summary these articals

Pergamon

Annals of Tourism Research, Vol. 22, No. 3, pp. 535-552, 1995

Copyright 0 1995 Elsevier Science Ltd Printed in the USA. All rights reserved

0160-7383/95 $9.50 + .OO

0160-7383(95)00004-6

TRAVEL LIFE CYCLE

Martin Oppermann

University of Nevada-Las Vegas, USA

Abstract: This article surveys changing tourism patterns along three time horizons: the last three decades, across the life cycle, and between successive generations. Based on empirical longitudi- nal data, the study suggests that the patterns and destination choice have changed with respect to all three perspectives. Younger generations gain different experiences as compared to previous generations and are likely to have different tourism patterns in later life stages. This has important implications for destinations that rely on the elderly market. Using longitudinal approaches, insights can be gained in changing destination choice across the life cycle and between successive generations, and it can complement traditional cross-sectional studies. Keywords: Life cycle, tourism pattern, destination choice, generation differences, longitudinal study.

R&urn& Le cycle de vie des voyages. Cet article examine les CaractCristiques changeantes du tourisme le long de trois horizons temporels: les trente dernikres an&es, ?I wavers le cycle de vie et entre g&&rations successives. B&e sur des don&es longitudinales empiriques, cette recherche suggere que les caracttristiques et choix de destination ant changC pour les trois perspectives. Les jeunes g&&rations acquierent des expCriences diffkrentes de celles des gCntrations prCcCdentes et auront done des comportements diff&ents plus tard dans la vie. Ceci a des implications impor- tantes pour les destinations qui dtpendent du march6 de troisi&ne Lge. En utilisant des mtthodes longitudinales, on peut mieux comprendre les choix changeants de destination pendant le cycle de vie et entre g&&rations successives; ces mCthodes peuvent complCter les Ctudes traditionnelles par Cchantillon. M&s-cl&: cycle de vie, types de tourisme, choix de destination, diff&ences entre gtnCrations, ttude longitudinale.

INTRODUCTION

In two separate accounts of tourism and recreation research, Dann, Nash and Pearce (1988) in their review of adopted methodology in tourism research and Mitchell and Smith (1989) in their overview ‘of the geography of recreation and tourism noted a lack of temporal studies:

Attention to the temporal dimension of tourism would seem advis- able, and the statistical treatment of time based data (e.g., time series analysis) could be an exciting direction of future research (Dann, Nash and Pearce 1988:25).

Related to replication is the problem of a lack of temporal studies. The examination of recreation phenomenon through time is almost

Martin Oppermann (Department of Tourism and Convention Administration, University of Nevada-Las Vegas, Las Vegas NV 89154-6023, USA. E-mail: “[email protected]”) cur- rently pursues his second Ph.D. in hotel administration. He has worked as Marketing Manager (GoHolidays), Research Associate (UniversitHt Tiibingen), and Research Director (Institut fiir Tourismus und Marktforschung International). His research interests include intranational tour- ist flows, Third World, rural and convention tourism, and destination images.

535

536 TRAVEL LIFE CYCLE

impossible because time-series data are seldom collected and made available to interested researchers (Mitchell and Smith 1989:401).

According to Zimmermann (1982), there is not only one temporal dimension, but three: period, life cycle, and cohort effects. Period effects refer to annual changes and/or specific events that influence tourism behavior (e.g., Gulf War). Life cycle adverts to variations of the behavior over the life-span of an individual, most often associated with changes in family structure. Cohort effects relate to behavior patterns that are influenced by some unique aspect of that cohort, such as its size or historic background (e.g., baby boom generation). In tourism research, all three dimensions have been addressed, albeit to a different degree. Annual changes in tourist arrivals are a common concern in all countries and are at the heart of the widely used “destina- tion area life cycle” concept (e.g., Butler 1980; Din 1992; Meyer- Arendt 1985; Weaver 1990). Family life cycle applications in tourism and leisure are comparatively few (e.g., Bojanic 1992; Cosenza and Davis 1981; Lawson 1991; Rapoport and Rapoport 1975). Studies of generational or cohort differences in tourism pattern are rare (e.g., Becker 1992).

This study discusses the changing travel patterns of German resi- dents with respect to all three dimensions. The variables used are travel frequency and intensity, destination choice, and succession. This re- search aims to give insights into the possibilities and problems of longi- tudinal tourism studies; examine destination succession patterns over the last 30 years; analyze the influence of the individual’s stage in the life cycle on travel to destinations; and discuss differences between successive generations with respect to their destinations.

THEORETICAL FRAMEWORKS

Cycle theory is a commonly used concept in a variety of research fields such as economics, marketing, demography, urban studies, and geography (Doling 1976; Roberts 1991; Vernon 1966; Wagner 1989; Wells and Gubar 1966). However, one needs to distinguish between two different life cycle concepts, namely studies related to the product life cycle (product understood as a generic term) and research on the family life cycle. The basic reasoning of the former is that a product or tourism destination moves through stages of inception, growth, maturity, stagnation, and eventual decline. Depending on the product and the locality, this curve may be steeper or longer but an eventual decline is assumed. The product life cycle of tourism resorts was intro- duced by Butler (1980) in his seminal article “The Concept of a Tourist Area Cycle of Evolution: Implications for Management of Resources.” According to Mitchell and Smith (1989), however, Christaller (1963) and Stanslield (1978) f orwarded similar notions. The destination area life cycle (DALC) describes the evolution of resorts/destinations from involvement over exploration and development to consolidation and stagnation. Eventually the tourism area faces declining tourist num- bers if it is not temporarily saved by rejuvenation. Associated with changes in visitor numbers are changes in visitor type, organization of

