Exploring the crowding-satisfaction relationship of skiers: the role of social behavior and experiences

ABSTRACT

Previous research has shown that crowding represents an essential factor in driving overtourism and highlighted the negative effects of crowding. However, there is limited knowledge of how crowding affects skiers’ experiences and what variables drive the crowding-satisfaction relationship. Our findings show that individual skill levels and the behavior of other skiers moderate this relationship. These findings are supported by a set of interviews, which focus on skiers’ crowding experiences. Our contribution provides a refined understanding of crowding in the ski context through highlighting the importance of tourist-to-tourist encounters to better understand the effects of crowding and suggests visitor management techniques.

KEYWORDS:

• Overtourism
• skiing
• crowding
• tourist-to-tourist encounter
• carrying capacity
• satisfaction

Introduction

Nowadays, overtourism is mostly associated with cities (Koens et al., 2018), such as Venice (Seraphin et al., 2018), Dubrovnik (Panayiotopoulos & Pisano, 2019), or Budapest (Smith et al., 2019). However, overtourism related issues, such as crowding, have also reached rural destinations and affect tourists’ satisfaction as well as destinations’ long-term development (Eckert et al., 2019; Jørgensen & McKercher, 2019; Lalicic, 2019; Zehrer & Raich, 2016). In the winter sport context, Steiger et al. (2020) highlighted the impact of variables such as crowding, experiences and other ski area-specific characteristics on customers’ demand. Since crowding is known to represent a phenomenon, which occurs in a defined geographical space or area during a specific period, we advocate exploring crowding within the concrete scenario of a ski area. While previous research has strongly focused on snow conditions and climate change (Steiger et al., 2019), only limited research explored experience-related conditions, such as crowding (Zehrer & Raich, 2016). Nevertheless, industry reports confirm the high importance of crowding as a potential factor limiting economic success of ski areas and skiers’ experiences (Vanat, 2019).

Thus, the study at hand attempts to close this gap and uses the theoretical lens of perceived crowding to explore visitors’ satisfaction and experiences in the ski industry. Since previous research highlighted the importance of social interactions for tourist experiences (White & White, 2008) and also underlined that “tourist-to-tourist interactions differ from interpersonal interactions context in terms of mode, content and depth (Lin et al., 2019, p. 153), we shed more light on the social encounters that drive crowding situations. For this reason, we use a mixed method design and combine a quantitative and qualitative phase (Onwuegbuzie & Leech, 2005), which helps us to expand the interpretation of the findings. First, we conducted a quantitative study (N = 224) in a large winter sports area asking skiers about their ski experience using the constructs of perceived crowding, crowding location, individual skill levels, behavior of other skiers and satisfaction. Second, we focus on explaining the tourist-to-tourist behavior (Yagi & Pearce, 2007) in a qualitative study (N = 27). This phase is particularly important since previous research highlighted the role of social interactions for tourists’ experiences (White & White, 2008) and also literature reviews emphasized the relevance of social components for perceived crowding (Lee & Graefe, 2003; Neuts & Nijkamp, 2012).

In summary, the findings of this paper offer new insights into the relationships that drive crowding and impact satisfaction in the ski context (Zehrer & Raich, 2016). The contribution highlights the relevance of context-specific crowding variables such as individual skill level (Matzler et al., 2008) and social factors, which are related to the behavior of other skiers (Lin et al., 2019; Nicholls, 2010). The mixed method design helps to better understand the drivers of crowding and highlights the necessity to start managing tourist-to-tourist encounters. The paper concludes with the identification of managerial implications for destination managers and ski area operators and recommendations for future research.

Conceptual background

Overtourism and carrying capacity

In the last years, publications, edited books (Dodds & Butler, 2019a; Innerhofer et al., 2018; Milano et al., 2019) and literature reviews (Capocchi et al., 2019; Dodds & Butler, 2019b; Perkumienė & Pranskūnienė, 2019) on topics related to overtourism were burgeoning. In general, overtourism is defined as “the impact of tourism on a destination, or parts thereof, that excessively influences perceived quality of life of citizens and/or quality of visitors experiences in a negative way” (UNWTO, 2018, p. 4). Peeters et al. (2018) offer a more detailed definition concerning capacity thresholds:

Overtourism describes the situation in which the impact of tourism, at certain times and in certain locations, exceeds physical, ecological, social, economic, psychological and/or political capacity threshold. (Peeters et al., 2018, p. 15)

Previously, overtourism was mostly identified as an urban phenomenon (Koens et al., 2018), but rural areas are also beginning to struggle with rising visitor numbers. In the rural setting, overtourism and overcrowding may be even harder to manage due to limited capacity thresholds and missing management and governance structures (Bichler, 2019; Butler, 2019). In addition, it represents a social issue, which is strongly related to social carrying capacity. This concept is crucial since it focuses on the quality of the experience that visitors will accept before seeking alternative destinations and on the level of tolerance, showed by the host population to the presence of tourists (Graefe et al., 1984; López-Bonilla & López-Bonilla, 2008). Therefore, social carrying capacity consists of two aspects: “a capacity issue,” e.g. how many tourists can be accommodated on the slopes before negative impacts occur, expressed in numerical terms, and a “perceived capacity issue,” e.g. how much ski tourism is acceptable before there is a decline in visitor satisfaction (Coccossis & Mexa, 2016; Saveriades, 2000). These two dimensions of social carrying capacity represent the underpinning idea of perceived crowding (Neuts & Nijkamp, 2012; Zehrer & Raich, 2016).

