Rethinking the Economic Contribution of Tourism

Abstract

The article introduces an integrated market-segmentation and tourism yield estimation framework for inbound tourism. Conventional approaches to yield estimation based on country of origin segmentation and total expenditure comparisons do not provide sufficient detail, especially for mature destinations dominated by large single-country source markets. By employing different segmentation approaches along with Tourism Satellite Accounts and various yield estimates, this article estimates direct economic contribution for subsegments of the UK market on the Mediterranean island of Cyprus. Overall expenditure across segments varies greatly, as do the spending ratios in different categories. In the case of Cyprus, the most potential for improving economic contribution currently lies in increasing spending on “food and beverages” and “culture and recreation.” Mass tourism therefore appears to offer the best return per monetary unit spent. Conducting similar studies in other destinations could identify priority spending sectors and enable different segments to be targeted appropriately.

Keywords

tourism yield market segmentation mature destinations tourism satellite accounts Mediterranean

Introduction

As a key factor driving tourism development, the economic impact of tourism has received considerable attention from tourism academics. One of the most important parameters that forms the basis of analysis of economic impact is visitor spending (Wilton and Nickerson 2006). Visitor spending typically includes expenditure on transportation, lodging, food and beverages, gifts and souvenirs, and entertainment and recreation (Jang et al. 2004). The resulting flow of currency into a destination’s economy has a significant influence on economic output indicators such as the value added by tourism, profit, national income, tax income, and employment (Klijs et al. 2012).

It is well established that, depending on the kinds of products or services being purchased and the direct and indirect linkages of the respective economic sectors (Cai, Leung, and Mak 2006), the same amount of overall expenditure can have a significantly different economic impact (or yield) (Dwyer and Forsyth 1997; Pratt 2012). Quantifying and understanding the economic impact of different types of tourism activities ultimately informs residents, consumers, businesses, and governments with regards to the most effective marketing techniques and best planning of facilities and amenities (Mok and Iverson 2000; Frechtling 2006; Dwyer and Thomas 2012).

The study of expenditure patterns inevitably leads to comparisons between different tourist types, each of which typically represents a distinct market segment. Expenditure patterns are used to estimate the size of each travel market in economic terms as well as to identify sets of attributes that account for differences in expenditure characteristics between market segments (Jang et al. 2004). An effective yield comparison between market segments relies on appropriate ways to characterize different tourist types in the first instance; otherwise, the comparison risks being meaningless. There is, however, no consensus over which yield measures or segmentation criteria are best employed in order to perform comparisons between tourist types. The choice will usually depend on the available data, the specific aims of the comparison, and the nature and stage of development of the destination. Segmentation of visitors has traditionally been based on their geographic origin, since country of origin (COO) has most often served as the basis for collecting and interpreting tourism data (Reid and Reid 1997; Andriotis, Agiomirgianakis, and Mihiotis 2008).

A critical challenge facing mature tourism destinations worldwide is how to maximize the economic impact of tourism spending (Alegre and Cladera 2006; Kozak and Martin 2012). Many mature Mediterranean “sun and sand” holiday destinations like Malta, southern Spain, and the Balearic Islands, Greek islands, and more recently, Turkey are following the well-documented path toward greater product diversification in an attempt to target the most profitable visitors (Bramwell 2004; Kozak and Martin 2012).

Established destinations tend to already have significant numbers of repeat visitors (Alegre and Cladera 2006), typically from large single-country source markets with a preference for the destination because of factors such as proximity, affordability, or other historical links. Most mature destinations in the developed world regularly collect expenditure data through passenger surveys, reflecting the long-term importance of tourism as a dominant economic sector. By combining appropriate segmentation techniques and tourism yield measures tailored to the destination in innovative ways, it is possible to explore the profitability of different subgroups within large source markets.

The aim of the study is to explore possible ways in which currently available tourism data in a Mediterranean mature destination could be used to supplement existing COO segmentation with tourist typologies based on other tourist characteristics and consumption patterns. The island of Cyprus, an established European Union (EU) destination in the eastern Mediterranean, represents an ideal case study as it is a small independent country, with a wealth of data from passenger surveys as well as fully fledged Tourism Satellite Accounts (TSAs) (Eurostat 2009). Furthermore, tourism is dominated by a large source market, the United Kingdom, which is chosen as the focus of this case study.

In a similar way to Pratt (2012), this article heeds the call made earlier by Dwyer et al. (2007) for more studies focusing on estimating yield for different destinations and different market segments. By comparing the results from different segmentation techniques and yield indicators to explore important COO markets, this study contributes to the existing body of literature by arguing for an integrated and flexible segmentation/yield approach, with an application to a mature destination with a dominant country source market. It is envisaged that this framework would potentially be generalizable to other mature destinations where, despite a relative abundance of expenditure-related data, a detailed understanding of the economic impact of different COO subgroups is currently lacking. This would provide a better insight on how to make the most out of existing tourism to the destination.

Market Segmentation

The tourism sector is highly heterogeneous and offers a plethora of different products that cater for different tastes, budgets, and times of the year—reflecting the fact that tourists are not homogeneous in terms of their desires and behavior (Hsu and Kang 2007). Understanding and explaining the motives and characteristics of different “types” of tourist has become a long-standing research objective, with a considerable number of empirical studies exploring similarities and differences in terms of travel patterns and attitudes between tourist groups (Kozak 2002). Forming and exploring market groups or segments has become a popular managerial practice known as market segmentation (Chen 2003). Dolnicar et al. (2012) argue that focusing on smaller and more manageable subgroups increases the chances of marketing success.

Market segmentation encompasses the choice of techniques and statistical procedures that allow the division of a heterogeneous market into relatively homogeneous subgroups (Mok and Iverson 2000). Depending on the available data and desired outcome, one or more segmentation criteria may be used with the main choices in the literature being geographical (such as COO segmentation), socioeconomic, demographic, psychographic (activities and opinions), and behavioral (Bigné, Gnoth, and Andreu 2007). Two principal approaches are recognized in the literature: a priori and data-driven (Dolnicar 2004). In a priori (or “commonsense”) segmentation, the researcher chooses a variable or variables of interest and then classifies tourists according to those predefined criteria. In data-driven (or post hoc) segmentation, a range of variables are used together to derive groups (segments) based on quantitative techniques data analysis such as cluster analysis (Najmi, Sharbatoghlie, and Jafarieh 2010).

The two approaches may also be combined to give rise to hybrid segmentation techniques. This usually occurs where a large commonsense segment (such as a large single-country source market) is further split into data-driven subgroups. According to Dolnicar (2004), combining a priori and post hoc segmentation in creative ways allows for a more original segmentation of tourists and, thus, for the identification of niche markets that have yet to be exploited. One of the aims of the present study is to explore this option.

Many tourism boards around the world, including the Cyprus Tourism Organization (CTO), rely on COO segmentation for targeting and marketing. The advantage of this straightforward a priori approach is the creation of segments that may be targeted through a common language and national media channels (Dolnicar 2008). However, according to Park and Yoon (2009), the purpose of segmentation is to promote improved efficiency in supplying products that meet the identified needs of target groups, with the ultimate objective usually being to make the most profit from selected target markets (Jang, Morrison, and O’Leary 2002). COO segmentation may capture some cultural differences in preferences and spending patterns between tourists (Kozak 2002; Becken and Gnoth 2004). Nonetheless, the intragroup variation within COO segments is often considerable and may result in other important descriptors being overlooked (Flognfeldt 1999).