MARTIN OPPERMANN 537

the tourism industry, marketing efforts, and local involvement. Al- though some aspects of the underlying product cycle concept have been mentioned by others in earlier publications (Christaller 1963; Cohen 1972; Plog 1973; Stansfield 1978), Butler’s contribution is the most widely recognized (Wall 1982). S ince then, an ever-growing number of studies have applied the concept to destination areas of various sizes (Benedetto and Bojanic 1993; Cooper and Jackson 1989; Dahms 1991; Din 1992; Foster and Murphy 1991; France 1991; Getz 1992; Hay- wood 1985; Hinch and Butler 1988; Hovinen 1981; Ioannides 1992; Meyer-Arendt 1985; Strapp 1988; Weaver 1990). The main trust of its critics is the concept’s too general nature (e.g., Dhalla and Yuspeh; Hart, Casserly and Lawless 1984; Haywood 1985; Wall 1982). It does not allow for a placement of a resort in a specific development stage. Hence, it is more a descriptive rather than a prescriptive tool (Cooper and Jackson 1989; France 1991). Nonetheless, the real contribution of Butler’s (1980) model is in its general approach and the potential for utilization and testing in different study areas. Attempts to build on theory are rare in tourism studies (Smith 1990). More often than not contributions are single case studies or models that are never tested by other researchers and/or applied to different areas (Mitchell and Smith 1989).

The distinctive category, family life cycle (FLC), explains differ- ences in individuals’ behavior at varying stages of their life and particu- larly their family life. For example, residential choice behavior has been linked to the family life cycle (Doling 1976). One of the first applications of the FLC to the field of leisure and tourism was by Rapoport and Rapoport (1975) who discussed the changing context of leisure activities over the FLC. According to Murphy and Staples (1979), FLC studies date back to the 1930s. Changing family struc- tures cast serious doubt on the future utility of this concept because the “normal” family of two adults and one or more children is presently an “abnormality” (Stapleton 1980). Changing demographic trends result in classification problems when the traditional family life cycle scheme is applied. In a study on tourists in New Zealand, Lawson (1991) encountered 40% unclassifiable cases. Although all sorts of modifica- tions have been proposed (e.g., Bojanic 1992; Murphy and Staples 1979), the increasing number of categories- for example, Zimmer- mann (1982) used 39 stages in five household types-made a simple concept too complex to be attractive. Despite the modifications, a substantial number of cases still remained unclassified, ranging from 10% (Bojanic 1992) to 17% (Murphy and Staples 1979). In addition, the application of the FLC concept in tourism faces a second challenge. Travel party size frequently does not coincide with household size (e.g., singles traveling with friends or married individuals going alone). Lawson (199 1) encountered 12.3 % of such cases in his study. Bojanic (1992) circumvented this problem in his household survey since his questions did not refer to any specific vacation trip.

Overall the number of FLC applications in tourism research are few as compared to the DALC concept. One reason for the lack of FLC studies in tourism is the difficulty in operationalizing or standardizing the family stage categories as well as collecting useful data. Another

538 TRAVEL LIFE CYCLE

problem is that as soon as additional categories, such as single parents, are introduced, the cyclical nature of the concept where one stage follows the other is disturbed. Several categories existing parallel to each other refute the underlying concept of unilinear change making the FLC concept less appealing than the DALC concept. Nevertheless, the potential of the FLC concept remains to be explored, for example as a tool for comparing travel behaviors of different nationalities.

An area of inquiry that both DALC and FLC studies have neglected is the integration of the tourist perspective with the destination (i.e., resort). In referring to tourists who first “discover” or visit a destina- tion, some related studies regard them as “allocentrics” (Plog 1973) or “explorers” (Cohen 1972). It is usually assumed that pioneering tourists start the development process in most locations, apart from the planned resort (Butler 1980). In other terms, they may be called trend- setters because they set the trend for the future and other tourists follow. This dynamic element was explicitly mentioned by Plog (1987) who asserted that allocentrics pick up new products first and then “introduce the products to the near-allocentrics, and then the near- allocentrics will “pass” the products on to the midcentrics, . . . ” (cited in Pearce 1993:124).

In a discussion of tourism motivation research, Pearce (1993) noted that motivation theory should be flexible enough to account for indi- vidual changes across the life-span. He also introduced the concept of a “travel career”:

Like a career at work, people may start at different levels, they are likely to change levels during their life-cycle and they can be pre- vented from moving by money, health or other people. They may also retire from their travel career or not take holidays at all and therefore not be a part of the system (Pearce 1993: 121).

Although Pearce’s work makes specific references to visitation of theme parks, it may also be adapted to general pleasure travel. The basic points are, in the view of this author, that travel patterns change as the individual moves through her/his life-span and/or family career, that tourism destination choice and patterns vary according to previous experience including the time of childhood, and that external and personal barriers influence annual tourism decisions. Thus, what is required is a longitudinal study inquiring into the tourists’ travel pat- terns over the tourists’ life-span. Adopting such an approach one can derive an individual’s travel career and relate it with variables such as age, family status, and previous tourism experience. This echoes Wilkinson’s call for research over long time periods, “including longitu- dinal travel patterns of individual tourists (Wilkinson 1987:144). Ex- amples of such a discussion were given by Mitchell (1983, 1984), who analyzed the changing leisure activities and their spatial distribution over the life cycle of an individual. He showed how leisure activity space varies during different life cycle stages and how different activi- ties take place at different distances from home.

