Tourism consumption revisited-an empirical study of Finnish consumers

According to postmodern theories, social divisions are based on identity and lifestyle rather than on sociodemographic background. In this paper, the effects of these both were examined. Tourism consumption was divided into two dimensions, tourism consumption and desire to travel more. Empirical analysis were based on two nation-wide surveys, Finland 1999 and Finland 2004. Sociodemographic factors were assumed to influence more on tourism consumption than on desire to travel more. The results were somewhat parallel with the hypothesis. However, both demographics and lifestyle determinants should be taken into account. The effects have remained quite stable regardless of the finding that desire to travel more has decreased while tourism consumption has increased.


Introduction
After spending money on compulsory necessities, people are left with many alternatives. Tourism is one of those discretionary alternatives competing with other leisure activities. However, it is widely agreed that tourism holds an important part in the lives of the Western citizens. Nevertheless, there are remarkable differences between consumers' travelling habits (Honkanen 2004). Some people are willing to spend more money and leisure time on tourism than other people. For these "real tourists" tourism has become part of their lifestyle.
In the recent research of tourism, the concept of lifestyle has been mainly taken into account in the context of destination choices or activities during the trip (see Chandler & Costello 2002;Cleaver & Muller 2002;Rajasenan & Ku-Acknowledgements: The authors would like to thank Jani Erola, Pekka Räsänen and Terhi-Anna Wilska for conducting and giving a permission to use the data. The authors are also grateful to Finnish Social Science Data Archive. Addresses: Antti Honkanen, VAMK, University of Applied Sciences, Business Economics and Tourism, Raastuvankatu 29, 65100 Vaasa, Finland. Pekka Mustonen, Turku School of Economics, Rehtorinpellonkatu 3, 20500 Turku, Finland. E-mails: antti.honkanen@puv.fi, pekka.mustonen@welho.com mar 2004; Reisinger et al. 2004). In this article, the main focus was to examine how consumption habits representing lifestyles and demographic background variables influence on consumption and willingness to consume more on tourism. The first refers to actual behaviour although here people's own perceptions of their consumption compared to other people was used as an indicator. The latter refers to desire to travel in an ideal situation where there would be no financial restrictions.
The fundamental aim of this study was to utilise empirical survey data and find rules behind consumer behaviour when tourism consumption is concerned. The effects of background variables and consumption habits were examined.
According to postmodern theories, the importance of traditional structures has diminished. Here, empirical results were mirrored to these often exaggerated theories (see e.g . Agger 1991;Lash 1990, 2;Ritzer 1999, 72) and it was assumed that social background has still, at least to some extent, effect on how people behave. It might however have more effect on actual behaviour than consumption preferences. 43 Tourism consumption Practically all the theories of postmodernity share the common view about receded importance of former dividing elements such as the class and the nation (e.g . Featherstone 1991;Lash 1995;Miles 1998;Toivonen 1992Toivonen , 1997Urry 1995). this context, Lyotard (1985) speaks of metanarratives that have crumbled away and that, as Mustonen (2006, 69) suggests, have been replaced by small narratives. At the same time, even though these trends are visible, old structures still exist. Contrary to the ideas of postmodern social systems and postmodern consuming styles, consumption has traditionally been structured by social differences between consumers (e.g. Bocock 1993, 21-22). Numerous empirical studies show that traditional factors still affect people's behaviour (Räsänen 2003, also when consumption of tourism is concerned (Honkanen 2004;Räsänen 2000;Toivonen 2001;Mustonen 2003Mustonen , 2006. There are two dimensions that must be taken into account when tourism consumption is concerned. On the one hand, consumption requires financial capabilities. On the other hand, there must be a desire to consume. Modernist thinking emphasises the first dimension. According to this view structural factors such as monetary restrictions and demographics are still important predictors. Postmodernist point of view instead plays with fantasies and finally claims that consumption choices and desires are in fact results of conscious choices. Thus, those who have adopted the postmodern approach, tend to be interested more in discourses of touristic practices instead of examining the demographics (e.g. Moore 2002) and other observable facts. Taking this further, according to extreme postmodern view, being an "adventure traveller" is possible even if the adventure is experienced at home (compare to Urry 1990).