This latter point is highly relevant to destinations where inbound tourism is dominated by certain large and extremely diverse COO segments. Some prominent examples are Caribbean and Mexican destinations for U.S. tourists; New Zealand for Australians; Thailand, Taiwan, Macau, and Hong Kong for the Chinese; and Malta, southern Spain, and Cyprus for the British. In such cases, enhancing COO with other segmentation techniques is not only beneficial but essential.

Tourism Yield Estimates Using TSAs

Segmentation studies focusing on the characteristics of big spenders have gained popularity in the tourism literature in recent years (Shani et al. 2010). Nevertheless, expenditure-based segmentation studies usually only consider total expenditure, with their main objective being to determine the demographic and behavioral characteristics of high-spending tourists, rather than to examine expenditure patterns and the economic contribution from different segments (Pratt 2012). Two travelers in the same area at the same time may choose to spend their money in entirely different ways (Legohérel 1998). The same amount of expenditure can, in fact, generate a different amount of gross value added (GVA)¹ as well as a different number of jobs depending on which products are being consumed (Salma, Suridge, and Collins 2004).

Although the precise definition of yield depends on the context and degree of resolution required (Becken and Simmons 2008), tourism yield most commonly refers to the financial or economic gain to a destination from attracting particular types of tourists (Dwyer, Forsyth, and Dwyer 2010). The simplest and perhaps still most commonly used measure for yield is total expenditure. However, total expenditure fails to indicate where and how the money is spent and how the benefits from the yield are spread throughout the economy (Dwyer and Forsyth 2008). Destination marketing organizations (DMOs) frequently collect expenditure data through various kinds of tourist expenditure surveys (Frechtling 2006). It is therefore possible to use such data to disaggregate total expenditure for each market segment into distinct spending categories. After matching spending categories to the TSA classification, expenditure data can be used to provide more meaningful estimates of tourism yield.

TSAs are satellite accounts used as an adjustment to a country’s national accounts to measure the economic significance of tourism activities, which are otherwise not separately identified in the conventional national accounting framework. TSAs are constructed using a combination of demand data from tourism surveys with data on the supply of goods and services taken from the System of National Accounts (SNA) framework (Smith, Webber, and White 2011). TSAs essentially aggregate the share of overall activity in an economy that is directly linked to tourism demand. Through effectively linking tourism supply and demand in a consistent and balanced framework that uses the same definitions and approaches as those agreed for the measurement of other economic activities (Bryan, Jones, and Munday 2006), TSAs provide a consistent tool for measuring the performance of the tourism sector.

From a country or DMO perspective, the wealth of information contained in TSAs can be used to identify the more profitable types of tourism, providing valuable insight that may then be used to inform and improve tourism policy (Jones, Munday, and Roberts 2003). By using the appropriate tables in TSAs, several measures of tourism yield can be estimated—such as tourism GVA (TGVA),² direct contribution to GDP, and contribution to employment in tourism industries.

Different measures of yield (such as GVA or employment generated) are unlikely to provide consistent rankings for different market segments (Dwyer et al. 2007; Becken and Simmons 2008), particularly since these measures can be estimated on a total market segment, contribution per visit, or contribution per night basis (Salma, Suridge, and Collins 2004). An analysis using a combination of different yield measures therefore provides a more complete understanding of economic impact. Total market segment contribution is not generally the preferred definition as it tends to merely highlight the importance of large markets. Recent yield studies (Lundie, Dwyer, and Forsyth 2007; Dwyer and Forsyth 2008) combine yield per day and yield per trip in order to identify high-yield markets dominated by long-staying high spenders.

The main limitation of yield estimates derived directly from TSA statistics is that they do not include indirect or induced effects (interindustry and economy-wide impacts) on income and employment (Ahlert 2008), which can only be estimated through economic modeling approaches. Despite this restriction—which does not allow TSAs to capture additional value added from industries producing intermediate inputs used to make goods and services consumed by tourists—they are considered to be an appropriate method where the study objective is to determine direct economic contribution from tourism (Frechtling 2010).

TSAs provide invaluable information with respect to where tourists spend their money, as well as the extent to which different sectors directly benefit and depend on their spending (Hara 2008)—making them an appropriate tool for describing the size and overall significance of the tourism sector. As the present study is an initial attempt to estimate yield without the use of economic modeling, TSA-derived yield estimates of direct economic impact are deemed satisfactory as a first step, and already represent a substantial improvement over simple expenditure measures.

Methodology

Context: Tourism in Cyprus

Cyprus is the easternmost island nation in the Mediterranean Sea. Similarly to other islands in the region, a favorable climate, natural beauty, and rich history have made the island a popular tourism destination. As in the case of many other Small Island Tourism Economies (SITEs),³ the development of tourism has been a remarkable success story, with the establishment of tourism as the dominant economic sector since the early 1980s (Sharpley 2001). However, the growth in tourism revenue experienced in previous years appears to have come to a halt after the turn of the millennium (Adamou and Clerides 2009). Some degree of recovery occurred in 2011 and 2012 (CYSTAT 2013), largely because of a “boom” in business-driven arrivals from Russia. However, there is currently no evidence that this trend will persist, especially in light of the recent financial crisis faced by Cyprus’s major banks.

As in other mature tourism destinations, key factors that have contributed to the fall in tourism revenue have been: rising costs as Cyprus has become wealthier (Ayres 2000; Clerides and Pashourtidou 2007), failure to attract higher-spending tourists (Adamou and Clerides 2009), emergence of cheaper mass tourism destinations in the area (Saveriades 2000), overreliance on foreign operators resulting in revenue leakage (Ayres 2000), ageing infrastructure (Sharpley 2003), and, in more recent years, the global financial crisis (Boukas and Ziakas 2013).

At present, Cyprus appears to be following the example of other mature Mediterranean destinations (Clerides and Pashourtidou 2007; Adamou and Clerides 2009), by moving toward a better quality tourism product in a bid to attract a “higher spending” clientele. While the drive toward diversification may be justified and somewhat overdue (Andronikou 1986), others have reasoned that it may be more prudent to focus on maintaining and improving the existing tourism product as part of a more flexible adjustment strategy (Ayres 2000; Sharpley 2003).

The present study makes a timely contribution to the debate by estimating tourism contribution from established segments, and providing management suggestions to allow the maximization of yield through potentially keeping and enhancing spending from the large pool of repeat visitors. In the absence of any sound segmentation and yield estimates, targeting and marketing to all UK tourists, who accounted for 37%–58% of annual inbound tourism to Cyprus over the last two decades (CYSTAT 2012), ignores the considerable heterogeneity within this large and important source market. Similarly, in the absence of any rigorous quantification, the a priori assumption that all mass or package tourism has a low yield remains an oversimplification. Exploring segmentation possibilities within the UK market should allow more useful management implications to be drawn.

General Overview

The methodological approach is composed of three main steps. Step 1 involved preprocessing the passenger survey data set in order to render it suitable for segmentation analysis. Step 2 involved COO segmentation followed by expenditure-based segmentation as well as cluster analysis of the UK market. The final step was to estimate tourism yield for each of the market segments established in step 2.

