Post on 18-Jun-2020
!"""
US touristic clusters: the impact of the geographic effect on hotel’s economic
performance.
Abstract
The importance of geographical concentration and, therefore, its synergistic effects, has been widely
studied through the years but mostly from a manufacturing perspective. In this paper, USA touristic
clusters are identified and mapped using the location coefficient. Then, touristic cluster identification will
be used to classify hotels as properties inside or outside touristic clusters. We analysed five-year
economic data from Smith Travel Research (STR) database using an event study technique to compare
the economic performance among a total of 27207 hotels, 4339 of those located in touristic clusters and
those which are outside, considering that the aim of the research is to determine if the cluster effect is
affecting the economic performance of hotels. Hotels are segmented and compared to similar groups in
terms of revenues, scale (luxury, upscale, mid-price, economy and budget), location (urban, suburban,
airport, interstate, resort and small metro/town) and affiliation (chain, independent or franchise) and then
each of the hotel inside a previously classified touristic cluster is compared to a similar group of hotels
that do not belong to any geographic agglomeration. Though economic performance mean values are
higher of those properties located in clusters, specific analysis suggest that the cluster effect is not
affecting all the hotels in the same way.
Key words: US touristic clusters, cluster effect, hotels economic performance, synergic concentration,
positive externalities
1.- Introduction.
Despite the popularity of the cluster concept and considering that the tourism
industry presents considerable spatial concentration levels, the study of touristic clusters
is quiet recent (Jackson and Murphy, 2002, Michael, 2003, Jackson, 2006, Brown and
Geddes, 2007, Miller et al., 2008, Bernini, 2009, Lazzeretti and Capone, 2009, Erkus-
Öztürk, 2010, Weidenfeld et al., 2010, Segarra-Oña et al., 2012) and it is at an early
stage (Debbage and Ioannides, 2008).
The touristic concentration has been traditionally considered due to the existence
of natural characteristics or a heritage attraction (Secondi et al., 2011). However,
several authors have shown that the geographical concentration of the hotel industry
was largely justified by the presence of significant economies of location (Capone and
Boix, 2008)." The importance of geographical concentration and the evidence of the
existence of industrial clusters and, therefore, the synergistic industrial concentration
that is superior to the sum of its individual effects, have been widely studied through the
#"""
years (Brusco 1982, Porter, 1985, Krugman, 1991a, 1991b, Porter, 1998) following
Marshall’s seminal work (1890).
These localization’s economies have been minus valuated at the service sectors,
given that industrial agglomerations have traditionally been justified by the presence of
externalities in supply, harder to find in the service industry. However, some research
shows that services’ industries, when clustering, reduce consumer’s search costs (Chun
and Kalnis, 2001), which would confirm the importance of the so-called externalities of
demand (Canina et al., 2005, Freedman and Kosova, 2012) and would help to explain
territorial agglomerations of the hospitality industry.
If, as expected after the economic crisis at 2007, the tourism industry will have a
slower recovery compared to other economic activities (Smeral, 2010), it is urgent to
deepen in the understanding of profitability and performance of hotels related to cluster
effects to aware hotel managers to face with needed changes and provide the
information to take optimum decisions to maintain and improve its competitiveness
(Neely, 1999, So et al., 2006, Manzanec et al., 2007, Cabral et al., 2008).
" Therefore, the crisis is having direct consequences for the tourism industry, but
it can also be an opportunity, since we can find a least resistance scenario to confront
changes (Scott et al, 2008) and greater intra-industrial cohesion (Quarantelli, 1998).
These advantages could be clearer in tourist clusters, where cooperation, partnerships
and existing networks in specialized destinations are a source of tourism innovation
(Pikkemaat and Peters, 2005) which can be reinforced by the need to cope with a crisis.
In this research, once the cluster mapping is done, we compare those hotels
located inside and outside touristic clusters to see if the fact of being located within a
tourist cluster has any effect on the economic performance of the hotels considering a
broad period of time (2007-2011) and analyzing if the category of the hotel (luxury,
upscale, midprice, economy and budget), the location (urban, suburban, airport,
interstate, resort and small metro/town) or the management type (chain, franchise or
independent) are acting as moderating factors and therefore affecting the hotels from
benefitting from the positive externalities generated from being in a cluster.
This is known in the literature as cluster (Porter, 1998), geographical or district
effect (Becattini, 1990, Signorini, 1994) or synergic concentration (Segarra-Oña and
De-Miguel, 2009) and has been widely demonstrated at the manufacturing industry but
the research on the services industries’ studies are still scarce.
$"""
Taking into account the above, the main objective of this work will be to assess
if the performance along the time of hotels belonging to touristic clusters differs from
those allocated outside. So, the work will review the existing literature on touristic
clusters, and, using labor data, we will identify and map the US touristic clusters for
later analyzing and evaluating one by one the economic performance of 4339 hotels
which are located in a cluster since 2007 until 2011 with regard to its comparison group.
After the statistical analysis using an event methodology, results and their discussion
are presented. The paper ends summarizing the conclusions and standing the limitations
of the study. Also further research is suggested.
2. Literature review.
The acceptance, in recent years, of the importance that clusters have for
competitiveness (Bannister, 1994, Saxenian, 1996, Wright and Burns, 1998, Karaev et
al., 2007, Michaelides and Papazian, 2007, Asheim et al., 2008, Puig et al., 2009,
Piperopoulos and Scase, 2009, Spencer et al., 2010), has supposed an important change
in the appraisal of the locating of the businesses.
Until date, studies on tourism clusters have focused mainly on the role played by
the territory, the different actors and social and productive relations (Van der Berg et al,
2001, Nordin, 2003, Flowers and Easterling, 2006, Hall, 2005), as well as in the
knowledge transfer produced (Hallin and Marnburg, 2008).
While firms when allocating consider economic issues, legal and political issues
to facilitate location decisions (Jiang et al., 2006) and also accessibility, basic services,
site costs, environmental regulations, industrialization, labor availability, host taxes and
incentives, host government cooperation or exchange controls (Bass et al., 1977,
MacCarthy and Atthirawong, 2003), there is an academic consensus standing that
some competitive advantages reside, in the “know-how”, in the capacities, in the
information, in the motivation or the geographic externalities produced (Lazzeretti and
Capone, 2009). All of them are aspects related to the local business’ environment
(Ingram and Roberts, 2000) and that the competitors located outside the cluster find
more difficult to obtain (Nassimbeni, 2003). Until date, studies on tourism clusters have
focused mainly on the role played by the territory, different actors and social and
productive relations (Van der Berg et al, 2001, Nordin, 2003, Flowers and Easterling,
2006, Hall, 2005), as well as in the knowledge transfer produced (Hallin and Marnburg,
2008).