Postmodern features of tourism have been relatively widely discussed. It has even been said that a "tourist" can be seen as a metaphor of a postmodern actor (Bauman 1993, 240-244;Featherstone 1995, 126;Jokinen & Veijola 1997). This is connected to the fact that tourism as a phenomenon has changed a lot during past decades and these changes are not solely due to demographic changes. Postmodern ideas play certain role in the discussions and new divisions and barriers must be taken into account (compare to Bauman 1996, 203,294;Lash 1995, 176;Scott 2002, 23). According to Bocock (1993, 27-28, also Mackay 1997 these new divisions are based on identity formed largely by consumption rather than on traditional social factors. Consumers are said to be living in the consumer society, where consumption is the main factor behind lifestyles and culture (Miles 1998). Lifestyle is something that is absorbed through socialization process but still, it can also be chosen, for example through consumption behaviour. According to Miles (2002, 137) lifestyle is a material expression of person's identity.
When tourism is under scrutiny, the emphasis on the individualised experiences is a central sign of the increasing meaning of lifestyles. Voase (2007) suggests that the most remarkable changes in "individualised" tourism are connected to the tendency to do more frequent visits to rural, urban and in particular to cultural destinations. In addition to these changes, duration and seasonal diversity have increased. Travel industries have also adapted their services to meet the increasing demand for independent travelling, customised package holidays and new and unusual destinations. At the same time, postmodern consumerism can be seen to have penetrated into totally new areas such as museums (Kirchberg 2000).
After all, social structures of these changes lie deeper than postmodern theories claim. People are looking for personal experiences and ways to fulfil their own desires, that are often created socially. Tourism is a social phenomenon and tourists search for communal experiences. Thus especially in the context of tourism, consumption choices depend on social environment that in western societies is very consumption driven (Sharpley 2002, 307-311). Bell's (1974) seminal discussion of post-industrial society was not based on consumption per se but rather on increased importance of service sector. Tourism is an excellent example of this. Although tourism is closely connected to all the sectors of economy, the service dimension is emphasised. In spite of necessary means of transport, tourism is largely based on services. The birth of large-scale tourism coincided with the more general change to which the growth of service sector perfectly joins.
Similarly occurred growth of immaterial consumption and interest in postmaterial issues is also considered one important feature of postmodern (see e.g. Inglehart 1977Inglehart , 1997Scarbrough 1995). Whilst in modern societies material val-ues are traditionally more important, in postmodern societies people are said to be more interested in personal quality of life. Instead of escaping the everyday life people search for themselves (see Smith & Kelly 2006).
It is only recently that the importance of tourism as a specific part of consumption has been fully understood (Sharpley 2002, 310). In modern societies, tourism was regarded as a separate part of social activities. Tourism consumption created divisions and reflected the social differences between consumers (ibid, p. 311). Now tourism, too, must be focused through a new postmodern perspective. Tourism and touristic practices have rapidly gained importance in the contemporary world during the past two or three decades. In addition to this, due to vertical de-differentiation (Uriely 1997;Uriely et al. 2003; see also : Lash & Urry 1994;Urry 1990Urry , 1995 touristic practices can be found in other fields of consumption as well. Thus, people may behave like tourists even if staying still. Tourism, after all, requires movement from one place to another (World Tourism Organization 1995). Although Munt (1994) has stated that "tourism is everything and everything is tourism", this "touristization" of the society does not mean that all the people are tourists (Sharpley 2002, 311). First of all, in the "postmodern" western consumer societies, everyone does not want to travel. In addition to this, travelling habits may change remarkably according to one's phase of life (e.g. Hall 2005, 77-98). According to Pearce (Pearce 2005;Pearce & Lee 2005), each and every tourist has a travel career; people's motivations tend to change according to their life span and accumulated travel experience.
People have now more leisure time and resources than ever before (see Harrison 2003, 28). Consumption of tourism is not compulsory and large scale international tourism is mainly possible only in wealthier nations. Therefore the discussion around postmodern consumption concerns primarily western countries which generate the bulk of tourist flows globally. People from the first world can, in principle, travel anywhere. Due to changes in consumer culture the kind of travelling behaviour which earlier was possible only to real forerunners, the allocentrics (Plog 1974), has become more common in today's world.
When contemporary tourism is discussed, it is difficult to think of a truly alternative form of tourism. For example, backpackers or so-called ecotourists, who are often consid-ered alternative tourists, are actually representing just another dimension of mass tourism (Cohen 1995;Mustonen 2005;Ryan et al. 2003;Scheyvens 2002;Wearing et al. 2002). Taking this further, those people who do not travel at all even if they could, are the most alternative "tourists".
Postmodern tourists do not necessarily even want to be different. The individual's possibilities to do something do not necessarily lead to similar behaviour.

