Source: Authors.

When comparing the different urban settings pairwise, the Kruskal-Wallis test statistics reveal significant differences for co-presence but only between Central and Sha Tin, with a mean rank score of 169.89 for Central, 136.82 for Sha Tin and 164.82 for Mong Kok. This is consistent with the hypothesis that residents in the suburban area have less co-presence than in the city centre area. In that respect, wandering in the area stands out for Sha Tin – with a significant difference compared to Central – which may be performed relatively less by residents in, for instance, the shopping mall (Figure 3e). While the difference is not significant, the results are not consistent with the hypothesis that residents in the new urban tourism area have more co-presence than in the city centre area. In that respect, sitting around stands out for Mong Kok – with a significant difference compared to Central – which may be performed relatively less when residents are living their local life and spending time in, for instance, the shopping mall and street (Figure 3f).

For focussed interaction, the Kruskal-Wallis test statistics do not reveal significant differences between urban settings, with a mean rank score of 163.92 for Central, 162.51 for Sha Tin and 148.48 for Mong Kok. While the difference is not significant, the results support the hypothesis that residents in the suburban area have less focussed interaction than in the city centre area. In that respect, inviting to the home stands out for Sha Tin – with a significant difference compared to Central – which may be performed relatively less by residents when encountering tourists in, for instance, the shopping mall (Figure 3e). While the difference is not significant, the results are not consistent with the hypothesis that residents in the new urban tourism area have more focussed interaction than in the city centre. In that respect, photo taking for tourists stands out for Mong Kok – without a significant difference compared to Central – which may be performed relatively less by residents in, for instance, the shopping mall (Figure 3f).

For experienced quality, significant differences show but only between Central and Sha Tin, with a mean rank score of 140.50 for Central, 176.44 for Sha Tin and 158.92 for Mong Kok. This is consistent with the hypothesis that residents in the suburban area experience a higher quality of the interaction than in the city centre area. In that respect, cooperative stands out for Sha Tin – with a significant difference compared to Central – as a relatively higher quality of interaction with tourists taking place in, for instance, the shopping mall (Figure 3e). While the difference is not significant, the results are not consistent with the hypothesis that residents in the new urban tourism area experience a lower quality of interaction than in the city centre. In that respect, profound stands out for Mong Kok – without a significant difference compared to Central – as a relatively higher quality of interactions taking place with tourists in, for instance, the shopping mall, market or street (Figure 3f).

The asymmetry in quality of tourist-resident interaction

The results for the asymmetry in the experienced quality of interactions between tourists and residents in the different urban settings can be found in Table 4. Overall, the Kruskal Wallis test statistics suggest that there are significant differences in the quality of interaction between Central, Sha Tin and Mong Kok. Looking at the mean rank scores reveals that tourists experienced a higher quality of interaction than residents in each of the urban settings. Moreover, the asymmetry in the quality of tourist-resident interactions appears to be the largest in Central (23.811) and the smallest in Sha Tin (18.279), with Mong Kok (22.086) taking a somewhat intermediate position. This is partly consistent with the hypothesis that the asymmetry between tourists and residents in terms of the quality of interaction is smaller in the suburban area and larger in the new urban tourism area than in the city centre area.

Table 4. Kruskal-Wallis test statistics for comparing quality of interactions in three urban settings. Source: Authors.

Quality of interaction	Central			Sha Tin			Mong Kok
	Tourist	Resident	K-W test statistics	Tourist	Resident	K-W test statistics	Tourist	Resident	K-W test statistics
	138.72	95.05	23.811*	125.23	88.54	18.279*	157.64	111.54	22.086*
Harmonious	141.19	92.04	32.299*	124.51	89.45	17.644*	156.84	112.72	20.995*
Friendly	139.69	93.86	27.831*	123.50	90.73	15.202*	157.74	111.39	23.138*
Interesting	139.62	93.95	27.595*	122.11	92.48	12.618*	153.23	118.03	13.378*
Equal	135.75	98.65	18.097*	116.93	99.01	4.623*	152.03	119.80	11.198*
Cooperative	140.25	93.19	29.013*	117.42	98.39	5.149*	162.05	105.04	35.038*
Close	135.87	98.50	18.361*	124.00	90.10	16.251*	153.92	117.02	14.608*
Profound	136.69	97.51	20.252*	122.58	91.89	13.488*	150.50	122.06	8.756*