Data Sets and Preprocessing (Step 1)

The proposed framework requires TSAs along with passenger survey data (which includes expenditure in different categories). According to the UNWTO (2010), 60 countries had already produced or were in the process of developing TSAs as of early 2010, with more countries likely to have been added to this list since then. Eurostat reports that, despite methodological discrepancies between different national TSAs, most countries in the EU have fully fledged TSAs (Eurostat 2010). This includes key Mediterranean mature destinations such as Portugal, Spain, Italy, Greece, and Cyprus, all of which have the immediate potential to adopt the framework outlined in the present study.

Passenger survey data that includes expenditure is typically collected through exit surveys: tourists are asked to provide an estimate of total and detailed expenditure during their entire visit, or, alternatively, an estimate of what they spent on the last day of their trip (Wilton and Nickerson 2006). According to Frechtling (2006), such expenditure data are widely available for many destinations. Ideally, expenditure categories in the survey data must match those of the TSAs, but even when this is not the case, sector aggregation and matching may be used to address the issue (as shown in this article; see Figure 2). Exit surveys also record a wealth of sociodemographic and behavioral characteristics that allow for examining the relationship between tourist choices and expenditure patterns.

The present study uses the latest edition of the Cyprus TSA from 2007 along with the corresponding passenger survey data for the same year. Both were supplied by the Cyprus Statistical Service (CYSTAT); the exit surveys were administered through questionnaires at the island’s two major airports (CYSTAT 2011). The survey data was obtained in a raw form to allow alternative methods of segmentation other than COO to be used. In the preprocessing stage, all desired variables were merged into one data set for the entire year using the software package R. All expenses were converted from reported foreign currencies to Cypriot pounds (used in the TSA) based on monthly currency exchange rates provided by the CTO. The total sample size is 30,849 cases (which corresponds to 60,456 individuals). This represents a very large data set compared to previous segmentation studies.

Segmentation (Step 2)

The study used two types of a priori segmentation (COO and expenditure-based segmentation) as well as a data-driven technique (see Figure 1). The present study is the only study in the literature, to our knowledge, that compares two different segmentation methods in order to divide up a large source market. The objectives of the segmentation stage were to explore the diversity within the UK market segment and to create subsegments of a size that would allow yield comparisons between them and with other COO segments. Journey costs⁴ were removed as these are more likely to benefit international companies and firms within the tourist’s COO as opposed to the local economy (Legohérel and Wong 2006). For package tourists, local expenditure consists of the cost of the package (minus 15% to reflect foreign agency profits), plus any additional expenses minus estimated journey costs. For nonpackage tourists, local expenditure consists of all costs minus journey costs. Data analysis was carried out using SPSS version 19.0.

Figure 1.

Segmentation procedure carried out in the study. Cross-tabulation of variables was used to create country of origin (COO) segments. The UK segment was subsequently segmented in two different ways.

The first a priori segmentation performed was the traditionally used COO segmentation that created tourist groups based on their COO (see Figure 1, left). Cross-tabulation of variables allowed profiling of the main COO segments to determine their mean expenditure in each of the spending categories.

The UK market segment was then further segmented into three groups using expenditure-based segmentation (shown in Figure 1, top right). This commonly used a priori technique divides visitors into three equal-sized segments based on the frequency distribution of the total expenditure variable (Mok and Iverson 2000; Shani et al. 2010). Following the recommendation by Legohérel and Wong (2006), daily expenditure was chosen as the preferred segmentation criterion. Analysis of variance (ANOVA) and chi-square (χ²) were used to confirm that there are significant differences in the means between the spending groups for all spending categories as well as for all the other categorical and continuous variables.

Finally, a data-driven segmentation of the UK market was performed using the SPSS TwoStep® cluster analysis (see Figure 1, bottom right). This technique was chosen because it can simultaneously handle both categorical and continuous variables and was also specifically designed to handle large data sets (SPSS 2001). It consists of preclustering cases into many small subclusters, followed by clustering the subclusters into the desired number of clusters (Hsu, Kang, and Lam 2006). In the preclustering stage, observations are read and processed to decide whether they should be combined with an existing precluster, or whether a new precluster should be created, based on a log-likelihood distance measure. In the clustering stage, preclusters are then grouped through an agglomerative clustering algorithm (Huang and Han 2008). The goal of the clustering procedure was to use variables other than the expenditure categories in order to produce three homogeneous segments. A number of variable combinations were tested in order to arrive at a consistent solution with a good degree (>0.5) of intragroup cohesion and intergroup separation.⁵ ANOVA and χ² were used to confirm that there are significant differences between the groups across all expenditure categories.

Tourism Yield Estimates (Step 3)

To estimate yield from different segments, the study employed the methodological approach introduced by Tourism Research Australia (Salma and Heaney 2004). A similar approach is also found in Becken et al. (2007) and Becken and Simmons (2008), who estimate yield for previously established tourist segments in New Zealand. The Cyprus TSAs contain data on both tourism value added (TVA) and employment contribution for nine different spending categories. TVA is the financial contribution (value added) that “tourism dollars” support within domestic industries, and employment contribution refers to the number of jobs directly dependent on “tourism dollars” (Jones, Munday, and Roberts 2009). Combining the value added and employment contribution per sector to give total values per segment provides two valuable measures of yield. These are presented in a matrix form (as in Figures 3 and 4) to compare daily yield with total trip yield for different market segments (Dwyer, Forsyth, and Dwyer 2010), providing an intuitive visual tool where trade-offs and synergies between indicators become apparent.

Value-added and employment contribution are also compared to tourism consumption⁶ in order to provide two composite indicators of yield, namely, contribution to GVA and employment per unit of currency of tourism consumption at the market segment level (Salma and Heaney 2004; Dwyer, Forsyth, and Dwyer 2010). Yield is therefore estimated as tourism GVA per Cyprus Pound⁷ (CYP) of tourism consumption (TGVA per CYP) and as the number of full-time equivalent (FTE) jobs generated per million CYP tourism consumption (FTE jobs per million CYP). The steps involved in performing these calculations were as follows:

The expenditure categories for each segment established in step 2 were aggregated and merged according to Figure 2. The final five spending categories used are accommodation and accommodation-related spending (1), food/beverage/tobacco (2), transport (3), culture and recreation (4), and all other products including shopping (5).

For each segment, the proportion of expenditure compared to the average tourist expenditure in each category was calculated.

The proportions were applied to TSA estimates of aggregate tourism consumption, GVA, and employment contribution in each sector/industry.

The individual sector estimates were added up in order to provide estimates of mean total consumption, mean total GVA, and mean employment contribution from each segment.

Finally, each market segment GVA and employment contribution was divided by tourism consumption to provide per visitor per day averages corresponding to each market segment.

Figure 2.

Sector aggregation and matching to reconcile TSA and expenditure categories.

Results and Analysis

Segmentation Results

Tables 1 and 2 show the average characteristics for each segment. The mean value is used for the expenditure categories and continuous variables whereas the mode (and its percentage within the segment) is used for categorical variables.⁸ The analysis concentrates on selected findings.

Table 1.

Results of the COO segmentation for the five main country market segments.