Empirical study
The empirical part of the study will examine answers to three research questions that can be presented in the form of hypotheses. First hypothesis claims that both lifestyle issues and sociodemographic background should be taken into account when examining the factors behind tourism consumption. Secondly, it is assumed that postmodern features are most likely to more visible in the case of consumption desires than in the case of actual consumption. In addition to these two hypotheses, the empirical significance of the so called postmodernisation hypothesis will be tested. According to the hypothesis the effects of life-style issues should be increasing while at the same time effects of social background determinants should be decreasing.

Data and variables
The empirical part of this study was based on two sur- In this article two different dimensions of tourism were looked into. Tourism consumption referred to respondents' evaluation of their own consumption of tourism comparing to the average consumer. The question behind the variable was: "How do you compare your consumer habits in relation to average consumer in the case of tourism?" The answers were measured using likert scale (1-5) where value 1 meant "a lot of less than average consumer" and value 5 "a lot of more".
Other dimension was desire to travel more in case of no monetary restrictions. The question was: "Would you travel more if you can afford?" Also here answers were measured using likert scale (1-5) where value 1 meant "not at all" and value 5 "a lot of more".
These two dimensions are different sides of the same coin but this does not mean that respondents should give parallel answers to the questions. For example someone who according to her/himself travels less than regular consumer, is not necessarily willing to travel more. She/he may find other ways of consumption more appealing.
Thus, tourism consumption can be analysed in many different ways. The variables that are used in this paper, are not expected to cover all the aspects of tourism consumption. By using just these variables, it is not possible, for example, to say anything about how much money people spend on tourism comparing to other consumer goods.
Usually quantitative research is concentrated on tourism consumption that can be observed. The questions like "how many trips have you made during the last year?" are common. From qualitative perspective, feelings and desires towards tourism are often more interesting than tourism consumption. When lifestyles are under the scoop, both actual tourism consumption and desires must be equally taken into account. The differences between these two are most probably due to the differences between traditional social background factors whilst postmodern "freedom of choice" may be seen through lifestyles and willingness to travel. Table 1.

Distributions of dependent variables are shown in
When tourism consumption was concerned, lower categories (1-3) were emphasised. Thus according to themselves, respondents generally spend less than on average on tourism. This was not a surprise because people often tend to underestimate their consumption (Wilska 2002, 199). Nevertheless, respondents would like to travel more. When desire to travel more was examined, the share of respondents in categories 3-5 was remarkably greater. By interpreting the table, also some changes between the years can be found. These Following statistical analysis deepens these descriptive results. The main research question was how consumption attitudes and on the other hand demographic background influence on consumption on tourism (tourism consumption) and willingness to consume more on tourism (desire to travel more). The purpose was to examine principles behind these two dimensions of tourism consumption. The central question was: Which one, consumption attitudes (lifestyles) or demographic background, a better explanation?
Based on the above discussion of lifestyles it was assumed that traditional social background variables would have more effect on tourism consumption than on desire to travel. Accordingly when desire to travel was concerned; it was assumed that the importance of consumption styles would be greater. Does examination of the data give answers to the question of what kind of consumers are most likely to be tourists? On the other hand, are the demographic background factors connected to people's desire to travel more? In addition to trying to find answers to these main questions, also changes were examined. It was also discussed whether occurred changes can be explained by increased importance of lifestyles and postmodern features often connected to them. At the same time it was examined whether changes in tourism consumption consumption and desire to consume more on tourism are due to the changes in respondents distributions to the different sociodemographic categories and on the other hand to the changes in consumption patterns.

Descriptive analysis
Taking lifestyle into consideration in tourism studies is not a new idea. As early as in 1980's a few studies in which tourism was examined through lifestyle were published (e.g. Bernard 1988;Mayo & Jarvis 1981). Nevertheless, usually lifestyles have been connected to the way how tourists behave and consume during their trip, and on the other hand to the destination choices (see Dolnicar et al. 1998). Unlike in the case of the widely used typologies like The Eurostyle System (ibid) and 'VALS' (e.g. McIntosh et al. 1995, 426-467;Shih 1986;Skidmore & Pyszka 1987), in this study consumption patters have been created using more limited data. Here the concept of lifestyle was based on consumption habits or rather the attitudes of consumers. The aim Table 1 Cross-tabulation for tourism consumption and desire to travel more (%).  (2341)  2004 12,3 7,9 38,3 13,3 28,2 100 (3506) a) = 15,37, df = 4, p. = ,004 b) = 502,39, df = 4, p. = ,000 1=totally disagree, 5=totally agree was to examine how consumption attitudes are connected to tourism consumption. Although consumption attitudes and habits represent only one dimension of lifestyle, according to postmodern theories, they are amongst the most important and dominant ones (Featherstone 1991;Bocock 1993).