The evolving “travel style” of the individual tourist or a category of travelers is mentioned by Becker (1992) as well. He argues that a

MARTIN OPPERMANN 539

longitudinal analysis of travel patterns provides for more accurate fore- casts of future patterns and intentions. One needs to distinguish such longitudinal studies of life long travel patterns (with all data collected in one survey) from the more commonly known longitudinal studies that are basically time-series data gathered in repeated data collections procedures. The latter is normally hampered by a lack of data cohe- siveness owing to changes in the questionnaire and the questions asked (Mitchell and Smith 1989). Time-series also does not allow for an analysis of an individual travel career because each year different re- spondents are chosen.

GERMAN TRAVEL LIFE CYCLE

To attain the objectives mentioned above, data were required on life-long tourism patterns of individual persons. Focusing on holiday travel, data such as holiday destination(s), mode of transportation, and length of stay were needed. In addition, information on the respon- dents family career (i.e., year of marriage, birth of children) was needed to relate travel patterns with life cycle stage. The data on life-long pattern allows for the analysis of destination succession pat- terns over the last 30 years. Further, to investigate differences between successive generations, respondents had to include a wide variety of age groups. To fulfil these requirements, this pilot survey was designed as a general household survey. In two stages, questionnaires were distributed to 2,000 households at two residential areas in the cities of Tiibingen and Reutlingen (Germany), along with an accompanying letter stating the purpose of the study. Although the two residential areas were selected based on the researcher’s familiarity with them, there was no preselection of individual households to be interviewed. Questionnaires were delivered by foot to all households, eliminating a possible bias due to nonlisting in sources such as telephone books. Both residential areas are planned suburbs built primarily in the 1960s and 1970s. The wide variety of housing (i.e., one family houses, double and multistoried buildings) and housing ownership (i.e., ownership of houses, apartments, leased apartments, and subsidized housing) were considered beneficial in attaining a representation of a wide range of socioeconomic classes. In fact, professions ranged from teacher to architect, gardener to retiree, housewife to public servant, and student to professor. Initially 1,000 questionnaires were distributed in Tub- ingen. The low response rate called for the distribution of another 1,000 in anticipation of higher returns and to detect if the low return rate was location specific. Disappointingly, there were even fewer re- sponses from the households in Reutlingen. On both occasions, a stamped return envelope was enclosed.

A total of only 138 ( or 6.9%) questionnaires were returned. Four- teen answers were dismissed because they were incomplete (e.g., year of birth was missing). The analysis is, therefore, based on 124 (or 6.2 W) completed questionnaires. The very low return rate was proba- bly caused by several factors. First, surveys in Germany are still ham- pered by what may be called the “census-” or “glasnost-syndrome.” In 1987, a highly criticized and controversial census took place after a

540 TRAVEL LIFE CYCLE

previous aborted attempt in 1985 because of legal problems. A consid- erable number of people did not participate in the 1987 census, many of whom were facing legal action by the German government. The publicity and controversy surrounding the population census led to suspicion and scepticism by the public for all other surveys. Similar tendencies have been observed by this author in other studies in Ger- many as well. Symptomatic of this was the reaction of one household who contacted the state Datenschutzbeauftragten (independent state ofli- cial responsible for supervision of data safety) to assure the legality of this survey.

Second, the nature of this questionnaire needed more time than other surveys. Anyone who has attempted to reconstruct all holiday trips over his/her life span knows how much time this requires. Becker (1992) achieved a response rate of 45% in his survey. His study, however, was based on personal interviews with an initial letter send to the target group and a phone call to agree on the time of interview. In addition, his target group included a large number of retired per- sons who may have had more time and, therefore, may have been more inclined to answer the questions than the average working popu- lation.

This study included questions about place and date of birth, other places of residence, occupation of respondent, spouse and parents, and number and age of children. The respondents were asked to name all destination areas visited for each year on trips that lasted for at least three nights. This time limit was used to avoid the inclusion of weekend trips and to focus on vacation trips. In addition, weekend trips were considered too short and too frequent as to be recalled accurately. In case the respondents had traveled several times in one year, which happened frequently in the 198Os, the trip with the longest duration was considered the main vacation. Trips with the second and third longest duration were defined as secondary and tertiary trips respec- tively. Besides the month and year, the survey asked for travel pur- pose, length of each trip, number of overnight destinations, main transportation and accommodation types used, and number of accom- panying travelers. These variables were included to obtain insights into the actual travel pattern. These variables were presumed to vary according to the life cycle, year of travel, and the destination area visited. Some responses were very specific (e.g., specific dates of travel and places visited) indicating the use of travel diaries. Others were more general, for example, giving Italy as tourism destination. This analysis uses the more general level (e.g., countries in Europe and continents outside Europe).

Data Analysis and Results

Data analysis is hampered by the very low return rate (6.2 %), small sample size (124 respondents), lack of randomness, and the fact that no information exists on the nonrespondents. Confounding the problem further is the assumption that the respondents remembered all trips correctly. It is most likely, however, that vacations in more recent years were recalled better than those 20 years ago. Lack of appropriate

MARTIN OPPERMANN 541

data limited the evaluation to the three decades from 1963 to 1992. Since the respondents were a mix of individuals from 18 to 82 years old at the time of survey, most data were available for the last decade with fewer persons being old enough to have traveled between 1963 and 1972. Since the responses are in no way representative for the German population, or even the Reutlingen or Tubingen population for that matter, only nonparametric data analysis methods are utilized. The results should be considered as initial findings that require further thorough investigation and systematic testing.

The data are analyzed on two levels. First, the responses are dis- cussed on a “respondent level.” This means that the variables are pre- sented with the “total” column representing the number of respondents (N= 124). For this level, data are only used on a decade basis (1963- 72, 1973-82, 1983-92). F or a meaningful comparison between the respondents, this required that information had to be available for the entire decade and not for a few years. Thus, if one respondent had given his/her travel destinations for the years 1975 until 1992, then only the years 1983 to 1992 were considered.