Note: Each item was asked on a 7-point Likert scale where 1 = ‘Strongly disagree’ and 7 = ‘Strongly agree’.

* Asymptotic significances (2-sided tests) are displayed. The difference is significant at the 0.05 level. Significance values have been adjusted by the Bonferroni correction for multiple tests.

Conclusions and discussion

This article examined and compared the frequency, intensity and quality of interactions between Mainland Chinese tourists and Hong Kong residents in different urban settings. Our main contribution to the field of tourism studies is making a systematic comparison of tourist-resident interactions in three typical settings within the urban area – i.e. a city centre area (Central), a suburban area (Sha Tin) and a new urban tourism area (Mong Kok). To do so, we build on the work by Reisinger and Turner (2003) arguing that differences between urban settings may affect the interactions taking place. In line with the work by Fan et al. (2017) and Yu and Lee (2014), for instance, we examined three dimensions of tourist-resident interactions – i.e. co-presence, focussed interaction and quality of interaction. Building on the work by Sutton (1967) and Su et al. (2016), for instance, we also further investigate the asymmetry in the experienced quality of tourist-resident interaction by testing it in the three different urban settings for interactions.

The results from the tourist perspective show that, as expected, less focussed interaction occurs in Sha Tin as suburban setting and the quality of interaction is higher, compared to Central as city centre setting. However, contrary to expectations, it is not the case that relatively less co-presence occurs in Sha Tin. This may be explained by tourists spending more time in Sha Tin than assumed, resulting in more co-presence with residents. For instance, tourists could be spending quite some time in the suburban area while waiting for a transport connection back to the mainland and, in so doing, hang out in the park, stroll the streets, dine in the restaurants and shop in the mall. For Mong Kok, as expected, the results show that more co-presence occurs and that the quality of interaction is higher than in Central. However, contrary to expectations, focussed interaction does not occur more than in Central. This may be explained by new urban tourists prominently adopting an ‘observation mode’ – as Wildish and Spierings (2019) also indicated – while trying to ‘blend in’ and behave as a local with the aim of experiencing mundane and authentic city life.

The results from the resident perspective show that, as expected, less co-presence and less focussed interaction take place and the quality of interaction is higher in Sha Tin as suburban setting, compared to Central as city centre setting. For Mong Kok, contrary to expectations, the results show that more co-presence and focussed interaction than in Central do not occur and that the quality of interaction is not lower. A possible explanation for the latter is that new urban tourists are not (yet) experienced as intrusive by residents living in Mong Kok, compared to studies on the new urban tourism settings we based the hypothesis on – i.e. in Berlin and Beijing (Dirksmeier & Helbrecht, 2015; Gu & Ryan, 2008). A possible explanation for the deviating findings for co-presence and focussed interaction is that Mong Kok may actually be closer to providing a city centre setting than a new urban tourism setting due to its high and increasing popularity among tourists.