However, it is important to notice Veal's (2000) review which points out the multitude of lifestyle concepts.
In Table 2  were found. They together explained 51,5 % of the variance.
The first factor was named as environment due to a clear emphasis of the questions concerning the environment, issues around food and on the other hand critique against consumption based lifestyle. Among the factors of this study, this factor represents best the ethical dimension of consumption.
In the second factor, pleasure, the questions were linked with restaurants and impulse purchases. The third factor, saving, describes goal-directed activity whilst the fifth factor (price) represents the group of consumers to whom the main intention is to consume commodities which are as cheap as possible. For the people who represent the price-factor, saving for some particular goal seems not to be important. This is the main difference between these two. The questions concerning fashion, appearance and interior decoration loaded most strongly to the fourth factor, which was accordingly named as fashion. For fashion-oriented people consumption habits are connected to the aesthetical values. The last factor, culture, was composed around the questions about "high class" consumption like classical music and wines.
For the forthcoming analysis, six variables named accordingly with the factors, were built using the factor scores. Fac-  Table 4. The method was chosen because comparing to ANOVA, it was possible to add continuous variables as covariates (see Tabachnick & Fidell 2001, 275).
Totally five variables were chosen to represent the sociodemographic background of the respondents (table 3)  18 got a multiplier 0,5. This method utilised here is widely used in official statistics (see Statistics Finland 2007). After weighting procedure, the variable was divided into five categories.
Residential area was divided into two classes, cities and countryside. The idea behind the variable was that those living in countryside were expected to be less postmodern than those living in cities. The variable defining the gender of the respondent was excluded from the analysis because surpris-ingly it had any effects on dependent variables. Other excluded variable was education. It would have had some minor effect to desire to travel more, but due to problem of multicollinearity with the variable income, it was also excluded from the analysis. Income turned out to be a significantly better explanant.

Explanatory analysis
In Table 4 the effects of socio-demographic background variables and consumption styles on tourism consumption and desire to travel were examined using the data from the year 1999. In the first two models, the dependent variables, tourism consumption (Tc) and desire to travel more (Dt) were explained using only socio-demographic variables as independent variables. It can be seen that both the variables, identification to a certain social class and income, explain tourism consumption quite well. Respondents who had identified themselves to the higher social classes were more likely to think that they travel more than on average. Also the effect of income level was similar. These both variables are traditionally linked with economic resources and thus the result was expected. Also variables concerning the number of children and residential area were statistically significant but their Partial Eta squared -values were small. Socio-demographic variables altogether explained 16,7 % of the variance, which was quite a remarkable result.
In the case of another dependent variable, desire to travel more, the effects of socio-demographic variables were generally weaker. However, some interesting exceptions could despite these few significant findings, the social background variables explained only 5,9 % of the variance. This is remarkably less than in the case of tourism consumption.
In the third and fourth model, only the factors, which here indicate the consumption habits, were analysed. In these models, the difference between tourism consumption and desire to travel more was considerably smaller than in the first two models. Nevertheless, independent variables were still better explanants in the case of tourism consumption, but it must be noticed that when desire to travel more was concerned, factors explained 9,3 % of the variance, which is remarkably more than in the second model.  (Poon 1993) is exaggerated (Honkanen 2004). Environmentally friendly tourism is often associated with these. Environmental consciousness and ethical issues that are often connected to postmodern (e.g. Bauman 1996; Beck 1995) do not necessarily imply the growth of "good" or more sustainable ethical behaviour (compare to Sharpley 2002, 305-306). It must be also mentioned that contrary to postmodern ideas, from 1992 to 1999 environmental concern amongst Finnish citizens had diminished ( tourism consumption than on desire to travel more. However it must be noticed that also consumption styles (factors) affected strongly on tourism consumption and on the other hand highest income and age categories differed from the others on the case of desire to travel more. It is also interesting to notice the direction of the effect of age variable in the case of tourism consumption. Younger respondents seem to consider their travelling behaviour lesser to older respondents. In contemporary "postmodern" consumer societies, especially younger people are expected to travel and even though in a real life they would travel a lot, they may think that other people travel more. For example media plays a major role when these pressures are born. This same phenomenon can be a reason why older people think that they travel more than on average. The older cohorts have absorbed travelling as a part of their lifestyle (Mustonen et al. 2004;Toivonen 2001) and therefore they might think that they travel more than on average. Further investigation of this interesting finding could be possible for example by testing the significance of factors separately in all age groups.
Nevertheless, this is beyond the scope of this study.