On the “individual vacation level,” all trips are used for which infor- mation on destinations, year and month of trip, travel purpose, length of stay, choice of accommodation and transportation types, and num- ber of travel companions was made available. In this part, it did not matter if a decade was “complete.” Hence, from the previous example, all trips between 1975 and 1992 would have been included as long as all the data were “complete.” Some 3,3 18 individual vacation trips are included in this analysis.

The total of 124 respondents comprised 60 females and 64 males. This represents a male bias since females make up 5 1.7 % of the Ger- man population (Voit 1992). The average age was 47 years with 34% of the respondents being younger than 35, 27% between 35 and 55, 23 % between 55 and 65, and some 16% were above 65. These figures compare to 45 % , 28%) 12 % and 15 % respectively. Thus, there is an overrepresentation of respondents around retirement age (55-65 years) and a smaller than average share of younger respondents (below 35). This deviation is partly due to the survey approach. Not all age groups are equally represented in households with younger persons and espe- cially children more likely to live in multiperson households. On the other hand, an increasing number of older persons live in one-person households and are, therefore, more likely to be oversampled in house- hold surveys. About 40% had no children and another 23% had only one. Fifty-five percent were white collar and 20% blue collar workers. The others (25%) were not formally working (i.e., students, house- wife/-husband, retired). The average tourism expenditure for the year 1992 was DM3,200 (US$l,900) per household with a median of DM2,500 (US$l,500), DM 1,5OO/US$900).

or DM2,OOO (US$1,200) per person (median: Th is is higher than the average DM1,3 18 per

person tourism expenditure in a 1991 national survey (STfT 1992) and also higher than the DM1,468 spending by West-German travelers. However, this national survey asked for expenditure on the primary vacation trip while this survey included all vacation expenditures.

One important aspect in travel pattern is travel frequency. The

542 TRAVEL LIFE CYCLE

Table 1. Total Number of Trips per Decade (in %)

1963-72 1973-82 1983-92 Number of Trips (N = 69) (N = 103) (N = 124)

o-9 65.2 43.7 26.6 10-19 33.3 42.7 50.8 20-30 1.4 13.6 22.6 Average Number of Trips 7.5 10.7 13.9

Primary 6.3 7.5 8.3 Secondary 1.1 2.6 4.0 Tertiary 0.1 0.7 1.6

analysis of this variable with respect to the three decades reveals an increasing inclination to travel. In the years 1963-72, 65% had fewer than 10 trips (primary, secondary and tertiary). In the years 1983-92, this percentage had decreased to 27 % , while 23 % had ventured on 20 or more trips (Table 1). Thus, a tendency to higher trip frequency was discerned. While the average number of trips per person was only 7.5 in the period 1963-72, it rose to 10.7 (1973-82) and to 13.9 in 1983- 92. Although the recall bias may be a contributing factor to the strong increase, the data are confirmed by secondary data. Annual surveys by the Studienkreis fur Tourismus (Stff) showed that the number of Germans who traveled for at least five days rose from 42% in 1970 to 58% in 1980 and 71% in 1992 (StfT 1993). Becker (1992) noted an increased number of primary, secondary, and tertiary trips during the 1980s as compared to the 1970s. The average number of trips in 1983- 92 consisted of 8.3 primary, 4.0 secondary and 1.6 tertiary trips for a total of 13.9 (Table 1). The increased number of trips in the last period is largely due to the higher incidence of secondary and tertiary trips. The year 1992 was an all time high with respect to primary (90% of all respondents), secondary (56%), and tertiary trips (22%). These fig- ures indicate that the respondents might not be representative of the German population. In 1992, some 7 1 ‘$J had traveled at least once, 16% had gone at least twice, and 3 $% had traveled at least three times for five days or longer (StfT 1993). The cause for the difference be- tween this study and the Stff survey is partly due to diverging defini- tions used by the Stff and this author. The former uses the five-day limit as minimum for inclusion, while this study included trips of three or more nights (or four days). Other reasons are the low return rate of the survey, the high number of nontravelers who did not return their questionnaires, and the biased sample with respect to socioeconomic characteristics. It was shown by the Stff (1993), that travel frequency increases with increasing salary and education.

A cross-tabulation of travel frequency in the 1980s with the 1970s shows a general tendency of respondents (78 ‘%) to increase the number of trips. A few respondents (18%) had traveled more often in the years 1973-82 than in 1983-92. This was partly influenced by age since just 26% of those traveling in 1983-92 were 65 or above in 1992 as com- pared to 19 % of the whole sample.

MARTIN OPPERMANN 543

Destination Choice

Turning to destination choice, the destination areas or countries were categorized into five classes based on the distance from, similarity to, and connectivity with Germany. These five regions are Central Europe (Germany, Switzerland, Austria, and Liechtenstein), Europe (all other West and East European countries, including Russia), North Africa (all non-European mediterranean countries as well as the Ca- nary Islands), North America (United States and Canada), and the Third World (all other countries not included in previous categories). Naturally, the heading Third World is not appropriate since it also includes Japan, Australia, New Zealand, and a wide range of develop- ing countries. Besides Germany, Switzerland, Austria, and Liechten- stein are the only other countries where German is the primary native tongue. In addition, these three countries have a long history of attract- ing German tourists. In particular, the Austrian Alps have been ex- plored, visited, and opened to the general public by German moun- taineer clubs for more than 100 years. This led to one of the earliest published research article on tourism. Rungaldier (1924) analyzed and discussed the geographical distribution of mountain huts in the Aus- trian Alps. Around this core region are the other European countries, in a sense forming a second belt. North African countries were rela- tively early integrated into the German package tour circuit and today Germany is one of the major originating countries for visitors to North Africa and the Eastern Mediterranean from Morocco to Turkey. The large number of transatlantic flights makes North America an accessi- ble destination and reduced prices have resulted in low flight costs.

