The need to adapt to travel expenditure patterns. A study comparing business and leisure tourists in Barcelona

An understanding of expenditure patterns is vital to travel organizers and destination marketers. Using surveys conducted by the Barcelona Tourism Board (Turisme de Barcelona), this study profiled the trip by purpose: leisure or business, and investigated the relationships between travel expenditure, length of stay, repeat visits and satisfaction. Different trip-related and socio-demographic variables, such as trip purpose, number of nights spent and repeat visits were found to contribute to explaining the expenditure variations. The theoretical model is based on the literature, and puts forward different hypotheses, which are tested using SEM. The findings show behavioural differences between those groups and are useful for tourism managers of urban destinations.


Introduction
Tourism is becoming the most relevant industry worldwide, and tourist expenditure is contributing decisively to economic growth (Brida and Pulina 2010). In recent years, urban tourism has been one of the fastest-growing forms of tourism. This social and economic phenomenon has caught the attention of several researchers, including Ashworth (1996), Murillo et al. (2013) and Vandermey (1984), among others. Ashworth and Page (2011) highlight the benefits that tourism brings to cities. Consequently, it is important to analyse the role of tourism in the local economy and to understand the behaviour of the tourist, with a view to adapting the city to their needs, if necessary.
Over the past 30 years, researchers as Dardis et al. (1981); Hong et al. (1996) and Gomes de Menezes et al. (2008) have used the socio-demographic profile of travellers, the characteristics of their trips and the nature of the activities undertaken.
In the current environment, competition among traditional destinations and cities is increasingly intense (Wöber 2000), and firms and destinations within the industry are vying with each other to gain and retain customers (Charterina and Aparicio 2015). Therefore, a better understanding of visitor expenditure patterns is extremely important for travel organizers, destination marketers and policy-makers when developing effective strategies to boost tourism receipts (Lee et al. 2015). In this respect, visitor expenditure patterns have been recognised as a good segmentation system to increase yields (Petrick and Sirakaya 2004).
The contribution of this present study is to explore the combination of tourist spending patterns and the motivational patterns of business and leisure tourists, and its effects on the relationships between the constructs of expenditure, satisfaction, length of stay, and repeat visits. These relationships are measured using structural equation models (SEM) with multi-group analysis, which allows us to describe and explain the differences between the groups studied, for example the different levels of purpose under tourist behaviour.
The structure of the paper is as follows: in the next section the constructs, the segmenting effect of purpose and the structural model proposed are reviewed; section three presents the case study and the methodology; finally, the results are described, and the findings presented, as along with the limitations and future research arising from this work.
2 Literature review and hypotheses 2.1 Purpose: business vs. leisure Literature on tourism has focused on explaining the specific reasons why people travel or the needs that they intend to satisfy by means of their trip. While for Iso-Ahola and Allen (1982) leisure activities such as tourism are motivated by the desire for personal rewards and the desire to escape their normal living environment, some authors, including Baloglu and Uysal (1996), Crompton (1979), Yuan and Mcdonald (1990) differentiate between destination-related factors (or pull motivation), and personal issues (or push motivation). Along these research lines, tourism is analysed from the perspective of travel for personal reasons, linked to leisure and entertainment. Nevertheless, the perspective of travel related to business and professional meetings also has to be taken into consideration.
According to UNWTO recommendations, the category of personal trips includes holiday, leisure and recreation, visiting friends and relatives, education and training, health and medical care, among others; and they differ from a second major motivation or group of trips, namely business and professional (United Nations 2008). This article adopts these definitions, focusing on trips for personal reasons: holidays, leisure and recreation; and trips for business and professional reasons. The first category includes sightseeing trips, attending sporting events or culture, resting, honey-mooning, fine dining, while business and professional trips include such activities as attending meetings, conferences, fairs, exhibitions or similar.
The present paper uses SEM techniques with multi-group analysis, and takes purpose as the segmenting variable. Two segments, or groups, are created: business and leisure. The first group, business, are tourists on a business-related or professional trip, attending a fair, congress, conference, symposium or meeting. The leisure group is made up of those on leisure or sightseeing trips, tourists, and people attending sports or cultural events.