As a conclusion, it can be stated that tourism consumption is strongly connected to the resources and class status but also to consumption styles, or lifestyles, whilst desire to consume more on tourism is connected mainly to the lifestyle de- Similar models for the year 2004 can be seen in Table 5.
The results concerning the sociodemographic variables were quite parallel with the ones from the year 1999; economic resources affected positively on tourism consumption. The greatest difference could be found in the effect of income categories, which in the case of desire to travel more were significant in 1999. When the data from 2004 were examined, there were no differences between respondents in the lowest and highest categories. Respondents from the lowest category wanted to travel as much as respondents from the highest category. This was a remarkable finding considering that the less wealthy people consider their tourism consumption clearly lesser when compared to average consumer.
The lack of economic resources did not seem to affect on desire to travel more. This finding does not follow Bourdieu's (1984) idea that the taste of lower class citizens drives people to choose the necessary. Thus the observation of the influence of socio-demographic class-variables on tourism consumption must be considered, at least to some extent, separated from the taste. One reason behind this interesting ob-servation could be the very important cultural role of tourism in the Finnish society. (e.g. Selänniemi 1996;Mustonen et al. 2004). Of course it must be noticed that the question in the questionnaire concerned only desire to consume more on tourism in general. Bourdiean taste differences would have been revealed within more detailed data.
The influence of consumption habits remained almost unchanged. There were some slight differences in R-squared values and some parameter estimates had changed a little.
However it can be stated that the effects have been somewhat static. This is in line with the assumption that the postmodernization process, which is here generalised to mean increasing importance of lifestyle determinants and decreasing importance of social background, has become somewhat stable. Economic resources and social class together with consumption habits influence on tourism consumption whilst in the case of desire to travel more, consumption factors are better explanants and the effect of traditional background factors is less significant.
Finally both the data-sets were combined to create a pattern where the occurred changes could be interpreted. The results of the analysis of covariance (ANCOVA) are seen in Table 6. The analysis was conducted by utilising all the independent variables shown in Tables 4 and 5. The variable indicating the period (year) was also added to the matrix. Only F-values, partial Eta squared -values and essential parameter estimates are presented because of economic reasons. Comparing to the analyses conducted before, the additional aim was to test the interactions between the year and background variables. Now in the case of tourism consumption, the year was not amongst the significant independent variables (model 1). This is interesting in the respect that in the simple crosstabulation presented in Table 1 the differences of tourism consumption between the years were significant. When income variable was dropped out of the analysis, the parameter estimate of year happened to be significant (the model is not shown here). In year 2004 more responders seemed to belong in middle and higher income categories. This is probably the reason why more people consider their tourism consumption equal to average consumer. There were no significant interactions between the year and other variables.
When desire to consume more on tourism was consid- The empirical analysis strengthened the hypothesis. Consumption habits had effect on both, tourism consumption and desire to travel more. In addition to this, the effects of social class (class-identification) and income were remarkably strong in the case of tourism consumption. When desire to travel more was concerned, the effect of consumption habits was stronger than the effect of sociodemographic variables. However, when either of these two sides of tourism consumption is concerned, both lifestyle issues and sociodemographic factors should be taken into account. The assumed postmodern change has not totally wiped away the significance of modern structures.
The influence of background variables was examined using the data from years 1999 and 2004. Although some minor differences were found, the results, in general, were quite parallel. Thus, it cannot be assumed that travelling habits would be now more postmodern than, let's say, five years ago. Instead, the "postmodern change" might have happened earlier. It is also worth mentioning that socio-demographic variables might influence on travelling habits through consumption patterns. However, these kind of causal examinations were beyond the scope of this study.
Compared to the year 1999, in 2004 more respondents considered their tourism consumption equal to average consumer. This is due to increased income levels. In the case of desire to travel more, the direction of the change was the opposite. The desire to travel more has diminished. It seemed that in 2004 the respondents that represented the middle income categories were less willing to consume in tourism than in 1999. In the literature the greatness of tourism is often exaggerated and it is assumed that together with growing international tourism also the desire the travel must rise. The fact that tourism has grown globally (World Tourism Organization 2008) can, however, be explained for example by growing tourism demand in other than the most common tourism generating countries. On the other hand, travelling is now cheaper than before and domestic tourism might have lost its share to the international tourism. However, it must be stated that the rise of tourism presented in statistics is far beyond the change that was observed in this study.
Consumption of tourism competes with the other discretionary forms of consumption. In postmodern consumption cultures people are sometimes expected to travel. However, be considered which one, tourism consumption or desire to travel more, is more interesting and worth studying in the particular case. These two dimensions are connected to each others but even in the postmodern world, desire to travel does not necessarily realise into action.