The cooperation and competition links are the basis of the cluster dynamics
(Porter, 1998) but this is a transversal concept that still keeps the researcher´s attention.
%"""
The benefits of the so called “coopetition” (Brandenburger and Nalebuff, 1996,
Bengtsson and Kock, 2000) have been applied to the analysis of the value chain (Yin,
2011) and the relations buyers-suppliers (Wilhelm, 2011) and is also considered a key
aspect in applied operations techniques as the lean manufacturing (Holweg, 2007).
The definition of the touristic clusters involves a high degree of complexity, on
one hand, it involves private investments as hotels, travel organizers, attraction and
leisure activities, but also public or hybrids (public-private) as railroads, roads,
museums, theatres, or municipal services (Rüdiger, 2004) and, on the other hand, public
policy implications regarding personal and political security (Vargas-Vargas et al.,
2010). So, the creativity and interaction of the different local partners are increasingly
playing a more important role (Richards and Wilson, 2006) and also the existence of
tourism clusters is enabling areas to compete globally while working together locally
(Erkus-Öztürk, 2009, Novelli, 2006, Ferreira and Stevao, 2009).
So far, much of the existing literature on touristic clusters is focused on
developing countries and in emerging tourist destinations (Sharma et al., 2007, Cabral et
al., 2008, Erkus-Öztürk, 2009 and 2010). But at the United States, as one of the top
world touristic destination (UNWTO, 2012), the hypothesis that says that there are
territories with high enough levels of tourism specialization for deemed tourist clusters
must be raised. This work will have as a first objective, to validate this hypothesis
through the illustration of a tourist cluster mapping of the United States, so we will
implement a simple ratio of territorial specialization, following the idea that the
geographical agglomeration is the essential of all geographical concentrated cluster
(O’Donoghue and Gleave, 2004).
Signorini (1994) was the first to attempt to quantify the district effect1, showing
that the productivity of companies within the district is greater than that of companies
outside it. Subsequently, other studies have made similar comparisons focusing on
different aspects such as technical efficiency (Hernández and Soler, 2003), the ability to
export (Melitz, 2003), the generation of externalities (Miret and Segarra, 2010) and the
ability to innovate (Cainelli, 2008).
In this line, Krugman (1991a, 1991b) focuses on the interaction between the
structure of the market and economic geography, considering the geographical
concentration as the most obvious fact of the existence of dynamic economic activity. In
""""""""""""""""""""""""""""""""""""""""""""""""""""""""1 In this work we consider it to be equivalent to dynamic concentration, synergic concentration and cluster.
&"""
this work we will interpret the existence and importance of specialized territories as
tourist clusters. There is a certain consensus regarding that the implementation of the
concept of cluster tourism is appropriate (Jackson and Murphy, 2002), however its use
is very recent and it still is in an embryonic stage (Nordin, 2003).
As known, the justification of a traditional manufacturing cluster is based on that
a specialized territory has access to more and better resources (Flyer and Shaver, 2003,
Tallman et al, 2004) and, albeit studies on these economies of localization have been
focused almost exclusively in manufacturing industries, due, mainly to the difficulty on
verifying the presence of external economies in the service industry (Canina et al.,
2005) references to the tourism cluster can be found at Porter cluster´s seminal work
(Porter, 1998). Yet studies that relate services sectors with clusters dynamics and
regional policy aspects are minority (Berg et al, 2001) and cooperation and
complementarities are important but understudied components of tourism clusters
(Weidenfeld et al., 2011).
Despite this, in recent years there is a growing interest in studying localization
economies based on the demand side. This idea was already present in Marshall’s work
(1890), who believed that crowds allowed consumers exploit economies posed by
reducing the costs of search.
This benefit is particularly important in industries with a high degree of
heterogeneity in their products because these require expensive searches for the
consumer (Fischer and Harrington, 1996, Freedman and Kosova, 2012). Specialized
territories allow the consumer to have and to evaluate a variety of different services
within the same area. These external economies based on demand are especially
important in the service industry, due to the effect that location has, as an intrinsic part
of the service offered, also considering related aspects as" '()*)*'" +,-" ./0+12-340"(++-*+)1*" 56" evoking desired emotional responses in customers (Voss et al., 2008).
When a company (public, private or hybrid) invests in making a location more
attractive, the rest of the businesses (say hotels) located in their vicinity can also benefit,
which also implies a positive externality.
These externalities of demand would justify the existence of clusters in the
service industries in general and, in particular, the touristic clusters, but in any case
should be noted two peculiarities of the touristic clusters:
1. Tourism clusters are the result of the location of complementary companies
which do not necessarily belong to the same sector (Novelli, 2006), but which
7"""
benefit from the existence of networks, alliances and other dynamics between
them (Weidenfeld et al, 2010, Saxena, 2005, Tinsley and Lynch, 2001, Presenza
and Cipollina, 2010), since the location of new business directly or indirectly
related to tourism not only add value to the companies inside the same cluster
but also the touristic experience itself.
2. The prominence of cooperation over competition. The traditional
microeconomic model of competition is not applicable because in this case
companies are required to cooperate to promote a destination (Von Friedrichs
and Gummesson, 2006). In the tourism market, the first thing sold to the
customer is the destination. Therefore, services and products must be combined
to offer the specific experience sought by the customer (Michael, 2003).
The concept of touristic cluster can be applied to a wide typology of service
industries as the financial cluster (Cook et al., 2007) and to other service sectors (Pandit
et al, 2008). Findings show that strong service clusters promote entrepreneurship, which
in turn promotes cluster strength in a self-reinforcing dynamic, that some firms are able
than others to better benefit from cluster location due to superior firm competencies and
absorptive capacity and that, cluster strength contributes to the ability of entrepreneurial
firms to expand overseas.
This new perspective has opened up the possibility of studying tourism clusters
(Michael, 2003; Jackson and Murphy, 2006). Most of these studies have focused either
on specific sectors such as tourism conventions (Bernini, 2009) or the food and wine
tourism (De Oliveira and Fensterseifer , 2003) and, specially, on studying touristic
clusters in emerging economies (Erkus-Öztürk, 2009, Sharma et al., 2007). But, from
the hotel industry, there are few studies done (Edgar et al., 1994, Baum and Haveman,
1997, Sharma et al., 2007, Brida et al., 2010) and there is still a research gap relating
touristic clusters and hotel profitability.
Our study focuses on the analysis of economic results considering the
agglomeration of service companies. And studies, particularly, the economic
performance of hotels belonging to a touristic cluster.
3. Research model.
The ability of a destination to skip a decline phase can come by the ability of
their hotels to take advantage of the externalities generated and by transforming them
into better economic results. This becomes especially important, on one hand, for policy
8"""
makers in charge of the touristic promotion and, on the other hand and, becoming of
particular importance, for hotel managers to set their strategies and their decisions.