		COO Segments
	Average Tourist	United Kingdom	Germany	Greece	Sweden	Russia
Sample (n)	30,849	10,841	2,254	1,982	1,413	1,635
Sample share (%)	100	35.1	7.3	6.4	4.6	5.3
Actual 2007 (%)	–	53.1	5.7	5.8	5.0	6.0
Mean expenditure (Cyprus pounds)
Accommodation	20.32	14.08	15.63	35.12	7.72	22.89
Food/drink/tobacco	20.20	20.22	11.63	24.13	17.19	18.70
Transport	7.75	5.79	6.34	17.95	3.09	6.84
Recreation/culture	2.97	2.07	3.55	4.90	1.81	5.18
Other	11.08	7.40	6.86	20.92	7.84	13.83
Mean total expenditure	56.21	42.14	51.57	78.60	27.80	64.18
Continuous variables
Mean party size (persons)	1.96	2.15	1.97	1.35	2.48	1.80
Mean LOS (days)	10.25	11.36	9.40	7.85	9.28	10.46
Categorical variables (mode)
Month	May–Oct	May–Oct	Oct–Nov	Dec–Mar	May–Oct	Jun–Oct
Month	62%	61.6%	22.5%	44.5%	70%	64.2%
Sex	Female	Female	Female	Male	Female	Female
Sex	54.3%	56.4%	50.3%	51.3%	58.5%	62.8%
Age	20–49	40+	30–49	20–49	20–39	20–39
Age	64.9%	68%	53.4%	67%	72%	74.3%
First time (yes/no)	No	No	Yes	No	Yes	Yes
First time (yes/no)	53.5%	67.8%	72.9%	76.1%	54.4%	58.8%
Purpose	Leisure	Leisure	Leisure	Business	Leisure	Leisure
Purpose	77.3%	86.7%	86%	36.7%	94.1%	81.8%
Area	A. Napa	Pafos	A. Napa	Lefkosia	A. Napa	Lemesos
Area	19.6%	25.4%	27.5%	41.8%	53.3%	41.4%
Alone (yes/no)	No	No	No	Yes	No	No
Alone (yes/no)	66.2%	80%	75.6%	73.8%	84.1%	57.7%
Package (Yes/No)	Yes	Yes	Yes	No	Yes	Yes
Package (Yes/No)	50.9%	51.4%	72.6%	90.7%	88%	65%
Accommodation	4-star	4-star	4-star	Friends	App. A	3-star
Accommodation	21.8%	20%	44.4%	46.8%	32.6%	26.5%

Table 2.

Results of the Expenditure-Based Segmentation and the Cluster Analysis of the UK Market Segment.

	Expenditure-Based			Cluster Analysis
	Low	Medium	High	1	2	3
Sample (n)	3,594	3,594	3,594	1,967	3,549	5,325
Sample share (%)	33.3	33.3	33.3	18.1	32.7	49.1
Mean expenditure (Cyprus pounds)
Accommodation	4.43	6.55	27.34	34.08	17.44	7.20
Food/drink/tobacco	15.94	17.07	27.62	25.08	18.86	19.44
Transport	3.25	4.51	9.13	11.93	5.63	4.27
Recreation/culture	1.65	1.86	2.74	3.10	1.46	2.28
Other	4.31	6.66	11.31	12.60	6.83	6.09
Mean total expenditure	12.06	26.64	87.05	73.09	49.46	26.01
Continuous variables
Mean party size* (persons)	2.21	2.25	1.99	1.05	2.55	2.29
Mean LOS (days)	14.00	11.48	8.62	11.76	13.43	9.83
Categorical variables (mode)
Month	May–Oct	May–Oct	May–Oct	May	May–Oct	May–Oct
Month	74.8%	74.5%	72.9%	10.7%	71.6%	83.2%
Sex	Female	Female	Female	Female	Female	Female
Sex	60.2%	56.5%	52.6%	52.3%	57.8%	57.0%
Age	40+	40+	50+	40–49	50–59	40–49
Age	65%	72.1%	45%	22.3%	26.7%	24.4%
First time (yes/no)	No	No	No	No	No	No
First time (yes/no)	65.2%	67.7%	70.3%	79.9%	80.7%	54.8%
Purpose*	Leisure	Leisure	Leisure	Leisure	Leisure	Leisure
Purpose*	88.5%	91.3%	80.6%	47.4%	89.9%	99.0%
Area*	Paralimni	Pafos	Pafos	Larnaka	Pafos	Paralimni
Area*	29.1%	30.7%	24.3%	26.2%	25.7%	32.9%
Alone (yes/no)*	No	No	No	Yes	No	No
Alone (yes/no)*	83.4%	85.1%	71.7%	95.2%	100%	94.3%
Package (yes/no)*	Yes	Yes	No	No	No	Yes
Package (yes/no)*	64.3%	59.1%	68.4%	99.4%	93.4%	100%
Accommodation*	Friends	4-star	4/5-star	Friends	Friends	4-star
Accommodation*	23.5%	25%	47.3%	49.9%	30.7%	31.3%

Note: Variables with an asterisk were used as clustering. Alone (yes/no) and package (yes/no) are both used as dummy variables.

The COO segmentation (Table 1) reveals that the average UK tourist is fairly similar to the average inbound tourist. However, the UK tourist is more likely than the average to be a repeat visitor, more than 40 years old, travel with family or friends, visit the island for a holiday, stay slightly longer and spend less. Other COO tourist characteristics appear to be consistent with those seen in CYSTAT national reports and are therefore not discussed further here.

In the expenditure-based segmentation of the UK market (Table 2), “low spenders” are defined as those who spent less than 18.78 CYP (approximately 32 EUR) per day, “medium spenders” as those who spent between 18.78 and 36.98 CYP (approximately 63 EUR) per day, and “heavy spenders” as those who spent more than 36.98 CYP per day. All ANOVA and chi-square results for differences between segments in all variables were significant at the 0.01 level. Low spenders mostly visit Cyprus on a package holiday during the high season and stay in mass tourism resorts. Medium spenders and high spenders share similar characteristics; these tourists predominantly visit family holiday resorts and tend to be more than 40 years old. Most high spenders travel nonpackage and spend more in all categories compared to the other two segments—with the most significant difference in spending occurring in the accommodation category.

Table 2 also shows the results of the cluster analysis. The best combination of clustering variables was achieved using the variables with an asterisk. This produced a 0.65 value of intragroup cohesion and intergroup separation. The largest segment (cluster 3), which represents 49% of the UK market (and 26% of total inbound tourism to Cyprus), is composed only of package tourists and can, therefore, be considered representative of UK package tourists. The characteristics and spending patterns of cluster 3 appear to be a mix of those of the low spenders and high spenders segments established previously. Cluster 3 is given the name “budget mass tourism” to reflect its characteristics and spending patterns.

Cluster 1 (18% of the UK market and around 10% of the total) is mostly composed of nonpackage tourists traveling alone. The spending patterns of this segment appear to be similar to those of high spenders. Cluster 1 exhibits less seasonality, and a higher percentage of repeat visitors (80%). Cluster 1 is representative of more upmarket tourism and is thus dubbed luxury tourism.

Cluster 2 (33% of the UK market and more than 17% of the total) has characteristics that are more similar to those of the average tourist or the average UK tourist shown in Table 1. Cluster 2 tourists are mostly repeat visitors (81%), come in season with family or friends, and are not on package deals. This cluster is labeled “average nonpackage UK” to reflect the highly average nature of this cluster.