The results for the asymmetry in the quality of interaction show that, overall, tourists experience a higher quality of interaction than residents. Moreover, the asymmetry in quality between tourists and residents in the suburban area is, as expected, smaller than in the city centre area but, contrary to expectations, the asymmetry in the quality of interaction in the new urban tourism area is not larger than in the city centre area. Once again, a possible explanation for the latter is that Mong Kok residents do not experience new urban tourists as intrusive as we found in studies on local neighbourhoods in other cities. Besides, new urban tourism is oriented towards authentic experiences and local amenities in ordinary local neighbourhoods, which usually have indistinct boundaries (Füller & Michel, 2014; Pappalepore et al., 2010). On the basis of the characteristics of new urban tourism, previous studies (e.g. Matoga & Pawłowska, 2018) often connected new urban tourism with areas that are off the beaten track. However, in this study, we started from the point of view of tourists seeking authentic and local experiences in neighbourhoods. The different starting point may have led to different results in Mong Kok, because Mong Kok is a fairly well-established tourism area and is located in the central urban areas, while at the same time it can represent the true identity of Hong Kong.

An important starting point for this study was the combination of two tourism development trends which to some extent counterbalance the overcrowding of city centres. The first trend is the decentralisation of tourists towards areas away from the city centre, including towards more peripheral and suburban sites. The second trend involves new urban tourists visiting local neighbourhoods on the edge of or close to the city centre. Many tourism destinations, including Hong Kong, support and promote these tourism development trends with the aim to redistribute and disperse tourists over the entire city. This is often done by urban planners and marketeers through further developing and raising awareness of urban settings beyond the city centre – as local destinations offering different types of facilities, interactions and experiences.

In this context, the findings of our study provide practical insights for tourism planning and marketing. Overall, investing in and marketing relatively underdeveloped and unknown tourism facilities in areas beyond the city centre in order to spread tourist flows has potential for generating both different types and a higher quality of tourist-resident interactions in the city overall. With respect to the quality of the interaction, our study suggests that both from the tourists’ and the residents’ perspective, the experienced quality is higher in a suburban area as well as a new urban tourism area than in a city centre area. More specifically, the quality of interaction from the bilateral perspective in the new urban tourism area seems higher than in the city centre area but the asymmetry between tourists and residents regarding the experienced quality of the interaction seems relatively high. Compared to both the city centre and the new urban tourism area, the quality of the interaction from the bilateral perspective appears to be the highest and also the least asymmetric in the suburban setting. As such, suburban settings provide the most potential for improving tourist-resident interactions and experiences by further developing and marketing the areas and their facilities as local destinations – when aiming to release some pressure from the overcrowding city centre at the same time.

For further investigation of tourist-resident interactions in different settings within the urban area, we suggest the following. Firstly, more items (e.g. participating in festivals, taking local tours and asking for information) with respect to co-presence and focussed interaction combined with the local context of tourism destinations could be added to the survey questionnaire for a richer analysis of tourist-resident interactions. Secondly, the survey should be repeated every couple of years to investigate potential long-term changes in co-presence, focussed interaction and experienced quality of interaction in the urban settings – combined with a longitudinal analysis of the number of tourists visiting these settings. Thirdly, more personal characteristics of tourists and residents could be included in the survey – including attitudes towards each other, years of residency, purpose of visit, whether tourists are first-timer or repeaters, and origin – to provide richer explanations for differences between tourists’ and residents’ activities and experiences in different urban settings across the city.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Notes on contributors

Xing Su

Xing Su is a PhD candidate in Urban Geography at the Department of Human Geography and Spatial Planning of University of Utrecht. Her current research interests focus on the areas of tourism mobility, tourist-resident interactions, tourist spatio-temporal behaviours and application of big data in urban tourism research.

Bas Spierings

Bas Spierings is Assistant Professor in Urban Geography at the Department of Human Geography and Spatial Planning at Utrecht University in the Netherlands. His research focuses on the nexus between urban consumption and public space with specific attention for touristification, leisure, retail developments, (cross-border) shopping, walking mobilities and encounters with difference.

Pieter Hooimeijer

Pieter Hooimeijer is professor of Human Geography and Demography at the Urban Futures Centre of Utrecht University. His main research interest is the recursive relation between population change and the dynamics of housing and labour markets at a variety of spatial scales ranging from neighbourhoods to metropolitan areas. Current researches include the effects of neighbourhoods on social mobility and participation in society.