Considering the stated below, our research question will expect to find better
economic performance along a period of time in hotels located in touristic clusters that
those of the same category, prize range, type of hotel and location that are outside the
cluster due to the benefits derived from the synergic concentration of firms and the
generation of positive externalities.
Given these arguments, we propose the following hypothesis:
H1 Economic performance of hotels located inside a touristic cluster remains better
that hotels located outside.
There are different methods that have been used for the cluster identification and
the subsequent preparation of cluster maps. Porter (2003) applied simple statistical
indexes based on the location quotient. This method of localization is simple and
flexible, although not free of criticism. New methods have emerged in recent years as
those based on the comparison of spatial distributions (Brenner, 2006), measuring
bilateral distances between companies (Marcon and Puech, 2003, Duranton and
Overman, 2005), the Herfindhal index (Sapir, 1996; Aiginger and Pfaffermayr, 2004) or
the absolute entropy index (Aiginger and Davies, 2004)."However, since our aim is not
to define the geographical scope of the cluster but rather to compare territories whose
boundaries have already been determined, the LQ method is considered to be
appropriate for our analysis.
In our case, we calculated the level of specialization of each territory, measuring
the greater or lesser weight of an activity in a specific territory, with regard to the
average for the total amount of the territories. When a region has a strong expertise in a
specific industry, we suspect the existence of a cluster in that territory calculating the
specialization of a territory (also called relative concentration) is through the location
quotient (LQ) or Hoover-Balassa index (Kim, 1995).
The term tourism industry is still used regarding industries directly related to
tourism and non-hotel businesses, in line with the Leidner’s work (2004). NCIS sectors
that identify the whole tourism industry can be considered: “Arts, entertainment, and
recreation” (NCIS 71) and “Accommodation and food services” (NCIS 72). Data for
LQ calculation was extracted from U.S. Bureau of Labor Statistics Database for the year
2010.
9"""
LQ forces to establish an arbitrary cutoff for cluster consideration. Miller et al.
(2001) considered that there is a cluster when LQ is above 1’25, and Malmberg and
Martell (2002) when LQ its over 3. O’Doneghue & Gleave (2004) tried to solve this
problem by developing SLQ (Standard location quotient), which identifies those
location that show extraordinary concentration values (LQ), that is their residuals show
statistical significance at a 5% confidence level (residual value over 1’96).
In this paper, LQ values higher than 3 where considered highly concentrated
touristic clusters. Then SLQ calculus determined a LQ cutoff of nearly 2 which was
considered as medium concentrated touristic clusters and finally, counties with an LQ
over 1.25 where considered as low concentrated touristic clusters.
So, being more specific, we split our first hypothesis:
H1a Economic performance of hotels located inside a low concentrated touristic cluster
remains better that hotels located outside.
H1b Economic performance of hotels located inside a medium concentrated touristic
cluster remains better that hotels located outside.
H1b Economic performance of hotels located inside a high concentrated touristic
cluster remains better that hotels located outside.
As previous works alert us from heterogeneity within the cluster (Freedman and
Kosova, 2012, Cook, 2009) and that there are considerable differences among the
different types of hotels (Segarra-Oña et al., 2012), we will analyse the effect of hotel
segmentation on the main hypothesis.
Segmentation in the industry has been used basically to identify consumer
characteristics (Bowie and Buttle, 2004) or user’s attitude (Bowen, 1998). In
geographical terms, segmentation can be by country, region city, town and even
neighbourhood (Lewis and Chambers, 1989), urban, suburban, rural and beach, by
population density, size of city or climate (Paley, 2001). In this paper, we segmented
US hotels by location. According to STR data classification, we classified hotels into
six groups: urban, suburban, airport, interstate, resort and small metro/town. Thus, we
offer the following:
H2 Property location influences economic performance of hotels within a touristic
cluster.
We will also analyse economic performance of hotels using a category
segmentation which validity has already been checked in previous works (Canina, Enz
:"""
and Harrison, 2005) that considers luxury, upscale, midscale, economy and budget, so
we propose the following:
H3 Property level influences economic performance of hotels within a touristic cluster.
Yet, to better complete the analysis and considering previous results warning
about the different performances of hotels depending on their management type
(O’Neill and Carlbäck, 2011, Perrigot et al., 2009, Botti et al., 2009), we will study the
performance of hotel in within the touristic clusters taking into account if they are a
franchise, belong to a chain or are independent, so we state the final hypothesis:
H4 Property management influences economic performance of hotels within a touristic
cluster.
The proposed research model is presented in Figure 1
Figure 1. Hypotheses representation.
3. Data selection and methodology.
3.1 Touristic clusters identification.
Since there is no global figure that describes the development of the tourism
sector as a whole and there is certainly an indicator’s problem definition of the tourism
sector. Some authors just identify tourism with hospitality as Lazzeretti and Capone
(2009), another studies include other industries as transport and recreational activities
(European Commission, 2003). We preferred to adopt an intermediate position as in
Leidner (2004), and to define tourism in a statistical way, NACE codes sectors and
subsectors, were matched to their equivalent NCIS code, available at the US Bureau of
Labour Statistics:
Table 1. NACE and NCIS equivalent codes. NACE codes NCIS codes
!;"""
-Restaurants, bars, canteens, catering
(NACE 553-555)
Arts, entertainment, and recreation
(NCIS 71)
-Travel agencies and tour operators
(NACE 633)
Accommodation and food services
(NCIS 72). -Hotel and Other Accommodations
(NACE 551-552)
The rate used to measure the specialization levels was the location quotient.
LQ=Eis/Es // Ei/E
Where: Eis is the number of employees in state S in sector i (i defined as tourism sector)
Es is the number of employees in state S
Ei is the number of employees in USA in sector I (i defined as tourism sector)
E is the total number of employees in USA
The same methodology had previously been used to analyze the performance of
clusters in the Spanish region of Valencia (Miret et al., 2010) and to study the Spanish
touristic clusters (Segarra et al., 2012). The data used in the analysis was taken from the
US Bureau of Labour Statistics (last version available 2010).
Regarding the geographical division, we thought about considering the local
labour markets as the location variable to measure, as it has been considered in previous
works to locate a regional tourism clusters in Valencia (Miret and Segarra, 2010) or at a
national level in Italy (Lazzeretti and Capone, 2009, Capone and Boix, 2008). However,
the resulting maps would be too much fragmented. On the other side, using state level
labor market shows big areas within states that rather don´t match with cluster findings
or that don´t show existing clusters because other industries employment within the
same state are hiding the evidence.
Therefore, we decided to use geographic county areas officially recognized,
which makes the analysis more simple, understandable and useful especially for agents
responsible for tourism planning and decision making.