In terms of the expenditure categories, luxury tourists (cluster 1) spent double on “accommodation” compared to average nonpackage UK tourists (cluster 2) and more than four times as much as budget tourists (cluster 3). They also spent considerably more in the “transport” category (112% more than cluster 2 tourists and 180% more than cluster 3 tourists) and in the “other” category (84% more than cluster 3 tourists and 107% more than cluster 3 tourists), which includes various kinds of shopping. By contrast, spending in the “food and beverage” category is only 33% higher than in cluster 2 and only 29% higher than in cluster 3.

Tourism Yield

Yield estimates for all market segments are presented in Table 3. In terms of total GVA (column 2) and employment contribution per day (column 3) the results show that, in general, market segments with high tourism consumption (column 1), such as Greece, UK (high), and UK (luxury), perform the best in these indicators. However, in terms of TGVA per CYP, it is Sweden (0.51), UK low spenders (0.51), and the UK budget (0.50) segments that rank in the top three positions. UK budget and UK low spenders are also the top performers in FTE jobs per million CYP, with 44.65 and 43.92, respectively. The segments with the highest total tourism consumption—Greece, UK (high) and UK (luxury)—rank lowest in terms of both composite yield indicators (columns 4 and 5).

Table 3.

Summary of All Daily Yield Indicator Results for All Market Segments.

	1	2	3	4	5
Market Segment	Tourism Consumption per Visitor	TGVA per Visitor	Employment Contribution per 1000 Visitors	TGVA per CYP of Consumption	FTE Jobs per CYP Million of Consumption
Average tourist	60.25	27.11	2.42	0.45	40.13
UK (average)	47.27	22.34	1.97	0.47	41.75
Germany	43.62	18.68	1.81	0.43	41.50
Greece	107.75	44.36	4.09	0.41	37.97
Sweden	33.38	16.98	1.39	0.51	41.72
Russia	62.14	28.08	2.53	0.45	40.69
UK (low)	27.36	14.02	1.22	0.51	44.65
UK (medium)	34.78	16.99	1.47	0.49	42.18
UK (high)	75.22	34.37	3.06	0.46	40.73
UK (luxury)	86.73	37.51	3.44	0.43	39.70
UK (nonpackage)	48.07	22.28	1.97	0.46	40.97
UK (budget)	36.35	18.25	1.60	0.50	43.92

Selected data from Table 3 have been used to create Figures 3 and 4. Figures 3a and 3b show that market segments placed in the top-right quadrant, UK luxury (cluster 1), Greece, UK (high), and Russia, have above-average TGVA and above-average employment contribution per night and per trip. Market segments such as UK (low), UK (medium), Sweden, Germany, and UK budget (cluster 3) in the bottom-left quadrant are those with below-average daily and per-trip contribution. UK non-package (cluster 2) is placed in the top-left quadrant because contribution per trip is above average but contribution per night is below average.

Figure 3.

Tourism GVA per night and per trip (a) and number of full-time equivalent jobs generated per night and per trip (b) for all market segments. The figure shows a linear correlation between per night and per trip contribution.

Figure 4.

Market segments plotted to show trade-offs between two different measures of yield.

In Figure 4, the top right quadrant shows the segments whose GVA and employment contribution exceed that of the average tourist. Market segments in the bottom left quadrant such as Greece and UK luxury (cluster 1) are those with below average GVA and employment contribution. The market segments in the top left (none) and bottom right quadrants (Russia and Germany) contain the segments that perform above average in one indicator and below average in the other indicator.

Germany performs above average in terms of employment contribution but has below-average contribution to GVA. This occurs because of the high relative spending of this segment in culture/recreation (the sector employing the highest number of people, as shown in Table 4) and a low relative spending in the food/beverage and accommodation categories (both of which have a high relative GVA contribution per CYP of tourism consumption, as shown in Table 4). The top right quadrant in Figure 4 shows the market segments that perform above average in both indicators. This includes most segments, with the UK low, UK budget (cluster 3), Sweden, and UK nonpackage (medium) as the top performers.

Table 4.

Detailed Yield Breakdown for an Average Inbound Tourist.⁹

	1	2	3	4	5
	Tourism Consumption per Visitor per Day (CYP)	Tourism GVA per Visitor per Day (CYP)	Tourism GVA per CYP of Consumption per Industry	FTE Jobs per CYP Million in Each Category	FTE Jobs per CYP Million of Total Consumption
Accommodation	17.49	6.83	0.39	38.53	11.18
Food and beverage	15.36	10.96	0.71	55.61	14.18
Transport	18.13	4.76	0.26	35.63	10.72
Culture/recreation	1.30	0.70	0.54	160.39	3.47
Other	7.97	3.85	0.48	4.38	0.58
Total	60.25	27.11	0.45*	294.54	40.13**

Table 4 summarizes GVA and FTE jobs per CYP consumption in each industry. The high yield contribution from sectors such as “food and beverages” as well as the relatively low contribution of “accommodation” expenditure explain most of the differences between segments shown in Figure 4. The implications of this are discussed in the following section, focusing on the UK market.

Discussion

UK Segment Yield Comparisons and Management Implications

The findings highlight the established fact that different yield indicators often result in different rankings of market segments (Dwyer et al. 2007; Lundie, Dwyer, and Forsyth 2007; Dwyer and Thomas 2012; Pratt 2012). As expected, GVA and employment contribution per night generally appear to be proportional to total expenditure and tourism consumption. However, the two composite indicators (TGVA per CYP and FTE jobs per million CYP) reveal that low-spending tourist groups spend a high percentage of their total holiday budget in high TGVA and employment contribution categories.

Expenditure-based segmentation of the UK market is a straightforward way to determine average characteristics of tourists depending on their total spending. Most characteristics of the expenditure-based groups are fairly close to those of the average UK tourist. However, this study shows that when expenditure-based segmentation is combined with other yield measures, some useful conclusions may be drawn—especially if related and compared to the cluster analysis results. Higher-spending UK tourists tend to devote most of their additional expenditure to accommodation. This is far from ideal, as accommodation has a below-average contribution to both TGVA and employment. As a result, high-spending segments (luxury UK segment, high spenders, Greece, and Russia) make a low contribution to total consumption despite their high overall spending.

It appears that in the case of Cyprus, encouraging more spending in both the “food and beverage” and “culture and recreation” categories is a strategy to maximize yield from higher-spending segments. It could also be argued that tourists who already have a high daily spend would be easier to target, and that providing higher-quality dining, cultural, and recreation options to existing visitors should be a priority over investing in offering more upmarket accommodation. These findings are consistent with previous research emphasizing the potential to enhance the benefits of tourism to the local community through creating more linkages between tourism and agriculture (Telfer and Wall 1996; Torres 2003). Food in tourism is often considered to be an important part of the cultural experience of visiting countries (Lin, Pearson, and Cai 2011), and presents a way to further increase yield from high-spending tourists currently spending more in less profitable sectors.

Market segments with mass tourism characteristics may be less profitable in terms of their overall contribution to the tourism industry of Cyprus, but offer the best return per monetary unit spent. Promoting more spending from mass tourism segments (or attracting more tourists at least in the short term) could, therefore, form part of a more flexible adjustment strategy as suggested by Ayres (2000) and Sharpley (2003), instead of concentrating exclusively on investment to attract and maximize revenue from higher-spending market segments.