The majority of all trips were to Central Europe (51 W) and a sub- stantial proportion to the rest of Europe (41%). Hence, all other cate- gories had a minor share with respect to trip generation and these statistics confirm the notion that frequency of travel decreases with increasing distance (Table 2). This distribution differs slightly when the vacations are segregated according to primary, secondary, and tertiary trips. Among the former, Europe was the destination region of as many trips as Central Europe (both 45 %). In the Stff (1993) survey for 1992, the primary trip distribution was 44% to Central Europe, 45% to the rest of Europe, and 11% to places outside Europe. Hence, the survey data resemble closely the nationwide survey of 1992. Gener- ally there is a tendency to choose less distant destinations on secondary and tertiary vacations. In part, this may reflect the definition of pri-

Table 2. Destination Region by Primary, Secondary and Tertiary Trips (in %)

Region Primary Secondary Tertiary Total

Central Europe 44.7 Europe 45.1 North Africa 5.3 North America 1.8 Third World 3.2

x2 = 114.0; df = 8; p = 0.00000.

60.7 69.7 50.6 33.7 27.3 40.8

3.9 2.3 4.7 0.2 0.1 1.2 1.6 0.8 2.6

544 TRAVEL LIFE CYCLE

Table 3. Farthest Destination Region by Decade (in %)

Region 1963-72 1973-82 1983-92

No Trips in Decade 5.8 2.9 1.6 Central Europe 20.3 7.8 5.6 Europe 53.6 54.4 44.4 North Africa 4.3 13.6 18.5 North America 8.7 8.7 6.5 Third World 7.2 12.6 23.4

x2 = 29.86; df = lO;p = 0.002.

mary, secondary, and tertiary destinations applied in this study, namely by length of stay. Since the travel costs to more distant destina- tions tend to be higher, travelers may stay longer to justify higher expenses.

In order to achieve initial insights into destination succession pat- terns, the destination with the longest travel distance for each decade is analyzed. Table 3 reveals a trend among the respondents to travel greater and greater distances over the last 30 years. In the years 1963- 72, less than 21% had visited at least one destination outside Europe. This percentage increased to 38% in the following decade and to 50% in 1983-92, while the percentage of those who took their longest trip within Europe decreased from 54% (73-82) to 44% (83-92). The proportion of persons who visited destinations only in Central Europe remained almost constant between 1973-82 and 1983-92. This indi- cates the existence of a market segment whose members are quite happy with destinations in the vicinity and do not feel the urge to venture outside Central Europe. Some of this trend toward longer trips can be attributed to a change in destination availability over the last few decades. Travel times and costs decreased dramatically making more and more destinations available to the general public.

Stage in Life Cycle

It has been argued that the stage in life cycle influences travel behav- ior (Bojanic 1992; Lawson 1991; Wells and Gubar 1966). To analyze the extent to which life stage affects travel patterns, the respondents were asked to identify some travel parameters for each of their trips. The discussion here is limited to those trips for which all information was available. In addition, it is further limited to vacations by the respondents who were 15 years and older at the time of the trip to facilitate a comparison with the Stff data. It is also an age when youth are likely to travel independently.

The seasonal distribution of all trips shows a clear peak in the sum- mer months July and August, with little traveling in the winter months from November until February. This seasonality is caused by a variety of factors such as school holidays. The traditional holiday months July and August account for 20.3% and 20.7% of all trips, respectively. These figures are somewhat below the national average in 1992. Based

MARTIN OPPERMANN 545

on the Stff (1993), July and August were chosen by 26.8% and 24.8% of the travelers, respectively, as the months for the main holiday.

The average trip length was 14.8 nights. However, two facts need to be considered. First, the minimum length of stay used in this analysis was three nights. Second, many responses were somewhat general in nature. Frequently the trip length was only given as 1, 2, or 3 weeks. This was translated into 7, 14, or 21 nights for the sake of analysis. A large percentage of the trips lasted 8-14 nights (42%), 4-7 nights (26%) and 15-21 nights (21%). Vacations comprising 22-28 nights (7 %) were also common. Although the average German employee has holidays of about 30 working days, these are rarely taken in one block and not necessarily used to travel away from home. In 1992, the aver- age length of main vacation trips was 16 nights and some 50% of such trips lasted 8 to 14 nights (StfT 1993). About 26 % lasted 15-21 nights and 13 % 4 to 7 nights. These figures are relatively similar to the statistics derived from this survey.

The data suggest that younger people tend to leave Central Europe more often than older persons. Particularly those between 19 and 33 years old appear to leave Europe most often for overseas destinations (Table 4). Interestingly enough, there seems to be a small dip in the ages 34 to 48 with respect to overseas travel. This may be caused by the stage of the family life cycle. In this age group, many adults would have children, which makes traveling overseas much more troublesome and expensive. Commonly, the respondents do not travel at all during years when their children are young.