Perceived crowding

Crowding can be defined as “a motivational state aroused through the interaction of spatial, social, and personal factors, and directed toward the alleviation of perceived spatial restriction” (Stokols, 1972, p. 275). Later, Shelby and Heberlein (1984) added that perceived crowding results from the individual assessment of density levels within a specific physical environment. Crowding issues become important when the usage of social resources within a given area exceeds its norms (Lee & Graefe, 2003) and negatively impacts the visitors’ experiences (Shelby et al., 1989).

While crowding implies density, the two concepts are distinctive because the former is a necessary but not sufficient antecedent of crowding (Stokols, 1972). Thus, crowding can result from human or spatial density (Machleit et al., 2000) and arises if there is either a restrictive limit of space or a high number of people (Machleit et al., 1994). While the amount of space available determines spatial crowding, the perception of human crowding is subjective and dependent on the individual’s evaluation of density (Machleit et al., 2000). The number of non-human elements in an environment and their relationships defines the extent of spatial crowding perceived by individuals, while the human dimension of crowding concerns the number of individuals as well as the rate and extent of social interaction among individuals in a given environmental setting (Zehrer & Raich, 2016).

Neuts and Nijkamp (2012) differentiate between three dimensions of crowding determinants, which influence the level of perceived crowding in a specific situation. First, situational attributes, which are the number of tourists, availability of resource and environmental quality as well as the design of spaces. Second, the on-site behavior of other tourists, which addresses the degree of similarity perceived between oneself and other tourists, and tourist interaction. Third, the personal characteristics that relate to socio-demographic variables, such as country of origin and length of stay, motivations and expectations of the individual, experiences and preferences.

In the theme park context, Grove and Fisk (1997) showed that almost 60% of respondents had been significantly affected, positively and negatively, by the presence of others while visiting an attraction. While protocol incidents refer to the violation of rules of conduct during an experience, such as waiting in line, taking photos or searching for lost children, sociability incidents relate to friendly and unfriendly incidents, such as rude behavior of others. Additionally, ambiance incidents refer to events such as crowding, encumberment or malevolence. Similar to theme parks, the available space for slopes in ski areas is limited by topography and specific laws and regulations (Amt der Tiroler Landesregierung, 2018). Therefore, we focus on human crowding and formulate the following research question:

RQ1: Which determinants drive crowding perceptions in the context of ski areas?

Perceived crowding and tourist-to-tourist interactions

Previous research has shown that the notion of crowding is closely related to users’ expectations (Ditton et al., 1983; Lee & Graefe, 2003) which are strongly influenced by the behavior and actions of other customers (Nicholls, 2010). This holds especially true for tourists from different cultures (Weiermair, 2000) with different demographics or diverse backgrounds (Fleishman et al., 2004). Watson (1988) claims that the perception of crowding relates to the nature of the visitor interaction within a particular service setting, including visitors’ attributes and expectations. Following Martin and Pranter’s (1989) study on customer-to-customer interactions, customers influence each other’s experience. Various factors, e.g. the appearance, verbal exchanges, behavior and the number of other consumers can affect service perceptions (Moore et al., 2005).

Fullerton and Punj (1993) identified a range of characteristics that affect inappropriate customer behaviors, including demographics (age, sex, education and economic status), psychological features (personality traits: e.g. need for affiliation, aggression, compliance or dominance), social/group influences and a consumers’ frame of mind. Social interactions between customers seem to be more critical when customers are in close physical proximity (e.g. urban areas) to each other or when verbal interaction is more likely, or when customers engaged in various activities (Martin & Pranter, 1989). Especially in the tourism context, where individuals from all over the world with many different backgrounds coincide, tourists may recognize inappropriate behaviors due to their heterogeneous needs and expectations (Cai et al., 2018; Torres & Orlowski, 2017). Consequently, interactions between tourists may result in an encounter that affects the tourist’s experience and satisfaction negatively (Johnson & Grier, 2013; Wu, 2007). Earlier studies also measured behavior patterns of other tourists (e.g. Beernaert & Desimpelaere, 2001; Meyer, 1999). So far, the concept of tourist-to-tourist interaction has received little attention in tourism research (Lin et al., 2019) and the assumption that these interactions have less negative impacts on perceived crowding in urban settings (see Neuts & Nijkamp, 2012) holds true for crowded natural settings. In the past, some studies have dealt with the impacts of tourist-to-tourist interaction on their experiences (Cai et al., 2018; Huang & Hsu, 2010) but neglected largely the context of perceived crowding. While Torres (2016) found evidence that group formation in guided tours does not take long, others reported on the importance of social interactions among backpackers for backpacking tourism (Murphy, 2001; Sørensen, 2003). In terms of group identity, Wei et al. (2017) prove the role of social interactions as mediators between customer-to-customer interaction and the experiences in a conference setting. Wu (2007) suggests that companies should actively manage their tourist-to-tourist interactions by appropriately grouping customers with similar characteristics and communication of a code of conduct before starting a trip. This discussion leads us to our second research question:

RQ2: How do skiers perceive social encounters and incompatible social behavior in crowded ski areas?