Tourist expenditure
Previous research on tourist spending enables us to differentiate between macroeconomic and microeconomic studies. According to Wang and Davidson (2010), it is crucial to understand the differences between them, and to Alegre and Pou (2004), to show the advantages and purposes of each. In keeping with the objective of this research, the review of the literature has focused on the microeconomic studies.
In recent years it is worth noting the appearance of studies that have described spending patterns using econometric models (Aguiló Perez and Juaneda 2000; Alegre et al. 2011a;Chang et al. 2013;Díaz-Pérez et al. 2005;Pulido-Fernández et al. 2015;Thrane 2014;Zhang et al. 2012). These studies aim to demonstrate the effects of a range of variables on tourism expenditure, and to predict the behaviour of tourists.
Another line of research corresponds to studies that apply other statistical techniques, such as multivariate models. Jang et al. (2005) use a path analysis to test the relationships between the activities of the trip and spending, and conclude that they generate different levels of spending. Seiler et al. (2003) test the relationship between tourism expenditure, traveller demographics and certain characteristics of the trips, such as length of stay. They conclude that travellers whose stays are longer tend to consume more and vice versa. Chaitip et al. (2008) show that the total cost of the trip is related to product attributes and product management. The study by Thrane and Farstad (2012) demonstrates a positive relationship between length of stay and expenditure. Based on the above review of the literature, we can generate the following hypotheses: (H1) Tourist expenditure has a direct relationship on satisfaction. The effect of this relationship will be different depending on the purpose. (H2) Tourist expenditure has a direct relationship on repeat visits. The effect of this relationship will be different depending on the purpose. (H3) Tourist expenditure has a direct relationship on length of stay. The effect of this relationship will be different depending on the purpose.

Satisfaction
The research on tourist satisfaction is based on different consumer behaviour theories. Traditionally, the expectation-disconfirmation theory is widely accepted (Oliver 1999). In tourism research, tourist satisfaction plays an enormously important part in the behaviour of the tourist. Kozak and Rimmington (2000) consider that tourist satisfaction is one of the key elements for successful destination marketing, and influences both the behaviour of the tourist at their destination, and their decision to repeat their visit.
Concerning tourist satisfaction, we can highlight the number of studies that have shown that there is a relationship between satisfaction and other constructs, including loyalty or image, among others (Campo and Yagüe 2008;Charterina and Aparicio 2015;Zhang et al. 2009;Reisinger and Turner 2002).
This article considers satisfaction in terms of overall satisfaction (Oliver 1980). Therefore, if we accept satisfaction as an antecedent of loyalty (Oppermann 2000), and the satisfaction level is high, the level of loyalty will also be high. Afterwards, satisfied tourists will be more inclined to repeat their visits in the future (Martínez-Ruiz et al. 2010;Oppermann 2000). For Moniz (2012), repeat visits are sustained by satisfaction with the destination, while Chi and Qu (2008) prove positive and direct relationships between satisfaction and loyalty. Thus, on the basis of these studies we put forward the following hypotheses: (H4) Satisfaction has a direct relationship on repeat visits. The effect of this relationship will be different depending on the purpose.

Length of stay
The tourism research literature suggests that the length of tourist travel is one of the most important tourist demand variables (Uysal et al. 1994), and it has been studied from different perspectives. One line of research comprises those studies that analyse the characteristics of travellers segmented by the length of stay and the behaviour of tourists during their trip. In relation to length of stay, the similarities and differences between different groups of travellers are analysed (Crompton et al. 1992;Downward and Lumsdon 2003;Oppermann 1995).
Another research line corresponds to those studies that apply econometric models to explain and predict the duration of the stay (Alegre et al. 2011b; Barros and Machado 2010;Eugenio-Martin 2003;Ferrer-Rosell et al. 2014;Martínez-Garcia and Raya 2008).
In addition, some studies have focused on market segments based on tourist profiles and travel characteristics in order to help destination managers to gain a better understanding of the consumer, better meet their needs and implement better marketing strategies (Oppedijk van Veen and Verhallen 1986). Davies (1990), Murphy et al. (2010) and Pike (2002) study short breaks, while Charterina and Aparicio (2015) and Dunne et al. (2011) analyse city breaks. Finally, it is important to highlight the contributions of Neal (2004), which demonstrate the effects of the length of stay on tourist satisfaction. The references above allow us to generate the following hypothesis: (H5) Length of stay has a direct relationship on satisfaction. The effect of this relationship will be different depending on the purpose.

Repeat visits
The role of repeat visits has been framed in the marketing literature in the afterpurchase behaviour. In this regard, Boulding et al. (1993) investigate the factors that influence the repurchase intention, while Oliver (1999), believes that the intention to repeat the act of purchase is part of loyalty, and Flavian et al. (2001) 2015) and Osti et al. (2012) differentiate between tourists who repeat and non-repeaters.
Regarding the concept of repeat visits we should also note that for Kozak et al. (2005) satisfaction and the number of visits influence the intention to undertake future visits, and for Moniz (2012) the intention to repeat arises from the overall satisfaction with the destination.
Finally, of particular interest to this study are those studies that demonstrate causal relationships between repeat visits and other concepts. Ledesma et al. (2005) which concluded that the length of stay may increase the number of repeat visits, are particularly interesting, so we can also put forward the following hypotheses: (H6) Length of stay has a direct relationship on repeat visits. The effect of this relationship will be different according to the purpose.