In the case of UK tourists, the spending structures of low spenders and budget tourists are efficient in the sense that they spend a high percentage of their overall budget on food. This reflects the fact that most of the package tourists are on half-board or bed-and-breakfast deals, which means that they still need to purchase some meals. The tendency of many agents and hotels in the past five years toward “all-inclusive” deals, as a way to address the fall in revenue and arrival numbers, appears to be an ineffective strategy as far as yield is concerned. This is similar to the Balearic Islands, where research has shown that as a result of a recent tendency toward all-inclusive deals, not only does total expenditure tend to be lower but also that spending on services outside the hotel and tourist complex tends to be significantly less (Alegre and Pou 2008).

Offering cheap accommodation-only deals to budget tourists is therefore a way to attract more tourists and at the same time ensure that they spend a greater share of their budget outside hotels. Mature destinations that find themselves in a stage of stagnation or decline often have an abundance of unoccupied apartments and hostels that could be recycled in this way. In the case of the UK market, two of the segments derived from cluster analysis, budget tourists and average nonpackage tourists, represent more than 80% of the UK market, which corresponds to 45% of all inbound tourism to Cyprus. These segments have large party sizes that indicate the presence of families. Increasing spending on food and beverage as well as culture and recreation should be prioritized by providing the relevant options for each of these market segments. There is certainly potential to perform further segmentation, such as focusing on different destinations on the island. This could be followed by studying the implications of shifting spending from local transport to food and beverage and culture and recreation.

Limitations in the Present Framework

Despite highlighting a promising approach that integrates segmentation techniques and yield estimates, significant assumptions and data limitations do exist. As noted by Salma and Heaney (2004), this approach does not take into account differential rates of profit by type of product within each TSA sector/industry. For example, different kinds of food or other purchases can be more or less capital- or labor-intensive than others. This is an aggregation issue, the TSA and survey expenditure categories being relatively broad. Furthermore, imported goods and services are associated with higher leakage in tourist expenditure (Dwyer and Forsyth 1997), something that is not accounted for in this study.

In addition, TSAs can only be used to estimate the immediate effects of expenditure made by tourists on the economy (direct contribution) (Frechtling 2010). As a result, any attempt to measure tourism contribution solely on TSAs would underestimate the importance of the tourism sector to the overall economy (Smeral 2006). The indirect effects of tourism consumption can be estimated through the use of economic models such as computable general equilibrium models (CGEs), which can trace the ripple effects of tourism consumption across entire supply chains.

Only yield indicators related to GVA and employment contribution are considered in the present study. Many more indicators exist, with their appropriateness depending on the priorities and desired outcomes of the given study. Furthermore, the 2007 passenger survey data set is now fairly dated, but was chosen as this corresponds to the latest edition of publicly available Cyprus TSAs. The data and results for 2007 are therefore best seen as a snapshot of the situation at the time. Lastly, the simple five-sector aggregation of the present study highlights the need for improved data collection, which would allow for more comprehensive estimates of direct and indirect economic and environmental impacts.

Future Research Avenues and Required Additional Data

Additional primary data for tourist expenditure at a more detailed product level are certainly required. This would involve asking tourists more about their consumption habits and where their money is spent. Ideally, the questions should be detailed enough to match the national accounts or input–output (I-O) table (where these are available) sector classification, in order to avoid the need for aggregation and consequent loss of detail. In their study, Becken and Simmons (2008) use a 20-category expenditure classification that distinguishes different kinds of accommodation and diverse cultural sites and recreational activities. This is certainly more detailed than the present study, but food and beverage is still not disaggregated into different venues or kinds of food—which as shown in the present study holds a lot of potential for maximizing revenue from all kinds of tourists, especially those in the high end of the market.

Expenditure categories with high GVA and employment contribution are likely to differ across destinations. An approach such as the one outlined here to highlight priority sectors with most potential to improve contribution for different segments can subsequently encourage the collection of more detailed expenditure data within those categories, allowing the consumption of more profitable products to be promoted at the destination. The surveys could also ask tourists where they would be spending their money assuming that their transport or accommodation was cheaper. This is important, because maximizing yield ultimately depends on how tourists would be likely to spend any remaining budget or savings.

Yield from tourism expenditure also depends on the amount of products and services that are locally produced. If more goods and services within a certain expenditure category need to be imported, there will be a higher leakage of tourist expenditure (Dwyer and Forsyth 1997). Mature island destinations in the Mediterranean and elsewhere tend to import many products, including food. Targeting products that are locally produced is likely to increase the potential benefits of tourism spending; this is another reason why more detailed tourism expenditure data need to be collected. Mature destinations are often found in developed countries, where the resources and infrastructure to pursue such additional data collection (as described in the previous paragraph) are more likely to be available. It is also important that any additional data on expenditure from tourists are collected locally, at the destination. Using country-level data for segmentation and yield analysis works well for Cyprus because it is a SITE, but larger countries require destination-specific expenditure data.

It is envisaged that the current study will be expanded through further data collection and the use of I-O tables. Matching tourism consumption to the I-O classification and environmental data such as carbon emissions and water use (recently highlighted in Hadjikakou, Chenoweth, and Miller [2013] as a critical resource in many Mediterranean island destinations) could allow for estimates of the environmental impacts associated with different kinds of tourism spending, following the example of other researchers (Jones and Munday 2007; Lundie, Dwyer, and Forsyth 2007; Munday, Turner, and Jones 2013). This should eventually form part of an integrated segmentation-yield approach that can estimate “sustainable yield,” which includes not only economic but also social and environmental costs and benefits to a destination (Tyrrell, Paris, and Biaett 2013). Such a comprehensive framework could inform tourism policy with regards to the social–economic–environmental trade-offs associated with different market segments.

Conclusion

The integrated segmentation-yield approach established in this study is readily applicable to other destinations where the necessary data (TSAs and expenditure information) are available. As illustrated here, a highly suitable potential application is in mature destinations, where large source markets dominate. An advantage of this theoretical framework is its potential to be used iteratively, with the purpose of not only determining the most profitable segments but, even more importantly, providing insight with respect to where the most potential for maximizing revenue lies in existing segments.

The findings of the Cyprus case study give rise to valuable management implications in terms of maximizing value-added and employment contribution from UK tourists. The findings appear to challenge the orthodox view that moving “upmarket” is the only way to rejuvenate the tourism product. Spending on food and beverage as well as culture and recreation appears to produce maximum economic benefits, and it is the lower-spending segments that currently spend most of their money in these categories. The potential to increase spending in these categories exists for all segments and the study has made some relevant suggestions. The findings also illustrate that if commercial providers continue to favor all-inclusive deals, this is unlikely to be in the best economic interests of the country/destination.

The study has finally discussed the importance of collecting more detailed expenditure and has considered how such additional knowledge would allow more informed procurement of goods by tourism establishments. More disaggregated expenditure data could also be used to extend the framework in order to capture indirect economic impacts as well as environmental and social impacts. This is a future aspiration, but is likely to become more important as resources become scarcer.

Footnotes

Acknowledgements

The authors would like to thank the Cyprus Statistical Service for kindly supplying the data used in the study.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

Author Biographies

Michalis Hadjikakou is a PhD student at the Centre for Environmental Strategy at the University of Surrey. His main research areas are water management, sustainable tourism, tourism yield, and environmental input–output analysis.