As we said before, we considered three deferent concentration levels and we
decided to set up a cluster colour scale so we can show higher concentrations and lower
concentration cluster in a graphic way. Counties with an LQ higher than 3, highly
concentrated touristic clusters, were coloured in red. Counties with an LQ higher than
SLQ value (1.98) and lower than 3, medium concentrated touristic clusters, were
!!"""
coloured in orange. Finally counties with an LQ between 1,25 and 1.98, low
concentrated touristic clusters, were coloured in yellow as shown in figure 2.
[Insert figure 2 here]
3.2 In cluster and out cluster comparison. Data collection and methodology.
Hotel performance data for the study was provided by Smith Travel Research.
Data for the US market with the variables supply, demand and revenues for each
property, year and month was provided, rising over 4+ million data points.
The main variables provided in the dataset and their definition are shown in table 2
Table 2 Variable definition. PropSup Number of rooms available that day (if daily data); number of room-nights available that month (monthly data)
PropDem Number of rooms sold that day (if daily data); number of room-nights sold that month (monthly data)
PropRev Room revenue ($US) for that day (or month)
CompSup Number of competitors rooms available that day (if daily data); number of competitors room-nights available that
month (monthly data)
CompDem Number of competitors rooms sold that day (if daily data); number of competitors room-nights sold that month
(monthly data)
CompRev Competitors room revenue ($US) for that day or month
#Rooms Number of rooms in a hotel
Zip US Zip Code
Operation
Code
Chain, Franchise, Independent, Leased, Owned
Scale Economy chain through luxuy chain; independents
Price Pricing level (Five levels: budget, economy, midscale, upscale, luxury)
Location Characteristics of location: Urban, rural, airport, interstate, etc.
As we don’t know how many properties are involved, if they vary along the time
or which characteristics have the properties in each case for the competitors variables
(CompSup, CompDem and CompRev), these variables were dismissed.
Our intention was to study hotel performance and evolution over de last years
between properties inside touristic cluster and those outside them. Therefore, previous
data treatment was needed to handle the study.
First, we aggregated data for each year, so variables supply, demand and
revenues were aggregated calculating the mean for the total months available in a year.
Identification data remained as reported originally.
Secondly, we calculated the main variables needed to make the study for each
property and year. Variables and their definition are shown in table 3.
Third, we restructured database so data for each property was considered as a
single case, that is, we joined all data available for each property in a single row
creating variables in a time basis (Pe. revPAR2007 will be the rooms revenue per
!#"""
available rooms for the year 2007, revPAR2008 will be the revPAR value for the year
2008 and so forth).
Table 3 New variables created from the original dataset. Percentage of
Occupancy
The percentage of available rooms occupied for a given period. It is computed by dividing the
number of paid guest rooms occupied for a period by the number of rooms available for the same
period.
Average Daily Rate Total guest room revenue for a given period divided by the total number of paid occupied rooms
during the same period.
RevPAR Room’s revenue divided by the annual number of available rooms.
Forth, as we were interested in the evolution in the short, medium and long term,
considering specially the crisis environment, we establish year 2007 as benchmark year
for the event study following Hendricks et al. (2007).
We selected 2007, because at that time subprime mortgage crisis started in the
US and that event has been conditioning the touristic market evolution in the US.
As a result, we dismissed data previous to 2007 and cases with missing data for
the year 2007. At this point, we have a total of 27208 properties for which data was
available.
Finally, we identified properties in and outside the touristic clusters. As we
explained before, US counties were identified as being touristic clusters or not attending
to de LQ value. STR data doesn’t provide the county in which the property is based, but
zip code, state and MSA variables were available. Then, we used zip codes to locate
properties into counties and subsequently into clusters. To match zip codes with
counties we used US Postal Service Zip Codes for 1999 available in the US Census
Bureau. We used 1999 data as Census Bureau no longer tabulates decennial data by
U.S. Postal Service ZIP Codes and there are not updated versions since then.
After matching counties and zip codes, cluster classification was immediate.
4339 properties were classified in touristic clusters and the rest, 22867 properties,
outside touristic clusters.
"Anova results on mean revPAR over de period 2007-2011 between the two
groups show significant differences overall and for many of the segmented groups made
attending scale, price, location and affiliation. Properties in clusters show grater revPAR
value over those outside clusters as shown in table 4.""" But these results might be hiding real performance evolution as they are
influenced by a higher structural revPAR. That is, higher values of revPAR in in-cluster
properties over outcluster properties doesn’t mean necessarily that their performance
!$"""
evolution has been better over the period as revPAR starting point in 2007 could not be
the same. Therefore other methodology is necessary to answer our research questions.
Table 4. Economic variables comparison. ANOVA results.
Segment Group N Mean Std.Dev F Sig.
Total Outside 22868 50,75 36,59 140,73 0,000
Incluster 4339 58,18 43,83
Location Urban Outside 2277 86,28 63,73 0,00 0,996 In-cluster 314 86,26 45,37
Suburban Outside 9590 49,69 28,61 0,41 0,521 In-cluster 998 50,32 33,30
Airport Outside 1421 57,62 27,46 10,96 0,001 In-cluster 137 49,62 21,83
Interstate Outside 3218 38,38 17,52 1,11 0,292 In-cluster 735 39,15 19,44
Resort Outside 574 96,24 61,98 18,04 0,000 In-cluster 941 82,19 62,76
Small Metro/Town Outside 3881 42,07 20,72 131,58 0,000 In-cluster 1214 51,26 33,39 Affiliation Chain Management Outside 3494 61,30 51,83 91,51 0,000 In-cluster 517 85,62 66,66
Franchise Outside 16181 47,92 25,91 0,72 0,396 In-cluster 3240 48,35 26,99
Independent Outside 1286 83,44 67,22 2,34 0,126 In-cluster 582 88,52 64,85 ClusterConcentration Outside (O) (L, M, H) 20695 52,36 36,55 67,48 0,000 Low (L) (O, M, H) 3043 55,31 37,13
Medium (M) (O, L, H) 916 59,70 49,38
High (H) (O, L, M) 380 77,45 67,75 Letters in parentheses in Cluster concentration indicate the group from which this group was significantly different at a p-
value<0.05 level according to the Shaffe’s pairwise comparison procedure. F-statistics and associated p-values are derived from
one-way ANOVA’s.
3.3 The event study methodology.
We used an event study technique to determine whether properties inside
touristic clusters perform better than those outside.
We followed in some way the methodology used by Barber and Lyon (1996)
and then followed by others like Hendricks et al. (2007) to select comparison groups for
each property, although special adjustments have been made to adapt it to our purposes
and our reality. The first step is to determine the way to estimate the performance.
On the contrary to other industries, hotel industry has peculiar aspects that
makes difficult to use traditional measures (as ROA, e.g.) to evaluate performance.
For example, many hotels are owned by chains, so individual results for each
property are unavailable in many cases or unrealistic in others as they may not reflect
individual property performance, also specific effects like location or price range are
being hidden by aggregated results.