Structural model proposed
The purpose of this analysis is to test the hypotheses previously proposed with reference to the relationships between different pairs of variables as a whole, in order to determine the effects and importance of these relationships, and the contribution of purpose as discriminatory in the multi-group model. Figure 1 shows the proposed theoretical model. Model components correspond to the above hypotheses.
The purpose construct differentiates between business and leisure tourist. Previous studies proved the relationships of expenditure on length of stay (Thrane and Farstad 2012), length of stay on satisfaction (Neal 2004) and length of stay on repeat visits (Ledesma et al. 2005). Therefore, our causal model proposed analyses the relationships of expenditure, satisfaction, length of stay and repeat visits with the principal contribution of using the purpose of stay as tourists' segmentation variable (business and leisure).

Case study
Barcelona is one of the most popular cities to visit in Europe, as well as being a successful example of urban tourism (Gómez et al. 2004) due, among other factors, to its unique heritage, its strategic location as an economic and administrative capital, its attractiveness, its Mediterranean character and the 92 Olympic Games (López Palomeque 2006).
Barcelona has been able to accumulate symbolic capital, which depends on the values of authenticity and uniqueness. This values are highlighted by the 7.9 million tourists who stayed in hotels in the city in 2014, and 17 million overnight stays. This popularity has also placed the city among the leading congress destinations in Europe, with more than 579,000 delegates participating in tourism-related events (Turisme de Barcelona 2014).
This empirical study was conducted among tourists who stayed at accommodation in the city of Barcelona between 2009 and 2011. The scarcity of scientific studies relating to tourism demand and consumer spending patterns confirms the relevance of this study, and highlights the importance of new strategies both for DMO and the private sector to improve their strategies.

Sample and data gathering
For the evaluation of the model we used the survey conducted by Barcelona Tourism through paper and pencil interviews, with an interviewer moving among The purpose of the visit is the discriminatory variable that allows to perform a multi-group analysis. Two groups were created: business and leisure. The business group comprises tourists attending meetings, conferences, congresses, fairs, exhibitions or similar. The leisure group is made up of tourists traveling for vacation, sightseeing trips, natural or man-made visiting sites, attending sporting or cultural events, resting, honey-mooning, fine dining or similar. Table 1 shows the distribution of the sample for business and leisure groups and their demographic characteristics. The data show a higher proportion of men than women, with a preponderance of middle-aged (between 35 and 44 years), skilled workers of European origin.
The main characteristics of the trips are shown in Table 2. Trip characteristics highlight some features of the tourists. For example, they indicate an average stay of four nights, a higher proportion of tourists staying in a hotel (68.1 %) travelling by plane (77.2 %) and with their spouse or partner (45.8 %).
Looking into the distinctive characteristics of business and leisure groups, we highlight below some significant features. For example, in the business group spending is significant higher (100 €), as well as the slightly shorter length of the stay, a preference for travelling by plane (87.6 %), staying at hotels (88.7 %) and accompanied by colleagues (50.1 %). In the leisure group, lower spending can be observed, as well as a longer length of stay (4.34 nights), a higher preference for using other types of accommodation than hotels (38.0 %), and a larger proportion of tourists travelling with their spouse or partner (56.2 %). Table 3 shows the variables used to estimate the structural model and to study the hypotheses and the model.