Jonathan Chenoweth is Senior Lecturer in Water Policy and Sustainability at the Centre for Environmental Strategy at the University of Surrey.

Graham Miller is Professor in Sustainability in Business and Head of the School of Hospitality and Tourism Management at the University of Surrey.

Angela Druckman is Senior Lecturer in Sustainable Energy and Climate Change Mitigation at the Centre for Environmental Strategy at the University of Surrey.

Gang Li is Reader in Tourism Economics in the School of Hospitality and Tourism at the University of Surrey.

References

Adamou

Clerides

(2009). “Tourism, Development and Growth: International Evidence and Lessons for Cyprus.” Cyprus Economic Policy Review, 3 (2): 3-22.

Ahlert

(2008). “Estimating the Economic Impact of an Increase in Inbound Tourism on the German Economy using TSA Results.” Journal of Travel Research, 47 (2): 225-34.

Alegre

Cladera

(2006). “Repeat Visitation in Mature Sun and Sand Holiday Destinations.” Journal of Travel Research, 44 (3): 288-97.

Alegre

Joaquin

Pou

Llorenc

. (2008). “Tourism Expenditure and All-Inclusive Packages: The Case of a Mature Mediterranean Destination.” Tourism Economics, 14 (3): 645-55.

Andriotis

Agiomirgianakis

Mihiotis

(2008). “Measuring Tourist Satisfaction: A Factor-Cluster Segmentation Approach.” Journal of Vacation Marketing, 14 (3): 221-35.

Andronikou

(1986). “Cyprus—Management of the Tourism Sector.” Tourism Management, 7 (2): 127-29.

Ayres

(2000). “Tourism as a Passport to Development in Small States: Reflections on Cyprus.” International Journal of Social Economics, 27 (2): 114-33.

Becken

Gnoth

(2004). “Tourist Consumption Systems among Overseas Visitors: Reporting on American, German, and Australian Visitors to New Zealand.” Tourism Management, 25 (3): 375-85.

Becken

Lennox

Fitt

Butcher

(2007). Yield Associated with Different Tourist Types. Canterbury, NZ: Lincoln University.

10.

Becken

Simmons

(2008). “Using the Concept of Yield to Assess the Sustainability of Different Tourist Types.” Ecological Economics, 67 (3): 420-29.

11.

Bigné

Gnoth

Andreu

(2007). “Advanced Topics in Tourism Market Segmentation.” In Tourism Management: Analysis, Behaviour, and Strategy, edited by Woodside

A. G.

Martin

Wallingford: CABI, pp. 151-73.

12.

Boukas

Ziakas

(2013). “Impacts of the Global Economic Crisis on Cyprus Tourism and Policy Responses.” International Journal of Tourism Research, 15 (4): 329-45.

13.

Bramwell

(2004). “Mass Tourism, Diversification and Sustainability in Southern Europe’s Coastal Regions.” In Coastal Mass Tourism: Diversification and Sustainable Development in Southern Europe, edited by Bramwell

Clevedon: Channel View, pp. 1-31.

14.

Bryan

Jane

Jones

Calvin

Munday

Max

. (2006). “The Contribution of Tourism to the UK Economy: Satellite Account Perspectives.” Service Industries Journal, 26 (5): 493-511.

15.

Cai

Junning

Leung

PingSun

Mak

James

. (2006). “Tourism’s Forward and Backward Linkages.” Journal of Travel Research, 45 (1): 36-52.

16.

Chen

J. S.

(2003). “Developing a Travel Segmentation Methodology: A Criterion-Based Approach.” Journal of Hospitality & Tourism Research, 27 (3): 310-27.

17.

Clerides

Pashourtidou

(2007). “Tourism in Cyprus: Recent Trends and Lessons from the Tourist Satisfaction Survey.” Cyprus Economic Policy Review, 1 (2): 51-72.

18.

CYSTAT. (2011). Tourism Statistics 2009. Nicosia, Cyprus: Cyprus Statistical Service.

19.

CYSTAT. (2012). Tourism Statistics 2010. Nicosia, Cyprus: Cyprus Statistical Service.

20.

CYSTAT. (2013). “Arrivals of Tourists by Country of Usual Residence.” http://www.cystat.gov.cy/mof/cystat/statistics.nsf/services_71main_en/services_71main_en?OpenForm&sub=1&sel=1 (accessed June 2, 2013).

21.

Dolnicar

(2004). “Beyond ‘Commonsense Segmentation’: A Systematics of Segmentation Approaches in Tourism.” Journal of Travel Research, 42 (3): 244-50.

22.

Dolnicar

(2008). “Market Segmentation in Tourism.” In Tourism Management: Analysis, Behaviour and Strategy, edited by Woodside

Martin

Cambridge: CABI, pp. 129-50.

23.

Dolnicar

Sara

Kaiser

Sebastian

Lazarevski

Katie

Leisch

Friedrich

. (2012). “Biclustering Overcoming Data Dimensionality Problems in Market Segmentation.” Journal of Travel Research, 51 (1): 41-49.

24.

Dwyer

Forsyth

(1997). “Measuring the Benefits and Yield from Foreign Tourism.” International Journal of Social Economics, 24 (1/2/3): 223-36.

25.

Dwyer

Forsyth

Fredline

Deery

Jago

Lundie

(2007). “Yield Measures for Special-Interest Australian Inbound Tourism Markets.” Tourism Economics, 13 (3): 421-40.

26.

Dwyer

Forsyth

(2008). “Economic Measures of Tourism Yield: What Markets to Target?” International Journal of Tourism Research, 10 (2): 155-68.

27.

Dwyer

Forsyth

Dwyer

(2010). Tourism Economics and Policy. Bristol: Channel View.

28.

Dwyer

Forsyth

Spurr

Hoque

(2012). “Economic Impacts of a Carbon Tax on the Australian Tourism Industry.” Journal of Travel Research, Online:1-13.

29.

Dwyer

Thomas

(2012). “Tourism Yield Measures for Cambodia.” Current Issues in Tourism, 15 (4): 303-28.

30.

Eurostat. (2009). Tourism Satellite Accounts in the European Union, Volume I: Report on the Implementation of TSA in 27 EU Member States. Luxembourg: Office for Official Publications of the European Communities.

31.

Eurostat. (2010). Tourism Satellite Accounts in Europe. Luxembourg: Office for Official Publications of the European Communities.

32.

Flognfeldt

(1999). “Traveler Geographic Origin and Market Segmentation: The Multi Trips Destination Case.” Journal of Travel & Tourism Marketing, 8 (1): 111-24.

33.

Frechtling

D. C.

(2006). “An Assessment of Visitor Expenditure Methods and Models.” Journal of Travel Research, 45 (1): 26-35.

34.

Frechtling

D. C.

(2010). “The Tourism Satellite Account A Primer.” Annals of Tourism Research, 37 (1): 136-53.

35.

Hadjikakou

Chenoweth

Miller

(2013). “Estimating the Direct and Indirect Water use of Tourism in the Eastern Mediterranean.” Journal of Environmental Management, 114: 548-56.

36.

Hara

(2008). Quantitative Tourism Industry Analysis: Introduction to Input-Output, Social Accounting Matrix Modeling and Tourism Satellite Accounts. Oxford: Butterworth-Heinemann.