Generation Dt#erences

To analyze existing differences between successive generations, the sample is divided into four categories. A cohort analysis commonly uses specific years (i.e., 50, 60, and 70 years old) to discuss issues of generation differences (Becker 1992; Wagner 1989). The 124 respon-

Table 4. Destination Visits According to Age Group by Region (in %)

Age Group

15-18 19-23 24-28 29-33 34-38 39-43 44-48 49-53 54-58 59-73 >63

Sample Size

226 340 367 357 329 300 300 303 324 261 211

CEU EUR NAF NAM TWO

50.9 45.6 2.2 0.4 0.9 38.8 54.1 4.4 0.9 1.8 42.0 45.2 6.0 2.2 4.6 48.7 41.2 4.8 2.2 3.1 56.2 39.8 1.8 0.6 1.5 58.7 36.3 2.3 1.3 1.3 53.3 40.0 3.7 0.7 2.3 52.5 36.6 6.6 2.0 2.3 51.5 38.0 5.6 1.5 3.4 52.9 34.5 6.9 0.8 5.0 56.9 33.6 7.6 - 1.9

Note: CEU = Central Europe; EUR = Rest of Europe; NAF = North Africa; NAM = North America: TWO = Third World.

546 TRAVEL LIFE CYCLE

Table 5. Changing Incidence of Secondary Trips by Age Group and Cohort (in % of primary trip generation)

Age Group Cohort 1 Cohort 2 Cohort 3 Cohort 4

(N = 520) (N = 783) (N = 1007) (A’ = 1008)

15-18 40.2 10.4 5.6 b 19-23 43.2 21.3 11.4 b 24-28 47.8 31.0 6.7 b

29-33 68.0 52.3 26.1 3.2 34-38 a 64.8 19.8 9.8 39-43 a 74.4 23.8 21.4 44-48 a b 34.8 30.5 49-53 a b 36.4 38.7 54-58 a b 38.1 49.0 59-63 a b 50.0 48.2 >63

a b a 54.9

“No observations. bNumber of observations below 30.

dents were divided into four about equally sized groups with similar age intervals to account for the limited number of returned question- naires. Thus, cohort I included 35 respondents born 1960-74, cohort II those 32 born 1945-59. Cohort III consisted of 30 respondents born in the years 1930-44 and the fourth category all others born before 1930 (29). These four age categories represent four different political and economic time periods in German history, which could account for some differences between the cohorts. Those born between 1930 and 1944, for example, grew up in a time of economic depression, under the impression of World War II and the hard years of rebuilding in the late 1940s. Consequently, they were unlikely to have gained much travel experience in younger ages as compared to a generation born and raised in economic prosperity (e.g., those born 1960-74). Despite its shortcomings, the approach adopted seemed to be the best possible way to inquire into generation differences with respect to travel patterns given the data limitations.

Two aspects were considered important in analyzing differences be- tween generations, namely travel frequency and destinations. In the absence of a better indicator, the number of secondary trips as a pro- portion of primary trips is used here as an indication of travel fre- quency. Table 5 shows that the four cohorts differ markedly with respect to travel frequency, if one compares the same age categories. The incidence of secondary trips at the age between 29 and 33 ranges from 3% (cohort IV) to 68% (cohort I). Similar differences occur for tertiary trip generation, although the number of observations is too small to be reliable. Hence, the travel frequency of those who were between 19 and 33 years old in 1992 is considerably different from the one of earlier generations. Thus, the younger generation has more travel experience at younger stages of their life course than their older counterparts and may, therefore, choose completely different sets of destinations in later stages of their life-span than their predecessors.

MARTIN OPPERMANN 547

To determine differences in destination choice between the four cohorts, visitation of Central Europe, Europe, and other destinations is compared. Data for all four cohorts was only available for the ages of 15 to 33. To obtain a sufficiently large number of observations of all cohorts for chi-square analysis, all four age categories (15-18, 19-23, 24-28, and 29-33) are collapsed and analyzed together (Table 6). Vacation destinations in Central Europe are less popular among the “younger” generation (cohort I) while Europe increases in importance. This suggests that the younger generation do not only have a higher travel frequency, but also a different destination choice than preceding generations. One factor contributing to the changing destination choice is the increased availability of advertising, transportation, and destinations. In the 1980s at a time when members of cohort I were 15-33 years old, a 2-week holiday in a Third World country was not necessarily more expensive than a vacation of the same length in an European resort. In the 1930 when members of cohort IV were in the same age, traveling overseas was not only relatively more expensive, but required a lot more time (sea instead of air transport). Hence, destinations that are easily accessible and affordable today (e.g., Ca- nary Islands) were out of reach of the average German population in the 1930s. Because the younger generation has so many destinations available to them, it is unlikely that they will follow in the footsteps of their predecessors in choosing places in later stages of their life. How- ever, age and especially physical health are likely to limit the older generation’s destination choice as much in the future as today.

CONCLUSIONS

This paper explored several topics which have not received much attention in the tourism literature. Although the life cycle theory is a widely used concept, its application in tourism studies is restricted to destination areas (Butler 1980). Its application to the tourists them- selves in the form of family life cycles has been limited. Applications of the life cycle concept in tourism are largely in the form of a cross- sectional analysis of a group of travelers to a selected destination (e.g., Lawson 1991) or in the origin area (Bojanic 1992), rather than a longitudinal study of individual tourists. However, only the latter can reveal differences between different generations.

This pilot study indicated that a number of changes in travel pat- terns occurred during the last 30 years, although the results should be

Table 6. Changing Destination Choice at the Age of 15-33 Years by Cohort (in %)

Destination Region

Central Europe Europe Other

Cohort IV Cohort III (N = 61) (N = 222)

68.9 59.0 29.5 35.1

1.6 5.9

Cohort II (N = 487)

43.3 47.4

9.2

Cohort I (N = 520)

36.7 52.5 10.8

x2 = 46.95; df = 6; p = 0.0001.