Results and discussion
In order to determine the relationships depicted in Fig. 1, a structural equation model (SEM) was carried out with Mplus 7. The structural model was studied using a multi-group analysis with the segmenting variable tourist purpose, where the professional and leisure categories represent mutually exclusive and independent groups.
The structural model was estimated using maximum likelihood. The results and the degree of compliance with the various hypotheses are presented in Table 4.
The results of the hypotheses have multiple interpretations, which are summarised below. Eurasian Bus Rev (2016) 6:253-267 259 Firstly, in relation to hypothesis H1, the effect of expenditure on tourist satisfaction is significant for leisure travellers, i.e. satisfaction is higher for tourists spending more at the destination. This effect is not significant in the business group (0.004) and therefore this hypothesis is not supported for that group.
Secondly, for H2 hypothesis, the effect of expenditure on repeat visits is statistically significant for leisure tourists. The results of this hypothesis suggest that leisure travellers with a higher spending at the destination, have a higher degree of  repeatability. For the business tourists this effect is not significant (0.031), so this hypothesis is not supported for that group. Thirdly, in hypothesis H3, tourist expenditure has a significant negative effect on the length of stay both for business (-0.247) and leisure (-0.201) travellers. This hypothesis suggests that tourists whose spend more, stay shorter at the destination.
Fourthly, for hypothesis H4, satisfaction level has a significant negative effect on the level of repeat visits in the case of business tourists (-0.077) and leisure tourists (-0.030). Therefore, the greater the satisfaction, the lower the number of repeat visits.
Fifthly, in H5, the effect of the length of stay on satisfaction is positive and significant for business (0.145) and leisure (0.037) travellers. In both cases, the greater the length of stay, the higher the satisfaction.
Finally, in H6, the effect of the length of stay on the repeat visits is negative and statistically significant only for business tourists (-0.111). The result seems to show that the greater length of stay, the lower the number of repeat visits. That is, tourists who have stay longer come less often. Interestingly, this effect is not significant for leisure (0.001).
Of the initial six hypotheses, all are supported for at least one of the groups of purpose. The above results demonstrate the importance of purpose as a key variable for practitioners. The differences found between the two patterns studied, professionals and leisure, are significant. For the business group, we highlight two aspects. The first relates to their travel characteristics: higher spending, slightly shorter length of the stay, preference for travelling by plane and staying at hotels. The second aspect is related to our working hypotheses. The effect of expenditure on satisfaction (H1) and expenditure on repeat visits (H2) are not supported by our results. These two hypotheses are supported for leisure travellers only.
For the leisure group we observe a lower spending pattern, a longer length of stay, a higher preference for using other types of accommodation and a larger proportion of tourists travelling with their partners. Moreover, the effect of length of  (Kozak 2001). It should be noted that the causal relationship between expenditure and length of stay (H3) is supported and negative for both groups, with a higher coefficient support for the business group (-0.244), than for the leisure (-0.210). This finding is consistent with the previous research of Barros and Machado (2010), although the urban nature of Barcelona and the technique used for the analysis of data are different. Furthermore, the effect of satisfaction on repeat visits (H4), is supported and negative, with an important difference between their coefficients, (-0.077) for business and (-0.030) for leisure. This finding is not consistent with other studies (Kozak 2001). Therefore, we can speculate that the level of overall satisfaction does not stimulate visits to the same destination. The explanation might stem from the location studied and the statistical technique used in our analysis. The results of the hypothesis of length of stay and satisfaction (H5) are consistent with Neal's study (2004), demonstrating the effects of the length of stay on tourist satisfaction.
The results of this paper also demonstrate that our theoretical model is validated, but the degree of support of each hypothesis varies depending on the purpose. This finding suggests that relationship between some constructs may differ for business and leisure purpose when travelling to an urban destination. For example, the relationship of length of stay on repetition (H6) varies according to whether the visitors are on a business or on a leisure trip. Therefore, the causal relationship between that pair of variables is supported and negative for business trips, and not supported for leisure travellers.
Although for destination managers the purpose of the trip is very important, limited research has been carried out to investigate the effect in structural relationships. Our results supported four hypotheses for the business tourist and five in the case of leisure travellers. This finding reconfirms that it is worthwhile to develop differentiated marketing strategies for the two groups.
The relationship between tourist expenditure and satisfaction and tourist expenditure and repetition for leisure travellers constitutes a remarkable finding. When increasing the expenditure, tourist satisfaction and repeatability also increases. Bearing in mind that more than half of visitors to Barcelona are on leisure trips, DMO and companies need to better define their products to this group. The higher spending, the higher degree of satisfaction and repeatability.
Finally, on the subject of repeat visitation, our results suggest that as repeat visits increase, satisfaction level decreases. We open the discussion about this matter. This might be linked to the fact that urban repeaters experience more things, and the knowledge of the destination generates a reduction of expectations. Therefore, the creation of new experiences, would help to generate new expectations.

Limitations and future research
The findings are subject to some limitations that must be taken into account. These results correspond to the city of Barcelona. To be able to generalise them, the same study would have to be carried out in other tourism cities.
This study uses data from an existing survey carried out by Turisme de Barcelona, the consortium that exists to promote the city. Both the questions and the responses of tourists present limitations. The use of multi-element scales measures in future studies would enhance the interpretation of results.
For future research, we recommend the use of the same criteria and variables to explore length of stay and tourist expenditure as a segmenting variable between the constructs.