Therefore, data in a property level is crucial to evaluate the performance in this
study. Demand, supply and revenue data provided initially is clearly inappropriate to
!%"""
evaluate performance. Supply is a variable with little evolution within a property as
only provides the rooms available on sale in each month, and this will change only
because property refurbishing interventions or property enlargements. Demand will be
an indicator of a property evolution but will be insufficient to compare within properties
as properties might not have the same available rooms (supply). Finally, revenue will
depend on the demand and the property scale between other aspects, so properties
revenue comparison won’t show property performance, neither property evolution.
Then, we built other variables that have been used to show property
performance. First, occupancy (demand/supply) shows the efficiency of the property in
filling the hotel. An industrial analogy will be how well are they doing in using full
production capacity. The Average daily rate, ADR (revenue/demand), is showing at
which price hotels are able to sell their rooms, and therefore it is showing how much
will the client pay for the service. The revPAR (revenue/supply) is showing the
economic efficiency of the property, how good is the revenue in the property per unit
(room), and have been traditionally used as a performance indicator in the lodging
industry (Chung & Kalnins, 2001; Ismail, Dalbor, & Mills, 2002,"Kalninsa and Chung,
2004). Note revPar is also ADR over occupancy.
We compared the performance of each firm belonging to touristic clusters
against a preselected comparison group.
To estimate if hotels in clusters had a better evolution that those outside the
cluster, we estimated abnormal performance as the change in the sample firm’s
performance minus the change in the median performance of the comparison group
(Hendricks et al., 2007). The change in the level of performance for both, sample firm
and comparison group, is calculated comparing the level of performance in the studied
year versus the level of performance in the base year (2007). This method is preferred
over comparing each year with the previous year as these might lead to bias the results.
The change can be measured as a variation in the level of performance or a
percentage of change in the level of performance. For measuring occupancy, ADR and
revPAR, both measures can provide similar information, but we decided to track the
disparity in the level of performance, as it seems easier to interpret. Then the
interpretation of an upturn of occupancy rate from 70% to 73% from one year to the
next might be that the change in the level of occupancy is +3% rather than the percent
change in occupancy is +4,28%. To better evaluate firm’s performance, we have chosen
besides, occ., ADR and revPAR, the % of variance in demand and revenue to help
analysis.
!&"""
As mentioned before, variance in supply is minimal, so there is no sense in
keeping that variable. Also, absolute variables demand or revenue have no sense as their
value for the sample firm and the comparison group might be quite different so bias in
either sense is possible.
To establish comparison groups we focused on revPAR value."It has been used
in previous works to study hotel performance and it is closely related to operating profit
per available room (Enz et al., 2005) which takes into account costs of operation. We
checked this relation as Enz et al. (2005) did using PKF Consulting Hospitality
Reasearch Group data. Pearson Correlation between revPAR and operational profit per
available room from a group of 2740 properties with data from 2007 to 2010 (10960
data points) was 0,813 (p<0.01). This result suggests that both measures are strongly
linked. As some error in this correlation might be due to different cost structures in each
segment, we controlled for costs by assuring comparison groups in the same scale and
similar ADR value.
Following Barber and Lyon (1996), we decided to evaluate each firm against a
portfolio of firms is better than using only one firm to compare and, therefore, we used
a step procedure to select the firms belonging to the comparison groups.
Some factors might be affecting revPAR that will be hiding the touristic cluster
effect, therefore we think that we should control for this variability in the comparison
groups. We controlled for location (urban, suburban, airport, interstate, resort and small
metro/town), affiliation (chain management, franchise and independent) and scale
(luxury chains, upper upscale chains, upscale chains, midscale W/ F&B chains,
midscale W/O F&B chains, economy chains and independents) within each comparison
group. Price (luxury, upscale, midprice, economy, budget) was not controlled directly
by the correspondent variable as hotels can be affected by the region in which it is
located. Hotels in high populated regions are expected to have higher ADR than those
located in other regions, therefore controlling ADR and scale we consider to be more
accurate in terms of the similarity between the sample and the comparison group than to
do it by price. For example, the ADR for a luxury hotel in central US might be way
lower than the ADR for a luxury hotel in Manhattan or Boston.
To determine the comparison groups we followed Barber and Lyon (1996)
guidelines to assure well-specified and powerful statistics and an adapted procedure
from Hendricks et al. based on revPAR value:
!7"""
Step 1. For each property in the sample (in touristic cluster) we identify all
firms outside touristic clusters with the same characteristics (location,
affiliation and scale) that have a revPAR within the 95-105% of the sample
firm for the benchmark year (2007). We also control for ADR to be in the 90-
110% filter to guarantee the maximum similarity of the firms in each group.
Step 2. If we don’t find any firms in Step 1, revPAR filter is increased 5% in
each side controlling also characteristics and ADR to be to be in the 90-110%
and property characteristics (location, affiliation and scale).
Step 3. If we don’t find any firms in Step 2, we maintain revPAR filter in the
90-110% and firms characteristics filter and we eliminate ADR filter.
Step 4. If we don’t find any firms in Step 3, we maintain firms characteristics
filter and we look for the closest match in revPAR.
After determining sample’s comparison groups, median values for each variable
were calculated as shown in table 5. As benchmark year was 2007, will have a total of 4
values for Occ, ADR, revPAR, %variance of demand and revenue for each sample firm.
Table 5 Median values for each variable analyzed.
!
"#$%&#'(&)!!(*!'+&!),-%.&!
/0&#,1&!1#$2%!!)(3&!
4'&%!5! 6789! 66:;5!4'&%!<! 5=8! 5:78!4'&%!6! 58<! 55:9<!4'&%!;! <65! 5!
Outliers can affect significantly mean values of abnormal performance
indicators, therefore, we ran non-parametric tests such as Wilcoxon signed-rank test on
the median values and binomial sign test of the percentage of firms experiencing
positive abnormal performance to help and focus the interpretation of the results.
Consistent with our hypothesis we tested significance using one-tailed test.
Finally, in order to test our hypothesis, we split the sample in 3 groups according
to the cluster concentration classification made earlier in this paper. We considered low,
medium and high concentrated clusters to check if the grater the concentration of the
cluster the better the hotel performance. Then, we ran the same statistics previously
explained within each group attending to location, affiliation and hotel category.
5 Empirical results.
In analysing the results we will get more focused in non parametric results to
avoid the influence that outliers might be causing in the mean results.
!8"""
The results on the overall sample (table 6) provide evidence that the first year
after the subprime mortgage crisis, hotels in the clusters were significantly (p-value !
0.01) more affected than those outside the clusters as it can be seen in all the
performance indicators. For example, the median change in revPAR from 2007 to 2008
have been -0,75$ in sample firms than those in their comparison groups, and only
45.23% of the sample firms got better results than those in their comparison groups.