Perceived crowding and satisfaction in ski areas

The majority of previous studies on perceived crowding have concentrated on retail and shopping (Andereck & Becker, 1993; Li et al., 2009; Maeng et al., 2013; Mehta, 2013; Pons et al., 2014; Popp, 2012), on urban areas and cities (Arnberger & Mann, 2008; Neuts & Nijkamp, 2012), or events (Lee & Graefe, 2003). Some research has focused on crowding related to outdoor settings (e.g. Fleishman et al., 2004; Kainzinger et al., 2015; Luque-Gil et al., 2018; Manning et al., 2009; Moyle & Croy, 2007; Rathnayake, 2015; Vaske & Shelby, 2008), but so far only Zehrer and Raich (2016) examined crowding in ski areas. It is necessary to differentiate between these different settings since perceived crowding is known to depend on specific constraints in terms of spatial, social and personal factors (Shelby & Heberlein, 1984; Stokols, 1972). Additionally, Coccossis et al. (2001) highlighted the need to consider different types of tourist destinations when assessing capacity considerations. Since the ski industry is a rural, seasonal, spatial and temporal concentration (Bicknell & McManus, 2006; Vanat, 2019), perceived crowding plays a tremendous role, especially during peak seasons. Crowding often burdens fragile environments in alpine areas and results in impairments for tourists during a ski-day (Pegg et al., 2012). Thus, when the carrying capacities regarding natural, economic and social aspects in ski areas are reached, too many skiers not only result in tourist dissatisfaction, but can also lead to severe problems regarding the attractiveness (Jacobsen et al., 2019; Weber et al., 2017). Existing research on crowding in rural areas (Fonner & Berrens, 2014) underlines the need to have a closer look at crowding in ski areas as research has demonstrated that e.g. waiting times matter to skiers (Bielen & Demoulin, 2007; Hui & Tse, 1996; Unbehaun et al., 2008; Won & Hwang, 2009).

Furthermore, crowding issues not only reduce the perceived attractiveness of an area (Matzler et al., 2008) but can also lead to coping behavior (Zehrer & Raich, 2016), resulting in dangerous situations. For example, skiers start to increasingly ride off-piste, which not only leads to conflicts regarding wildlife but also puts themselves in dangerous situations, increasing the risk for injuries and avalanches increases. In contrast, sometimes, crowding effects are also interpreted as a positive sign, indicating that the ski area is worth visiting (Weber et al., 2017).

In order to determine whether a destination is worth a visit, previous studies have highlighted the importance of satisfaction, which is known to ensure tourists’ post-trip behavior (Kozak, 2001; Oh, 1999). Besides, satisfaction is also one of the most frequently studied variables related to retail and tourism crowding (Budruk et al., 2002; Eroglu et al., 2005; Machleit et al., 2000). The relationship between crowding and behavioral intentions has been widely investigated in the past, but it remains one of the major issues which affect enjoyment and satisfaction at tourism sites (Kim et al., 2016; Lee & Graefe, 2003; Rathnayake, 2015). Thus, also previous tourism studies in the area of crowding and skiing focused on satisfaction as a central concept (Matzler et al., 2008; Zehrer & Raich, 2016). The underlying notion of satisfaction in the recreational context is that pleasing, exciting or relaxing experiences result in satisfaction and boring or frustrating activities result in lower levels of satisfaction (Chhetri et al., 2004). Since also Getty and Thompson (1995) suggest that behavioral intentions represent a function of perceived overall quality, we define overall satisfaction as the likelihood to engage in certain behaviors, which includes revisit and recommend intention (Oliver, 2014). This is also supported by Hui et al. (2007) showing the compatibility of satisfaction, revisit and recommend intention to form a single construct (Hui et al., 2007).

Development of hypotheses

The literature review highlighted that overtourism always occurs within a spatially restricted setting (Peeters et al., 2018) and showed the special importance of crowding location for perceived crowding (Graefe et al., 1984; Neuts & Nijkamp, 2012). We showed that there exists a discussion on how crowding is perceived differently in varying situations and sites (Jacobsen et al., 2019; Peeters et al., 2018; Won & Hwang, 2009), Based on these findings, we develop the following hypothesis:

H1. The perception of crowding a) at the valley station, b) on the slopes, or c) at the gastronomic facilities differently influences the overall crowding perception.

Our discussion on satisfaction has shown that it is a central concept to explain tourists’ post-trip behavior (Kozak, 2001). Therefore, it has also been widely used in the general crowding literature (Budruk et al., 2002; Machleit et al., 2000), but also in previous studies on ski area crowding (Zehrer & Raich, 2016), highlighting the adverse effects of perceived crowding levels on satisfaction. In this context, we defined the following hypothesis:

H2. Perceived crowding is negatively related to satisfaction.

Other studies focused on extending the understanding of the relationship between perceived crowding and satisfaction by integrating additional variables. For example, the literature review showed the importance of tourist-to-tourist interactions (Lin et al., 2019; Nicholls, 2010; Torres, 2016). Therefore, we focus extensively on the effects of social interactions (Johnson & Grier, 2013; Lee & Graefe, 2003; Moore et al., 2005; Wu, 2007). In addition, the importance of skill levels has been emphasized in previous leisure and recreational literature (Creyer et al., 2003) and also the effects of skill levels on satisfaction have also been discussed in the skiing literature (Matzler et al., 2008; Won & Hwang, 2009; Zehrer & Raich, 2016). We derive the following hypotheses:

H3. The relationship between perceived crowding and satisfaction is moderated through the individual skill level.

H4. The relationship between perceived crowding and satisfaction is moderated through the behavior of other skiers.

A large body of literature has discussed the influence of personal characteristics on the crowding-satisfaction relationship (Fleishman et al., 2004; Jacobsen et al., 2019; Neuts & Nijkamp, 2012). In the ski area context, Zehrer and Raich (2016) and Matzler et al. (2008) paid special attention to the impact of age, gender and type of visit but did not analyze country of origin. However, in an urban context, Neuts and Nijkamp (2012) found evidence for differences among Asian and European tourists, while Johnson and Grier (2013) demonstrated that perceived cultural compatibility amongst customers influences consumers’ satisfaction in a service setting. Vanat (2019) also pointed out that international flows of ski visitors are concentrated within Europe. Based on these findings and taking into account the diverse results for country of origin, we formulated the following hypotheses:

H5. Demographics such as age, gender and country of origin correlate with the perceived crowding.