37.

Hsu

C. H. C.

Kang

S. K.

Lam

(2006). “Reference Group Influences among Chinese Travelers.” Journal of Travel Research, 44 (4): 474-84.

38.

Hsu

C. H. C.

Kang

S. K.

(2007). “CHAID-Based Segmentation: International Visitors’ Trip Characteristics and Perceptions.” Journal of Travel Research, 46 (2): 207-16.

39.

Huang

S. J.

Han

W. M.

(2008). “Exploring the Relationship between Software Project Duration and Risk Exposure: A Cluster Analysis.” Information & Management, 45 (3): 175-82.

40.

Jang

S. C. S.

Bai

Hong

G. S.

O’Leary

J. T.

(2004). “Understanding Travel Expenditure Patterns: A Study of Japanese Pleasure Travelers to the United States by Income Level.” Tourism Management, 25 (3): 331-41.

41.

Jang

S. C.

Morrison

A. M.

O’Leary

J. T.

(2002). “Benefit Segmentation of Japanese Pleasure Travelers to the USA and Canada: Selecting Target Markets Based on the Profitability and Risk of Individual Market Segments.” Tourism Management, 23 (4): 367-78.

42.

Jones

Munday

Roberts

(2003). “Regional Tourism Satellite Accounts: A Useful Policy Tool?” Urban Studies, 40 (13): 2777-94.

43.

Jones

Munday

(2007). “Exploring the Environmental Consequences of Tourism: A Satellite Account Approach.” Journal of Travel Research, 46 (2): 164.

44.

Jones

Calvin

Munday

Max

Roberts

Annette

. (2009). “Top Down or Bottom Up? Issues in the Development of Sub-national Tourism Satellite Accounts.” Current Issues in Tourism, 12 (4): 301-13.

45.

Klijs

Jeroen

Heijman

Wim

Maris

Diana Korteweg

Bryon

Jeroen

. (2012). “Criteria for Comparing Economic Impact Models of Tourism.” Tourism Economics, 18 (6): 1175-202.

46.

Kozak

(2002). “Comparative Analysis of Tourist Motivations by Nationality and Destinations.” Tourism Management, 23 (3): 221-32.

47.

Kozak

Martin

(2012). “Tourism Life Cycle and Sustainability Analysis: Profit-Focused Strategies for Mature Destinations.” Tourism Management, 33 (1): 188-94.

48.

Legohérel

(1998). “Toward a Market Segmentation of the Tourism Trade: Expenditure Levels and Consumer Behavior Instability.” Journal of Travel & Tourism Marketing, 7 (3): 19-39.

49.

Legohérel

Wong

K. K. F.

(2006). “Market Segmentation in the Tourism Industry and Consumers’ Spending.” Journal of Travel & Tourism Marketing, 20 (2): 15-30.

50.

Lin

Y. C.

Pearson

T. E.

Cai

L. A.

(2011). “Food as a Form of Destination Identity: A Tourism Destination Brand Perspective.” Tourism and Hospitality Research, 11:30-48.

51.

Lundie

Dwyer

Forsyth

(2007). “Environmental-Economic Measures of Tourism Yield.” Journal of Sustainable Tourism, 15 (5): 503-19.

52.

Mok

Iverson

T. J.

(2000). “Expenditure-Based Segmentation: Taiwanese Tourists to Guam.” Tourism Management, 21 (3): 299-305.

53.

Mooi

Sarstedt

(2011). A Concise Guide to Market Research: The Process, Data, and Methods Using IBM SPSS Statistics. Heidelberg: Springer.

54.

Munday

Turner

Jones

(2013). “Accounting for the Carbon Associated with Regional Tourism Consumption.” Tourism Management, 36:35-44.

55.

Najmi

Sharbatoghlie

Jafarieh

(2010). “Tourism Market Segmentation in Iran.” International Journal of Tourism Research, 12 (5): 497-509.

56.

OECD (Organisation for Economic Co-operation and Development). (2001). “Glossary of Statistical Terms” [database online]. http://stats.oecd.org/glossary/detail.asp?ID=1184 (accessed November 2, 2012).

57.

Park

D. B.

Yoon

Y. S.

(2009). “Segmentation by Motivation in Rural Tourism: A Korean Case Study.” Tourism management, 30 (1): 99-108.

58.

Pratt

(2012). “Tourism Yield of Different Market Segments: A Case Study of Hawaii.” Tourism Economics, 18 (2): 373-91.

59.

Reid

L. J.

Reid

S. D.

(1997). “Traveler Geographic Origin and Market Segmentation for Small Island Nations: The Barbados Case.” Journal of Travel & Tourism Marketing, 6 (3-4): 5-21.

60.

Salma

Heaney

(2004). “Proposed Methodology for Measuring Yield.” Tourism Research Report (The Journal of Tourism Research Australia), 6 (1): 73-81.

61.

Salma

Suridge

Collins

(2004). “Economic Contribution by Inbound Market Segments.” Tourism Research Report (The Journal of Tourism Research Australia), 6 (1): 83-93.

62.

Saveriades

(2000). “Establishing the Social Tourism Carrying Capacity for the Tourist Resorts of the East Coast of the Republic of Cyprus.” Tourism Management, 21 (2): 147-56.

63.

Shani

Wang

Hutchinson

Lai

(2010). “Applying Expenditure-Based Segmentation on Special-Interest Tourists: The Case of Golf Travelers.” Journal of Travel Research, 49 (3): 337-50.

64.

Shareef

Riaz

Hoti

Suhejla

McAleer

Michael

. (2008). The Economics of Small Island Tourism: International Demand and Country Risk Analysis. Cheltenham: Edward Elgar.

65.

Sharpley

(2001). “Tourism in Cyprus: Challenges and Opportunities.” Tourism Geographies, 3 (1): 64-86.

66.

Sharpley

(2003). “Tourism, Modernisation and Development on the Island of Cyprus: Challenges and Policy Responses.” Journal of Sustainable Tourism, 11 (2-3): 246-65.

67.

Smeral

(2006). “Tourism Satellite Accounts: A Critical Assessment.” Journal of Travel Research, 45 (1): 92-98.

68.

Smith

Webber

White

(2011). The Economic Importance of Tourism: The UK Tourism Satellite Account (UK-TSA) for 2008. London: Office for National Statistics.

69.

SPSS. (2001). The SPSS TwoStepCluster Component. Chicago: SPSS Inc.

70.

Telfer

D. J.

Wall

(1996). “Linkages between Tourism and Food Production.” Annals of Tourism Research, 23 (3): 635-53.

71.

Torres

(2003). “Linkages between Tourism and Agriculture in Mexico.” Annals of Tourism Research, 30 (3): 546-66.

72.

Tyrrell

Timothy

Paris

Cody Morris

Biaett

Vernon

. (2013). “A Quantified Triple Bottom Line for Tourism: Experimental Results.” Journal of Travel Research, 52 (3): 279-93.

73.

UNWTO. (2010). TSA Data around the World: Worldwide Summary. Madrid: UNWTO Statistics and TSA Programme.

74.

Wilton

J. J.

Nickerson

N. P.

(2006). “Collecting and Using Visitor Spending Data.” Journal of Travel Research, 45 (1): 17-25.