These results are due to the reduction in the median occupancy level (-0,59%) and in the
median ADR value (-0,55$), which obviously have been affected by the negative
evolution of the demand and revenues (-1,04% and -1.50% respectively).
Although this negative evolution in the first year continues on 2009, a small turn
over occurs over the following years but overall there is not a significant difference over
the period 2007-2011. Therefore, we can reject our first hypothesis that stated that in-
cluster properties will perform better on the long term in an economic crisis
environment.
To better understand what is happening and to check for hypothesis 2, 3 and 4,
we segmented our sample attending to their industry sector and quality levels (scale),
their location and their affiliation (chain, franchise or independent), and we used the
three cluster concentration categories already mentioned in this paper, low, medium and
high concentrated. The results on the segmented samples for revPAR are shown in
tables 7-11. Results on ADR, occupancy, % change in demand and % change in
revenues, although not reported where used to explain the results and they are available
from the authors.
Results for hotel level segmentation show quite a few differences that must be
underlined. Properties in-cluster belonging to Luxury segment show consistently better
and grater results every year than those outside the clusters, as mean, median and % of
positive abnormal results are always positive. For example, in the overall period
revPAR increased in median $2.87 (p-value ! 0.01, mean $3.29, 58.03% positive) per
year more in in-cluster properties that those outside the cluster when other variables
were controlled (location and affiliation).
For this segment and period median ADR increased $0.93 per year and
occupancy 0.12% per year, which is telling that revPAR increased is due because of an
ADR improvement rather than an occupancy improvement. Going further, if we look to
% variance of demand and revenues in this case, we see a significant increase on
revenues.
!9"""
Assuming supply is almost constant we will conclude that in-cluster properties
in luxury hotels have increased significantly their revPAR over those outside the cluster
because of a larger increase on revenues.
On the other side, from midprice to budget in-cluster properties have stronger
negative results that those outside clusters. In all cases those differences are statistically
significant as shown in table 7. Therefore, hotel level segmentation evidences how
strategic orientation is affecting the results, as upper priced in-cluster properties have
significantly better results that those outside the cluster and in mid-low priced works the
other way around. Figure 3 shows more clearly this differences in accumulated revPAR
change.
Table 8 suggest that chain management or franchise is not making a difference
on the results in and outside the cluster. According to results shown in table 8, we
should underline the significantly better results of in-cluster independent hotels over
those independents outside the cluster.
When we segmented the sample by property location, immediately raised an
extraordinary difference on the results in urban properties in cluster over those outside
the cluster. By the year 2011 almost 75% of in-cluster urban properties had better
results that their comparison group with a median increase on the revPAR of $6.86 (per
year from 2007 (p-value ! 0.01). Again, ADR, occupancy, %change in demand and
revenues were also significantly (p-value ! 0.01) higher, although the increase was
higher on revenues than on the demand. On the other hand, suburban and airport hotels
in-cluster, perform significantly (p-value ! 0.01) worse than those outside the cluster,
while in-cluster Resort and Metro/town properties over the studied period seem to
perform similarly as their comparison groups (See table 9 and figure 4).
Finally segmentation of the sample by cluster concentration (LQ level) show
that properties in low concentrated clusters have better results over their comparison
groups that those in medium or high concentrated (See table 10).
Our analysis of the performance of in-cluster over out-cluster properties shows
that negative economic environment affects more to in-cluster properties specially in the
short term. The first effect might be caused because of the relative impact that negative
economic evolution has on undiversified economies. Touristic clusters rely on single
touristic attractions (for example, gaming in Las Vegas or Atlantic city or theme parks
in Orlando), therefore hotel properties are more sensible to a variance in the touristic
attractions demand as no other reason is supporting the industry. On the other hand,
touristic clusters have more strength to face adverse conditions, they have more
!:"""
resources, higher economy scales and a higher influence in the sales channel, which
should lead to a quicker recovery on the performance indicators as economy takes again
the path of growth.
Even general performance is not positive, some in-cluster properties in selected
segments have shown extraordinary better results over their comparison groups, while
other segments behave the opposite. Analysing results will lead us to say the upper
priced urban hotels in touristic cluster have performed way over similar properties
located outside clusters, even in the first steps of an economic crisis. This might be for
various reasons; urban hotels are often located near touristic attractions, tourist wants to
be within walking distance of tourist attractions so there is usually a higher demand for
this type of properties and results are in concordance with previous studies that focus on
the necessity to identify differences between segments, indicating a need for unique
combinations of skills and assets within each segment (Shea and Roberts, 2008) and
also confirm the complex scheme of interaction between the geographic location, prize
and services (Urtasun and Gutierrez, 2006). Accordingly to these authors, in an urban
context greater benefits can be found in geographic agglomerations with competitors
with different services but greater costs that benefits were found in geographic
agglomerations with competitors with similarly priced hotels, what would explain why
luxury hotels, which are more differentiated, are able to get higher economic revenues
inside a tourism cluster than those located outside.
Low concentrated clusters seem to have better results than those medium or high
concentrated. We think that this might be caused by market saturation. That is, heavily
concentrated markets are force high competitive environments when there is a demand
restriction due to an adverse economic situation, hence, pressure to captivate clients
forces to focus strategies on price (reflected in the ADR) or in volume (reflected on
occupancy) which, eventually, may affect revPAR.
Figures 3 to 5 show accumulated revPAR change from 2007 through 2011 for
each level, location and affiliation.
Figure 3. revPAR change from 2007 through 2011 depending on the hotel
category.
#;"""
Figure 4. revPAR change from 2007 through 2011 depending on the hotel location.
Figure 5. revPAR change from 2007 through 2011 depending on the hotel
affiliation type.
"
6 Conclusion, limitations and further research.
This study identifies the existing touristic clusters in the US using labour data
from the US census and applying a LQ standard indicator. The map shows strong
#!"""
clusters in some well-known touristic areas as Las Vegas, Atlantic City, Orlando,
Hawaii islands, but also in others that we had hardly consider that it conforms a touristic
agglomeration. And can also be seen that “hot touristic spots” have influenced nearby
areas (counties) and high concentrated areas are work as seeds where other less
concentrated clusters have been developing.
Based on the former justification of a cluster and its dynamics regarding that a
specialized territory has access to more and better resources (Tallman et al, 2004) and
creates what some authors call "externalities resource-based" (Flyer and Shaver, 2003)
and that similarly, the concept of cluster had been previously and mainly used in
manufacturing industries, we analysed a key part of the touristic clusters, hotels,
analysing its performance compared to, in all extend, alike hotels allocated outside
clusters.