H6. Demographics such as age, gender and country of origin correlate with satisfaction.

To determine how all of these variables regulate perceived crowding and satisfaction, we performed a series of quantitative data analysis procedures. To expand the interpretation of these results, we used a mixed-method design and gathered new qualitative data to enhance the quantitative data with qualitative insights (Collins et al., 2006).

Study design and methods

This paper investigates the relationship between site-specific crowding, perceived crowding and satisfaction from a pragmatic mixed method perspective (Onwuegbuzie & Leech, 2005; Watson, 1997). Since crowding represents a social issue that occurs at different places, times and on different scales, we have used a triangulation mixed method design (Creswell & Plano Clark, 2011; Doyle et al., 2009) to obtain findings that are plausible, insightful and relevant. This approach builds on the combination of an objective and subjective phase for data analysis and then combines quantitative results with qualitative insights (Creswell & Plano Clark, 2011). The rational for choosing a mixed method design was to balance between breadth and depth by expanding the interpretation of the quantitative results with qualitative findings (Collins et al., 2006).

First, to analyze the quantitative data, we use a structural partial least square (PLS) approach (Hair, 2010; Hair et al., 2019). The PLS approach was chosen because it represents a “soft modeling approach” (Hair et al., 2012, p. 416) with several benefits: PLS is used for prediction-oriented research that aims to maximize the explained variance of dependent variables and can be used if less rigid theoretical backgrounds are available (Hair et al., 2012; Henseler et al., 2014). This approach does not require normally-distributed data and is well suited for smaller sample sizes (Henseler et al., 2014). In combination with the mixed method design (Doyle et al., 2009), the PLS approach is more “straightforward” and also offers the opportunity to explore the data in greater detail.

Second, to get more insights and explanations for the quantitative data, we used narrative interviews, which were transcribed and then analyzed with the template analysis approach (King et al., 2019). Central to template analysis is the development of an initial coding template with a priori codes, which is then applied to a subset of the data and refined and adapted until it “capturers as full a picture of the analyst’s understanding as possible” (King et al., 2019, p. 219). We constructed the initial coding template based on the reviewed literature but also on the quantitative results. Based on the coding process, we formed three main categories: themes discussing interviewees’ individual experiences and perceptions of crowding, the role of other tourists’ social behavior and tourist-to-tourist encounters and the role of crowding at specific sites/events. According to King et al. (2019, p. 200), themes are “recurrent and distinctive features of participants’ accounts, characterizing particular perceptions and/or experiences, which the researcher sees as relevant to the research question.” The quantitative and qualitative findings are then integrated within the interpretation stage of the study (Onwuegbuzie & Teddlie, 2003).

Sampling strategies

Following Onwuegbuzie and Collins (2007), we used random sampling for the quantitative phase and combined it with non-random sampling for the qualitative phase. The quantitative sample (N = 224) was collected in a large and well-known ski area in Western Austria in February 2019, which is known as one of the most visited periods of the year due to the holiday season (Statistics Austria, 2020). On average, it took participants five minutes to complete the survey.

For the qualitative phase, we adopted information-oriented selection (Flyvberg, 2011) and only interviewed experienced skiers to guarantee a high probability of information. Therefore, we carried out 27 interviews with skiers, who have recently been skiing in a Tyrolean ski area with more than 100 km of slopes. Furthermore, we only gathered qualitative data in Germany because it represents the most important sending market for Austria’s ski industry (Statistics Austria, 2020; Vanat, 2019). Besides, this was supported by our quantitative data which showed no differences and effects for the country of origin. In this context, the size of the ski area seemed essential to ensure that skiers have previous crowding experiences as well as experiences of tourist-to-tourist encounters.

Measures and interview guideline

All items were adapted from previous research, but were adjusted to the ski context (Machleit et al., 1994). Since previous crowding studies in the winter sport context did not provide multi-item measures for crowding (Zehrer & Raich, 2016) or did not solely focus on the measurement of crowding (Matzler et al., 2008; Steiger et al., 2020), we adapted four multi-item constructs and three single-item measures from previous literature (see Appendix A for list of items). This is in line with Neuts and Nijkamp (2012, p. 2149), who called for alternative measures of perceived crowding.

Perceived crowding and sites

Previous research has shown the importance of human-induced crowding in the ski context (Zehrer & Raich, 2016). In line with Zehrer and Raich (2016), we understand perceived crowding as “the maximum number of people who can use a site without an unacceptable alteration in the physical environment and the social, cultural and economic fabric of the destination and without an unacceptable decline in the quality of the experience gained by visitors” (Wall & Mathieson, 2006, p. 33). Thus, we measured perceived human crowding with four items from Machleit et al. (1994). An exemplary item is “There were a lot of skiers in the ski area.” We also asked people about their perception of crowding at specific locations in the ski area. For these items, we used single-item measures (Rossiter, 2002) because the context was narrow in scope and the item to be judged concrete (e.g. crowding at valley stations, slopes and gastronomic facilities).

Moderating variables

Previous research indicated contrasting findings concerning demographics and visiting behavior in the ski context (Zehrer & Raich, 2016). Therefore, we adapted three items to measure self-reported skill levels of participants. An exemplary item is “I really enjoy skiing”. Additionally, previous research showed that the inappropriate behavior of other customers affects the intensity of perceived crowding (Neuts & Nijkamp, 2012). However, there was only limited knowledge of the moderating effects of other skiers’ behavior on satisfaction. Therefore, we used three items from Brady and Cronin (2001) to assess the moderating role of other customers’ inappropriate behavior on satisfaction. An exemplary item is “I find that this ski area’s other customers leave me with a bad impression”.