548 TRAVEL LIFE CYCLE

viewed cautiously owing to data limitations. It has been suggested that there is an increasing travel frequency (i.e., more trips), particularly with respect to secondary and tertiary trips and a tendency to travel longer distances, in this way exploring new horizons. At the same time, however, there is a market segment that prefers destinations in the vicinity of home. This finding indicates the continuing significance of the costs overcoming distances (i.e., the distance decay principle) and travel restrictions caused by increasing age in the latter stages of the life cycle. A cross-sectional analysis of travel pattern indicated that destination choice, travel purpose, and month of travel varies accord- ing to age. A cohort analysis disclosed that successive generations have different travel patterns. The younger generation appears to be travel- ing more frequently and farther than previous generations at the same age. Through their traveling the youth gain different travel experi- ences, which makes it unlikely that they will select the same destina- tions as previous generations in later stages of their life span. This could have an enormous impact on tourism destinations, particularly on those places that, at the moment, depend on the older generations as their main market. Thus, such destinations cannot rely on the younger generation to sustain their market share. If such destinations do not change their marketing strategies to attract younger visitors, they most likely face declining visitation and decreasing revenues in the future.

Destination choice seems to vary along two axes: with the stage of the life course of the individual and with successive generations. If one visualizes these two axes as a coordinate system with the origin representing a point back in time, there are four sectors into which each destination could be classified according to changing preferences for destinations over time. Figure 1 shows a simplified coordinate sys- tem. Some destinations gain in visitation through the life course of the individual and through generation succession (+ , +) sector. These are destinations that are likely to experience increasing tourist numbers in the short and medium term. A classification according to the results of this pilot survey suggests that trendy places such as Turkey, Canary Islands, and Portugal belong into this sector. Those destinations with decreasing preferences on both axes (- , - ) will face declining visita- tion if they do not manage to change their image and attract new clientele. Traditional destination areas such as Austria/Switzerland, and particularly Southern Germany appear to belong into this sector. Destinations in the two other sectors hold an intermediate position with respect to their future. Those in the (+ , -) sector gain through the life course but do not fare well among the younger generation. To change their fortune, they may want to reorient their promotion efforts to younger people. Destinations in the (- , +) sector appear to be very attractive among the younger generation and during earlier stages of the individual’s travel cycle. Not surprisingly, this sector includes Southern European destinations such as Italy and Spain that became fashionable in the 1960s and 1970s. They are still typical “sunlust” destinations for German residents and especially among the young people, although other destinations may be more trendy.

Hence, using a longitudinal analysis of travel pattern with respect to

MARTIN OPPERMANN 549

Jeneration axis

+ 1 -,+ I

I +,+

I

1 Greece

1 Turkey Spain I Canary Islands

Italy : Portugal France I Northern Germany -____)__-_-___--_- ---_-

United Kingdom , Scandinavia Austria 1 East Germany

Switzerland 1 Belgium, Netherlands Southern Germany 1

I I

-,- I +,- age axis \

Figure 1. Distribution of Selected Tourist Destinations with Respect to Changing

Preference According to Life Cycle and Generation Succession

individual travel life cycles and changing destination choices of suc- ceeding generations provides destination areas with valuable insights into their markets. It may prove to be a very effective short and me- dium term forecasting tool which does not replace cross-sectional anal- ysis of travel choice, but adds a valuable supplement. Based on a larger and more scientific sample, a market segmentation according to destination choice over an individual’s life span appears feasible. This would improve the identification of target markets for the individual destinations and, therefore, increase the effectiveness of their market- ing efforts.

Tourism researchers may profitably expand their research focus from traditional cross-sectional studies to include longitudinal ap- proaches as well. Becker’s (1992) research and this study have shown that the complexities of longitudinal tourism studies of individual travel careers can be overcome, and that they contribute valuable insights on the multiple space and time horizons of the travel life cycle. q 0

Acknowledgment -Valuable comments of Siti Zabedah Amri on an earlier version of this paper are appreciated.

REFERENCES

Becker, C. 1992 Lebenslanges Urlaubsreiseverhalten-Erste Ergebnisse einer Pilotstudie. Ma-

terialien zur Fremdenverkehrsgeographie 25:70-82. Bojanic, D.C.

1992 A Look at a Modernized Family Life Cycle and Overseas Travel. Journal of Travel and Tourism Marketing 1(1):61-79.

550 TRAVEL LIFE CYCLE

Butler, R. W. 1980 The Concept of a Tourism Area Cycle of Evolution. Implications for the

Management of Resources. Canadian Geographer 24:5-12. Christaller, W.

1963 Some Considerations of Tourism Location in Europe: The Peripheral Re- gions - Underdeveloped Countries-Recreation Areas. Regional Science Associ- ation Papers 12:95-103.

Cohen, E. 1972 Towards a Sociology of International Tourism. Social Research 39:164-182.

Cooper, C., and S. Jackson 1989 Destination Life Cycle: The Isle of Man Case Study. Annals of Tourism

Research 16:377-398. Cosenza, R. M., and D. L. Davis

1981 Family Vacation Decision Making Over the Family Life Cycle. Journal of Travel Research 20: 17-23.

Dahms, F. 1991 St. Jacobs, Ontario: From Declining Village to Thriving Tourist Community.

Ontario Geogranhv 36:1-13. Dann, G., D. Naih, and P. Pearce

1988 Methodology in Tourism Research. Annals of Tourism Research 15: l-28. Dhalla, N. K., and S. Yuspeh

1976 Forget the Product Life Cycle Concept. Harvard Business Review 54:102- 110.

di Benedetto, C. A., and D. C. Bojanic 1993 Tourism Area Life Cycle Extensions. Annals of Tourism Research 20:557-

570. Din, K. H.

1992 The “Involvement Stage” in the Evolution of a Tourist Destination. Tourism Recreation Research 17( 1): 10-20.

Doling, J. 1976 The Family Life Cycle and Housing Choice. Urban Studies 13:55-58.