ANOVA analysis of hotels located in and out of these touristic clusters gives
preliminary results showing significant differences and suggesting that being located in
a touristic cluster can be related to the better economic performance of hotels, that
would be aligned with previous studies results applied to services industries (Lazzeretti
and Capone, 2008) and would confirm hypothesis 1, but when we segmented the
sample, results showed how economic performance differed depending on the sample
heterogeneity, what reinforces previous studies such Freedman and Kosova (2012).
ANOVA analysis showed better results for luxury and upscale hotels in-cluster
compared to those located outside the cluster but results wouldn’t be the same for mid-
price to budget hotels. In the same way, when studying hotel economic performance
considering the property location, we found out that airport hotels located in clusters
were performing much better that those located outside, but there were not any
significant difference when considering urban, suburban or interstate hotels, and resort
and town hotels were performing significantly worse than those that would not belong
to a touristic cluster.
When analysing highly, medium and low concentrated clusters results and
according to Shaffe’s pairwise comparison, there are significant differences among out-
cluster properties and in-cluster properties. We can see that the higher the concentration
is, the higher the level of revPAR over the period 2007-2011. As we said before, greater
revPAR value over the period considered not necessarily mean better evolution over the
period considered, so, as we can say that touristic agglomerations are somehow
affecting the economic hotel performance, the first hypothesis could not be fully
accepted. When analysing the sample considering cluster concentration, we found out
##"""
that hotels located in low concentrated clusters were performing better than those
located in medium or high ones, so we can accept H1a hypothesis and reject H1b and c,
which have important managerial implications when considering the benefits and costs
of where to locate a new hotel.
Data provided by STR, allowed us to determine whether there are differences in
the major variables as occupancy, average daily rate, revenues, expenses or margins
between hotels belonging to a US touristic cluster and hotels outside. Segmented studies
were also run attending to the property location (airport, city centre, resort or beach), if
it belonged to a chain or was a self-managed hotel and its price (from luxury to budget).
In addition, year variation was calculated to determine if the variables evolution
were significantly different.
Our analysis of the economic indicators using the economic crisis as an initial
point using an ES (event study) analysis shows mixed results.
First, the only group that is performing significantly better when analysing the
property location is the urban one.
Secondly, it reinforces the previous findings that showed that luxury in-cluster
hotels were performing much better than those located outside the cluster, in line with
previous studies that determined the importance of differentiation within a cluster
(Freedman and Kosova, 2012, Urtasun and Gutierrez, 2006, Canina et al., 2005), and
the heterogeneous performance within agglomeration domains (Chung and Kalnins,
2001, Yang et al 2012), also showing that luxury hotels have a higher ability to benefit
from the agglomerations externalities. Even if their prices are eroded easily when
collocating (Enz et al., 2008), it still seems to be more profitable. This result allows us
to accept the third hypothesis that affirms that property level influences economic
performance of hotels within a touristic cluster.
Those findings are consistent if we consider that usually luxury hotels are
located in the city center (Egan and Nield, 2000).
Thirdly, there is an interesting conclusion showing that independent hotels
perform better in-cluster that those out-cluster. This could be because those hotels have
a more agile managerial structure and are able to take its own decisions, which better
allow them to react to the environment and better benefit of the agglomeration
externalities and reinforces previous findings (Chung and Kalnins, 2001).
This results have important managerial implications, considering that the
allocation decision is one of the most important. Taking into account the possible
cluster effect or the importance of the destination will not be enough, it is also important
#$"""
to consider the differentiation level of the property and the property management. We
can conclude that the luxury and urban hotels are clearly benefitting from being
allocated in a touristic cluster, especially if it is a low concentrated touristic cluster.
That hotel manager should try not to allocate their properties in a highly
concentrated cluster has also been clarified. Also that in an economic crisis situation
hotel in-cluster are, in general, performing better that those out-cluster.
Albeit findings have a solid basement due that we analysed more than 27000
properties using different years data, the study has also its limitations. First, we are not
able to explain why against all forecast, airport and suburban in-cluster hotels are
performing worse that those out cluster and second that the period of time studied, from
2007 to 2010 may be biasing results.
To solve this, the study should be replicated in a period with more economic
stability and should also be interesting to analyse deeper the dynamics of US touristic
clusters, especially focusing on the modern key topics regarding innovation and
sustainability to see if better economic performance is also aligned with better
environmental and innovative attitude.
!
"#$%&'()*(+(,-.!The authors would like to thank the Spanish Economy and Competitiveness Ministry for its financial
support through the research project (EC02011-27369), the Universitat Politècnica de València for
supporting the sabbatical research of M. Segarra at Cornell University and The Center for Hospitality
Research at Cornell University that hosted A. Peiró and M. Segarra as Visiting Scholars. Also the
Universitat Politècnica de València for its research funding to the project “Impact of innovative practices
on the environmental performance of companies” (PAID-06-2011-1879). The authors also want to thank
Donald Schnedeker from the Nestle library at the School of Hotel Administration at Cornell University
and Russ Lloyd and STR managers for providing us with the data.
References
Aiginger, K., Davies, S.W., 2004. Industrial specialization and geographic concentration: two sides of the same coin? Not for the European Union. Journal of Applied Economics 7 (2), 231–248. Aiginger, K., Pfaffermayr, M., 2004. The Single Market and geographic concentration in Europe. Review of International Economics 12 (1), 1–11. Asheim, B., Cooke, P. and Martin, R. (2008), Clusters and Regional Development: Critical Reflections and Explorations. Economic Geography, 84, 1,109–112. Bannister, G.J., Stolp, C., 1994. Regional concentration and efficiency in Mexican manufacturing. European Journal of Operational Research, 80,3, 672–690. Barber, B.M., Lyon J.D, (1996).Detecting abnormal operating performance: The empirical power and specification of test statistics, J. Finan. Econ., 41(3): 359-399. Bass, B.M., McGregor, D.W., Walters, J.L., 1977. Selecting foreign plant sites: economic, social, and political considerations. Academy of Management Journal 20, 4., 535–551.