Satisfaction

In this study, we define overall satisfaction as the likelihood to engage in certain behaviors (Oliver, 2014), which is of relevance since previous research showed that satisfaction is essential ensuring post-trip behavior (Kozak, 2001; Oh, 1999). Thus, overall satisfaction was measured with five items, which we adapted from previous literature (Eroglu & Machleit, 1990; Machleit et al., 1994). Exemplary items are “I am satisfied with this skiing day” or “I am satisfied with my decision to visit this ski area”.

Interview guideline

To gather qualitative data, we carried out in-depth narrative interviews. For this purpose, we used a narrative style interview guideline (Beth, 2002), which was designed to explore individual perceptions of crowding as well as social behavior and situations of tourist-to-tourist encounters. In the first part, we asked about the behaviors of other skiers (e.g. When talking about other skiers, what comes to your mind? Which behaviors of other skiers do you like/dislike? Which behaviors influence your ski day positively and which negatively?). In the second part, we asked participants about perceived crowding on the slopes (e.g. Do too many skiers disturb you and if yes, when, where and why?) and finally in the third part we focused on the types of skiers (e.g. Which types of skiers/which groups disturb you especially on the slopes and why?).

Results

This study uses a triangulation mixed method design, therefore we first present the quantitative findings and synthesize it with the qualitative findings in the interpretation stage (Creswell & Plano Clark, 2011; Doyle et al., 2009). We assessed the relationships among the constructs with the partial least square (PLS-SEM) approach (Hair et al., 2012). First, we assessed the validity and reliability for each item and used composite reliability (CR) and average variance extracted (AVE) to test it. Table 2 shows the CRs (ranging from.838 to .935) and AVEs (ranging from .680 to 742) for each construct and additionally the loadings on the individual item level. One item was excluded since the factor loading did not exceed .60 (Hair, 2010). To check for discriminant validity, we used the Fornell-Larcker ratio (Fornell & Larcker, 1981) and found that the square roots of the AVEs are greater than the construct correlations (Appendix B). Cross-loadings were not a significant concern for the data and all items loaded the highest on the proposed factor. To test for common method variance, we used Harman’s one-factor test and included all items (Podsakoff & Organ, 1986). In this unrotated exploratory factor analysis, no single factor emerged (four factors explaining 68.25% of the variance) and the first factor did not explain the majority of the variance (the first factor explained about 33.18% of the variance). For the evaluation of the inner model (Q²_PC = .394, Q²_SAT = .221), we ran blindfolding using an omission distance of six, since literature suggests to select a value between five and ten but not a divisor of the number of observations (Hair et al., 2012).

Table 1. Qualitative sample overview

Interviewee	Age	Gender	Ski days per Year	Preferred ski resort
Int.1	25	male	30	Zillertal Arena
Int.2	31	male	20	Wilder Kaiser
Int.3	48	male	10	Wilder Kaiser
Int.4	51	female	10	Stubaier Gletscher
Int.5	54	male	5	Ischgl
Int.6	26	female	15	Kitzbühel
Int.7	25	male	15	Sölden
Int.8	26	male	25	Ski Arlberg
Int.9	27	male	15	Sölden
Int.10	53	male	5	Kitzbühel
Int.11	18	female	5	Ski Arlberg
Int.12	19	female	10	Stubaier Gletscher
Int.13	28	male	5	Fieberbrunn
Int.14	55	male	15	Ischgl
Int.15	27	male	10	Ischgl
Int.16	20	male	5	Wilder Kaiser
Int.17	50	female	10	Serfaus-Fiss-Ladis
Int.18	48	female	5	Stubaier Gletscher
Int.19	48	male	5	Zillertal Arena
Int.20	25	male	15	Sölden
Int.21	24	male	20	Kitzbühel
Int.22	26	male	5	Stubaier Gletscher
Int.23	49	female	5	Fieberbrunn
Int.24	25	female	10	Ischgl
Int.25	48	female	10	Serfaus-Fiss-Ladis
Int.26	29	male	30	Sölden
Int.27	27	female	35	Stubaier Gletscher

“Int.27, 4” etc. refers to interviewee “27”, paragraph “4” of the transcript

Table 2. Psychometric properties of the constructs

Constructs and items	Loading
Perceived crowding (CR = .890, AVE = .669)
PC1 The ski area was a little too busy	.860
PC2 There were lot of skiers in the ski area	.834
PC3 There was much traffic in the ski area during my skiing trip	.802
PC4 The ski area seemed very crowded to me	.771
Social behavior (CR = .838, AVE = .722)
SC1 I find that this ski area’s other customers leave me with a bad impression	.856
SC2 Other skiers have attracted my attention through bad manners	.843
Skill level (CR = .863, AVE = .680)
EL1 I really enjoy skiing	.927
EL2 I am a good skier	.785
EL3 I go skiing a lot	.651
Satisfaction (CR = .935, AVE = .742)
SAT1 I am satisfied with my decision to visit this ski area	.910
SAT2 I am satisfied with this skiing day	.895
SAT3 This skiing day met my expectations	.892
SAT4 I would revisit this ski area	.812
SAT5 I would recommend this ski area to other people	.792
Crowding location (single items)
CL1 The ski slopes were crowded	-
CL2 The valley stations of the ski lifts were crowded	-
CL3 The gastronomic facilities were crowded	-

Note: The item “This company’s other customers do negatively affect my skiing experience” was dropped because of factor loading .582

Figure 1 shows that crowding at the valley station (β = .200, p < 0.01) and at the slopes (β = −0.613, p < 0.000) significantly increases perceived crowding. For crowding at the gastronomic facilities, no effects were found. The results support H1a and H1b but not H1c. H2 predicted a negative relationship between perceived crowding and satisfaction. We found that perceived crowding has a significant negative impact on satisfaction (β = −.159, p < 0.05). Additionally, we observed that individual skill levels (β = −.182, p < 0.05) and social behavior (β = −.105, p < 0.05) moderate the relationship between perceived crowding and satisfaction. Hence, H3 and H4 are fully supported.