Foster, D. M., and P. E. Murphy 1991 Resort Cycle Revisited: The Retirement Connection. Annals of Tourism

Research 18:553-567. France, L.

1991 An Application of the Tourism Destination Area Life Cycle to Barbados. Tourist Review 46(3):25-31.

Gem, D. \ I

1992 Tourism Planning and Destination Life Cycle. Annals of Tourism Research 19:752-770.

Hart, C. W., G. Casserly, and M. J. Lawless 1984 The Product Life Cycle: How Useful? The Cornell Quarterly 24(4):54-63.

Haywood, K. M. 1985 Can the Tourist-Area Life Cycle be Made Operational? Tourism Manage-

ment 7:154-167. Hinch, T., and R. W. Butler

1988 The Reiuvenation of a Tourism Center: Port Stanlev. Ontario. Ontario Geoa- raphy 32:i9-52.

Hovinen, G. R.

,

1981 A Tourist Cycle in Lancaster County, Pennsylvania. Canadian Geographer 25~283-286.

Ioannides, D. 1992 Tourism Development Agents. The Cypriot Resort Cycle. Annals of Tourism

Research 19:711-731. Lawson, R.

1991 Patterns of Tourist Expenditure and Types of Vacation Across the Family Life Cycle. Journal of Travel Research 29(4):12-18.

Meyer-Arendt, K. J. 1985 The Grand Isle, Louisiana Resort Cycle. Annals of Tourism Research 12:

449-465. Mitchell, L. S.

1983 Family Life Cycle Leisure Activities: A Geographic Case Study. Paper pre- sented to the Geography Session of the South Carolina Academy of Science.

MARTIN OPPERMANN 551

1984 A Geographic Analysis of Leisure Activities: A Family Life Cycle Case Study. In Leisure, Tourism and Social Change, pp. 35-44. Edinburgh: Dunfermine College of Physical Education, Centre for Leisure Research.

Mitchell, L. S., and R. V. Smith 1989 The Geography of Recreation, Tourism, and Sports. In Geography in

America, G. L. Gaile and C. J. Willmott, eds., pp. 387-408. Columbus: Merrill. Murphy, P. E., and W. A. Staples

1979 A Modernized Family Life Cycle. Journal of Consumer Research 6:12-22. Pearce, P. L.

1993 Fundamentals of Tourist Motivation. In Tourism Research-Critiques and Challenges, D. G. Pearce and R. W. Butler, eds., pp. 113-134. London: Routledge.

Plog, s. c. 1973 Why Destination Areas Rise and Fall in Popularity. Cornell Hotel, Restaurant

& Adminstration Quarterly 14(3):13-16. 1987 Understanding Psychographics in Tourism Research. In Travel, Tourism, and

Hospitality Research, J. R. B. Ritchie and C. Goeldner, eds., pp. 203-214. New York: Wiley.

Rapoport, R., and R. N. Rapoport 1975 Leisure and the Family Life Cycle. London: Routledge & Kegan Paul.

Roberts, S. 1991 A Critical Evaluation of the City Life Cycle Idea. Urban Geography 12:431-

449. Rungaldier, R.

1924 Die geographische Verbreitung der Schutzhiitten und Unterkunftshauser in den Ostalpen. Mitteilungen der Geographischen Gesellschaft in Wien 67:33-38.

Smith, S. L. 1. 1996 A Test of Plog’s Allocentric/Psychocentric Model: Evidence from Seven Na-

tions. Journal of Travel Research 28(4):40-45. Stanslield, C. A.

1978 Atlantic City and the Resort Cycle: Background to the Legislation of Gam- bling. Annals of Tourism Research 5:238-251.

Stapleton, C. M. 1980 Reformulation of the Family Life-Cycle Concept: Implications for Residential

Mobility. Environment and Planning A 12: 1103-l 118.

Strapp, J. 1988 The Resort Cycle and Second Homes. Annals of Tourism Research 16:504-516.

Stff 1992 Urlaubsreisen 1991. Kurzfassung der Reiseanalyse 1991. Starnberg: Studien-

kreis fur Tourismus. 1993 Urlaubsreisen 1992. Kurzfassung der Reiseanalyse 1992. Starnberg: Studien-

kreis fur Tourismus. Vernon, R.

1966 International Investment and International Trade in the Product Life Cycle. Quarterly Journal of Economics 80:190-207.

Voit, H. 1992 Haushalts- und Familientypen 1972 und 1990. Wirtschaft und Statistik4/1992:

223-230. Wagner, M.

1989 Raumliche Mobilitat im Lebensverlauf. Eine empirische Untersuchungsozi- aler Bedingungen der Migration. Stuttgart: Enke.

Wall, G. 1982 Cycles and Capacity. Incipient Theory of Conceptual Contradiction? Tourism

Management 3:188-192. Weaver, D. B.

1990 Grand Cayman Island and the Resort Cycle Concept. Journal of Travel Research 29(2):9-15.

Wells, G., and G. Gubar 1966 Life Cycle Concept in Marketing Research. Journal of Marketing Research 3:

355-363. Wilkinson, P. F.

1987 Tourism in Small Island Nations: A Fragile Dependence. Leisure Studies 6: 127-146.

552 TRAVEL LIFE CYCLE

Zimmermannt C. A. 1982 The Life Cycle Concept as a Tool for Travel Research. Transportation 11:51-

69.

Submitted 28 July 1993 Resubmitted 9 February 1994 Resubmitted 2 May 1994 Final version submitted 25 July 1994 Accepted 12 August 1994 Refereed anonymously Coordinating Editor: Lisle S. Mitchell