Figure 1. Crowding location, perceived crowding, skill levels, social behavior, demographics and satisfaction

Note: n.s. = not supported, p.s. = partially supported

The variables such as age also showed significant impacts on perceived crowding (β = −.086, p < 0.05) and satisfaction (β = .143, p < 0.05), while gender has significant effects on satisfaction (β = .103, p < 0.05) but not on perceived crowding (Table 3). Furthermore, no effects were found for the country of origin. In summary, H5 and H6 are only partially supported.

Table 3. Crowding location, perceived crowding and satisfaction

			Estimate	T-value	P-level
Valley station	→	PC	.200	2.816	.005
Slopes	→	PC	.613	11.177	.000
Gastro	→	PC	.073	1.244	.214
PC	→	SAT	−.159	2.509	.012
SL	→	SAT	.413	4.315	.000
SB	→	SAT	−.153	2.330	.020
PC*SL	→	SAT	−.182	2.056	.040
PC*SB	→	SAT	−.105	2.085	.037
Sex	→	PC	−.024	.542	.588
Sex	→	SAT	.103	2.184	.029
Age	→	PC	−.086	1.982	.048
Age	→	SAT	.143	2.248	.025
Country	→	PC	.081	1.932	.053
Country	→	SAT	.027	.474	.635

Discussion

The Alps absorb about 44% of the world’s skiing attendance and the Alpine countries are the dominant players in terms of market share (Vanat, 2019, p. 27). Austria is visited by 51.8 million skiers, of whom 49% of visits are recorded by the 79 ski areas within Tyrol (Statistics Austria, 2020). As a result, crowding is known to represent an issue for ski areas (Vanat, 2019; Zehrer & Raich, 2016). The quantitative findings show that perceived crowding has a negative effect on satisfaction and that this relationship is affected by skill levels and the social behavior of other skiers (Table 3). These findings are in line with Shelby and Heberlein (1984), who showed that crowding represents a subjective perception that is driven by several variables.

To better account for these subjective perceptions, the following sections provide a more detailed exploration of perceived crowding. This is supported by the triangulation mixed methods design (Creswell & Plano Clark, 2011; Doyle et al., 2009), which helped us to improve the significance of our findings by contrasting the results of the quantitative phase with the qualitative findings and discussing them against the background of existing literature (Collins et al., 2006).

Individual crowding perceptions

We adapted the crowding construct from Machleit et al. (1994), who defined it as human and spatial crowding. Therefore, we were particularly interested in exploring the crowding understanding of our interviewees in the qualitative phase. Asked about “a perfect skiing day” many interviewees related this to factors such as weather, service quality or snow availability. Frequently, attractive slopes and fast service at the ski lifts were perceived as essential. However, interviewees also related it with uncrowded slopes and mentioned the vital role of other skiers for their ski experience. Therefore, it became evident that crowding in ski areas represents a social phenomenon (Yagi & Pearce, 2007). An interviewee explained:

A successful skiing day is made up by the fact that there is no traffic jam or crowding in ski buses. That parking is guided and that I have no stress. Shops that offer service are always good in case something breaks down. (…) In general, the feeling of safety is important to me and of course, I find it cool when there is less activity. But for me the other people are already part of it. You can talk and chat. A successful skiing day is all about surrounding, other people, good weather” (quote “Int.29, 4”: interviewee “29”, paragraph “4” of the transcript, see Table 1).

In line with the findings of Zehrer and Raich (2016), some interviewees reported coping behavior, such as going and finishing skiing early. An interviewee added:

For me, the lift must open at 8 o’clock, not 9 o’clock but 8 o’clock. The first 1.5h are the best because there are fewer people on the slopes, there are no wait times, no queues and you have the long descents and wide slopes to yourself. (Int.2, 4)

While earlier literature has strongly focused on the effects of crowding and our findings confirmed this relationship (Table 3), we wanted to learn more about the individual meaning of crowding for skiers. The interviewees explained that crowding has adverse effects on their experiences (e.g. following Matzler et al., 2008 waiting times at ski lifts are unfavorable for satisfaction) and an interviewee made a particularly demonstrative example:

Yeah, it just bothers me when there’s a lot going on. First of all, you have long waiting times, meaning you have to wait 5-10 minutes for the ski run and if it is an old or slow lift you have to queue for 10 minutes and it takes another 10 minutes to get to the top. Therefore, it is annoying when there is a lot of crowding, time is lost. And on the slopes, it’s no fun anymore, you have to be concentrated and can’t go fast or use the full slope. (Int.1, 22)

Crowding and tourist-to-tourist encounter

Previous studies highlighted that the number of tourists and their behavior plays an essential role in the overtourism context (Namberger et al., 2019; Neuts & Nijkamp, 2012). Our findings confirm the significant role of other skiers’ social behavior to moderate the effect of perceived crowding (Table 3). An interviewee explained the high importance of other skiers.

In my opinion, other skiers are almost more important than the weather. You can ski in any weather; even if it snows, it is not that bad. But, if the other skiers’ skills or behavior doesn’t fit, it’s just annoying. (Int. 12, 48)

Furthermore, these findings confirm Neuts and Nijkamp (2012, p. 2149), who identified the number of tourists and the behavior of other tourists as a driver of crowding. In this context, the quantitative findings also confirmed this relationship (Table 3), but the qualitative data also offers a more nuanced view. For example, other skiers are understood to be important, but interviewees also recognized other factors such as the weather or quality of slopes to be essential.

I think it is more important that the weather is good and the slopes are properly prepared because I think you can ignore the behavior of the other skiers. If the snow and everything are good, then everything is perfect and the rest is not so important. And it is not so often that you have bad experiences with other skiers. (Int.11, 48)

This is interesting since previous research has shown the importance of snow reliability for regional economies (Steiger & Scott, 2020). Additionally, Steiger et al. (2020) showed that improvements in the price-performance ratio by lowering ticket prices and avoiding crowding can compensate for marginal snow conditions. Thus, yield management (e.g. dynamic pricing models) will become an important tool in managing customers’ expectations and crowding issues.

Sites of crowding perceptions

Since previous literature showed that overtourism can result from temporal, spatial overuse of resources (Peeters et al., 2018), we were particularly interested in the interviewees’ crowding perceptions at three different locations. These locations (valley station, slopes and gastronomic facilities) are highly important since they represent either the touchpoints for the service encounter (Bitner et al., 1990), determine part of the ski experience (Won & Hwang, 2009) or play an important role for overall quality perceptions (Muskat et al., 2019). The quantitative findings showed that crowding at the valley station and the slopes contributes significantly to the crowding perception (Figure 1). One interviewee explained his crowding experiences at the valley station in the following way:

Everybody knows it, just before getting into the gondola there is always pushing and shoving, and that always annoys me. Once they get past the turnstile, which is like a boundary, then they get nervous. Everybody starts pushing, everybody seems to be afraid that they won’t get a seat in the gondola. (Int.1, 16)

This confirms that crowding at the valley station contributes to higher levels of crowding and highlights the necessity to not just reduce waiting times (Bielen & Demoulin, 2007) but also invest in innovative and stress-reducing queuing techniques (Mattila & Hanks, 2012). Nevertheless, the results (Table 3) show that crowding on the slopes has the strongest impact on crowding perceptions. This aspect was also reflected by the interviewees, where we found that the waiting time at the valley station is less critical than crowding at the slopes. An interviewee explained:

My opinion is that it is better when there are fewer people on the slopes. You simply have more freedom. I am a skier who is always alert and careful. In other words, when there are fewer skiers on the slopes, you have more freedom. Then it is nicer to go skiing. When there is a lot going on, you really have to be careful. (Int.13, 26)

While previous research (e.g. Steiger et al., 2020) has not investigated the role of secondary activities such as gastronomic facilities, we discovered in our interviews that crowding in these facilities contributes to perceived crowding, however our quantitative findings did not confirm this relationship (Table 3). An interviewee explained:

It is often the case that people starting fighting over places. Then two people occupy the whole table and do not want anyone to sit with them. Anyway, there are always too few seats. (Int.22, 18)

The quantitative findings showed that older skiers have stronger perceptions of crowding and thus we confirm Zehrer and Raich (2016), who showed that older visitors significantly perceived the ski slopes more crowded (2016, p. 94). However, these findings are in contrast to previous studies outside the ski context, which indicated higher sensibility of younger people (Fleishman et al., 2004; Rasoolimanesh et al., 2016). The interviewees highlighted that for younger people, crowding is less an issue because for them it represents more of a social event, compared to older generations who used to ski on even more crowded slopes and therefore perceived and evaluated crowding differently. In contrast to Zehrer and Raich (2016), gender had no effect on crowding perceptions. Nevertheless, our findings show that female participants tend to be more satisfied than male participants (Table 3). While Rasoolimanesh et al. (2017) found no significant effects of age on crowding, Matzler et al. (2008) found only weak influences of age as a moderator for the overall satisfaction with ski resorts. In the more diverse urban context, Neuts and Nijkamp (2012) showed that the country of origin has an impact on the crowding perceptions but our findings do not show effects for country of origin (Table 3). This contrasting finding can be understood as a result of the currently rather homogenous ski market within the Alps (Vanat, 2019), where cultural differences are less applicable than in other studies, which compared e.g. European, American, and Asian-African visitors (Fleishman et al., 2004; Neuts & Nijkamp, 2012). We conclude that the effects of age, gender and country of origin are strongly dependent on the setting and the cultural backgrounds and therefore these results need to be interpreted with caution.

Conclusion, implications and future research

Overtourism can be deconstructed to spatial dimensions, excessive numbers of tourists and negative effects on the environment, economy and residents (Dodds & Butler, 2019a; Peeters et al., 2018). Crowding represents a real issue for ski areas (Vanat, 2019; Zehrer & Raich, 2016) and thereby highlights the particular relevance of this study. The study set out to investigate the crowding-satisfaction relationship and found that crowding at strategic sites such as the valley stations, the slopes or the gastronomic facilities contributes to increased levels of overall crowding perceptions. The findings show that crowding, which involves the spatial overuse of available capacity, too many visitors for the existing infrastructure or local conditions (Peeters et al., 2018), results in adverse effects for satisfaction of skiers (Kim et al., 2016; Rathnayake, 2015; Zehrer & Raich, 2016). Another major finding was that crowding is perceived differently based on individual characteristics (Neuts & Nijkamp, 2012). Thus, the novel contribution of this study relates to the moderating role of social behavior and skill levels (Namberger et al., 2019; Yagi & Pearce, 2007). These findings were also supported by the qualitative results, which underlined that inappropriate social behaviors such as jostling and grumbling affect skiers’ satisfaction. However, while the quantitative data show a strong negative and moderating effect, the qualitative findings emphasize the particular importance of tourist-to-tourist encounters for the overall ski experience. Nevertheless, the crowding-satisfaction relationship varies according to age and gender.

The theoretical contribution strengthens the importance of social factors in defining carrying capacity (Coccossis et al., 2001) and highlights the need to extend current quantitative overtourism measures (e.g. defining upper limits for Venice) with qualitative insights (Saveriades, 2000). This study and previous work (Graefe et al., 1984; Neuts & Nijkamp, 2012) also showed that the presence and behavior of other tourists has significant effects on the crowding-satisfaction relationship. In this context, it could be the case that despite increasing visitor numbers, tourists who exhibit compatible behavior, result in lower perceived levels of crowding and ultimately increased satisfaction.

The managerial implications of this study are threefold. First, since our qualitative results also confirm the importance of tourist-to-tourist encounters, the contribution of this paper is to launch a plea for a stronger focus on the latter. Destination management of the future cannot consist of solely defining quantitative measures and upper limits for tourists, but rather in continuing to guarantee a high experience level for tourists, which seems to also depend on the behavior of other tourists in crowded settings. These social interactions between different tourists should be considered, analyzed and finally coordinated to increase tourists’ satisfaction at crowded sites (Martin & Pranter, 1989). This can be supported by smart technology and customer involvement to better manage the innovation process (Pikkemaat et al., 2019). Second, research has shown that people tend to visit higher regions in times of scarce snow (Steiger et al., 2020). Simultaneous to increased demand, customers demand higher quality and these ski areas need to counteract crowding with e.g. price-related distribution mechanisms (Fonner & Berrens, 2014) or focus on further improving service quality (Brady & Cronin, 2001). In the latter context, ski areas will need to further invest and innovative their offers in order to stay competitive (Pikkemaat et al., 2019). Overall, it will become increasingly important to better manage the flow of visitors within ski areas (Albrecht, 2017), which can help to reduce site-specific crowding effects. Real-time flow management tools, which use placed-based information from lift use (Kah et al., 2011), will help to better distribute skiers. Third, since the ski market is consolidating (Vanat, 2019), ski areas need to better manage the dissemination of destination-related information such as e.g. snow conditions or the crowding status to secure satisfaction (Scholl-Grissemann et al., 2019). In the long run, it might be more promising to supplement the quantitative crowding measures with management recommendations for tourist-to-tourist encounters, not only in ski areas, but also beyond.

We acknowledge that crowding represents only one of the many aspects, which impact quality and satisfaction, but we argue that it will represent an even more critical discipline for ski areas who want to thrive in the competitive markets of the future (Steiger & Scott, 2020; Vanat, 2019). Further work is required to establish the effects of distribution mechanisms, which aim to reduce crowding. Additionally, research needs to consider the potential of upstream processes such as skiers’ information-seeking behavior (Scholl-Grissemann et al., 2019) and downstream processes such as social media interactions. Future research can use a sequential mixed design (Onwuegbuzie & Leech, 2005) and cross-validate the findings with objective data such as sold tickets, waiting times at ski lifts or available parking lots.

This paper also shows several limitations. First, the quantitative sample is relatively small and based on a single survey of crowding within a particular ski area. To overcome these issues, we used a PLS-SEM instead of a covariance based approach (Hair et al., 2019), but caution must be applied, as the findings might not be transferable to other settings. Second, additional measurements at different dates and locations could lead to more detailed findings. Third, the qualitative phase provided detailed insights into crowding perceptions but only included a few international skiers. Finally, the used sample includes major outbound countries such as Germany, but other countries with different cultural backgrounds were underrepresented. In this context, including countries with other cultural backgrounds could lead to interesting findings concerning the role of social encounters, primarily since Eastern Europe, Central Asia and Asia & Pacific represent markets with future growth potential (Vanat, 2019).

Acknowledgments

The authors would like to thank Benedikt Burger, MSc and Wolfgang Weissenbach, MSc for their support in the initial project phase and for supporting the data collection process.

Disclosure statement

No potential conflict of interest was reported by the authors.

Appendices

Appendix A. List of items

Items and constructs	Sources
Satisfaction
I am satisfied with my decision to visit this ski area	Eroglu & Machleit, 1990; Machleit et al., 1994
I am satisfied with this skiing day
This skiing day met my expectations
I would revisit this ski area
I would recommend this ski area to other people
Skill level
I really enjoy skiing	Matzler et al., 2008; Won et al., 2008
I am a good skier
I go skiing a lot
Perceived crowding
The ski area seemed very crowded to me	Machleit et al., 1994
There was much traffic in the ski area during my skiing trip
There were lot of skiers in the ski area
The ski area was a little too busy
Crowding location (single item measures)
The valley stations of the ski lifts were crowded	Machleit et al., 1994
The ski slopes were crowded
The gastronomic facilities were crowded
Social behavior
I find that this ski area’s other customers leave me with a bad impression	Brady & Cronin, 2001
This company’s other customers do negatively affect my skiing experience
Other skiers have attracted my attention through bad manners
Demographics
Gender Age Country of origin	Neuts & Nijkamp, 2012; Zehrer & Raich, 2016

Appendix B. Construct descriptives and correlations

Construct means, standard deviations and correlations
	Mean	SD	PC	SB	SL	SAT	CL1	CL2	CL3
PC	3.52	1.34	1
SB	2.81	1.46	.288**	1
SL	5.41	1.23	.030	.136*	1
SAT	6.35	0.82	−.273**	−.228**	.263**	1
CL1	2.88	1.75	.520**	.180**	.067	−.229**	1
CL2	3.17	1.57	.739**	.283**	.034	−.185**	.452**	1
CL3	2.99	1.63	.374**	.195**	.002	−.235**	.393**	.340**	1

N = 224, *p < .05,**p < .01,***p < .001