#%"""
Baum, J. A. C. and Haveman, H. A., 1997, Love The Neighbor? Differentiation and Agglomeration in the Manhattan Hotel Industry, 1898-1990, Administrative Science Quarterly, 42, No. 2, pp. 304-338 Becattini, G. (1990). The Marshallian industrial district as a socio-economic notion. In F.Pyke, G.Becattini und W.Sengenberger (Ed.), Industrial districts and inter-firm co-operation in Italy. Geneva: International Institute for Labour Studies Bengtsson, M., Kock, S., 2000. “Coopetition” in business networks: to cooperate and compete simultaneously. Industrial Marketing Management 29 (5), 411–426. Berg L.; Van Den Braun, E. and Windem, W (2001): “Growth Clusters in European Cities: an integral approach”. Urban Studies, 38 (1) pp.185-205. Bernini, C. (2009). Convention industry and destination clusters: Evidence from Italy. Tourism Management, 30(6), 878-889 Botti, L., Briec, W., Cliquet, G., 2009, Plural forms versus franchise and company-owned systems: A DEA approach of hotel chain performance, Omega, 37, 3, 566-578. Brandenburger, A.M., Nalebuff, B.J., 1996, Co-opetition. Doubleday, New York. Brenner, T (2006): “Identification of Local Industrial Clusters in Germany”. Regional Studies, 40. pp 1-14. Brida, J. G., Parte Esteban, L., Adrián Risso,W., Such Devesa, M. J., , 2010, The international hotel industry in Spain: Its hierarchical structure, Tourism Management, 31, 1, 57-73. Brown, K. G.; Geddes, R. (2007), Resorts, culture, and music: the Cape Breton tourism cluster, Tourism Economics, 13, 1, 129-141. Brusco, S. (1982). The Emilian model: productive descentralization and social integration. Cambridge Journal of Economics, 6: 167-184. Cabral Flecha, A., Alves Fusco, J. P., Damiani, W., Fernandes Amaral, H. (2010), The Economic Impacts of Tourism in Ouro Preto, MG, Brazil, Brazilian Journal of Operations & Production Management, 7, 2, 29-46. Cainelli, G. (2008), Spatial Agglomeration, Technological Innovations, and Firm Productivity: Evidence from Italian Industrial Districts. Growth and Change, 39:414–435. Canina, L, Enz, K. and Harrison, J. (2005) Agglomeration Effects and Strategic Orientations: Evidence from the US Lodging Industry, Academy of Management Journal, 48 (4), pp 565-581. Capone, F. and Boix, R. (2008) Sources of growth and competitiveness of local tourist production systems: an application to Italy (1991–2001), The Annals of Regional Science, 42, (1), 209-224. Chung, W. and Kalnins, A. (2001), Agglomeration Effects and Performance: A Test of the Texas Lodging Industry, Strategic Management Journal, 22, pp. 969–988. Cook, G. A. S., Pandit, N. R., Beaverstock, J. V., Taylor, P. J., Pain, K. (2007) The role of location in knowledge creation and diffusion: Evidence of centripetal and centrifugal forces in the City of London financial services agglomeration, Environment and Planning , 39(6), 1325-1345. Cook, G.A.S, Pandit, N. R. (2009), Clustering and the internationalisation strategies of SMEs in the media industry, Journal International Journal of Globalisation and Small Business, Vol. 3 (3), pp. 306-330 De Oliveira, E., Fensterseifer, W. J. E., (2003) "Use of resource-based view in industrial cluster strategic analysis", International Journal of Operations & Production Management, Vol. 23 Iss: 9, pp.995 - 1009 Duranton, G., Overman, H.G., 2005. Testing for localization using micro-geographic data, Review of Economic Studies 72 (4), 1077–1106. Edgar, D. A., Litteljohn, D. L., Allardyce, M. L. (1994) ,"Strategic Clusters and Strategic Space: The Case of the Short Break Market", International Journal of Contemporary Hospitality Management, Vol. 6 Iss: 5, pp.20 – 26. Egan, D. J., Nield, K., (2000), Towards a Theory of Intraurban Hotel Location, Urban Studies,37, 3, 611-621. Enz, C.A., Canina, L., Liua, Z. (2008), Competitive Dynamics and Pricing Behavior in US Hotels: The Role of Co-location, Scandinavian Journal of Hospitality and Tourism, 8, 3, 230-250. Erkus-Öztürk, H. (2009) The role of cluster types and firm size in designing the level of network relations: The experience of the Antalya tourism region, Tourism Management, 30(4), pp. 589-597. Erkus-Öztürk, H. (2010) Emerging Importance of Institutional Capacity for the Growth of Tourism Clusters: The Case of Antalya, European Planning Studies, 19, 10, 1735-1753. European Commission (2003) Structure, performance and competitiveness of European Tourism and its enterprises, EC, Luxembourg.
#&"""
Ferreira, J.M., Estevão, C.M.S., 2009, Regional Competitiveness of a Tourism Cluster: A Conceptual Model Proposal, Encontros Científicos-Tourism & Management Studies, 5, 37-51. Fischer J.H and Harrington J.E. (1996), “Product Variety and Firm Agglomeration” RAND Journal of Economics. 27, 2, 281-309. Flowers, J. and Easterling, K. (2006) Growing South Carolina’s Tourism Cluster, Business and Economic Review, 52, pp. 15-20. Flyer, F. y Shaver, J.M. (2003): Location Choices under agglomeration externalities and strategic interaction., in Joel A.C. Baum and Olav Sorenson (ed.) Geography and Strategy (Advances in Strategic Management, Volume 20), Emerald Group Publishing Limited, pp.193-213. Freedman, M. and Kosova, R. (2012). "Agglomeration, Product Heterogeneity, and Firm Entry" Journal of Economic Geography, Available at: http://works.bepress.com/matthew_freedman/10 Hall, C.M (2005) Rural Wine and food tourism cluster network development. In D. Hall I Kirkpatrick and M. Mitchell, (Eds), Rural Tourism and Sustainable Business, Channel View Publications, Clevedon, pp. 149-164. Hallin, C.A. and Marnburg E (2008):Knowledge management in the hospitality industry: a review of empirical research. Tourism management. 29 (2). 366-381 Hendricks, K.B. Singhal, V. R., Stratman, J. K. (2007) The impact of enterprise systems on corporate performance: A study of ERP, SCM, and CRM system implementations, J. Oper. Manag., 25(1):65-82. Hernández, F. and Soler, V. (2003) Quantification of the district effect through non-radial technical efficiency measures (In Spanish), Investigaciones Regionales, 3, pp. 25-40. Holweg, M., 2007. The genealogy of lean production. Journal of Operations Management 25 (2), 420–437. Ingram, P. and P. W. Roberts. (2000).“Friendships among Competitors in the Sydney Hotel Industry.” American Journal of Sociology. 106:2, pp. 387–423. Jackson, J. and Murphy, P. (2002) Tourism Destinations as clusters: Analytical experiences from the New World, Tourism and Hospitality Research, 4 (1), pp. 36-52. Jackson, J. and Murphy, P. (2006) Clusters in Regional Tourism Destinations: an Australian Case, Annals of Tourism Research, 33 (4), pp. 1018-1035. Jackson, J. (2006) Developing regional tourism in China: The potential for activating business clusters in a socialist market economy, Tourism Management, 27, 4, 695-706. Jiang, B., Frazier, G. V., Prater, E. L. (2006), "Outsourcing effects on firms' operational performance: An empirical study", International Journal of Operations & Production Management, Vol. 26 Iss: 12, pp.1280-1300. Kalnins, A., Chung, W. (2004), Resource!seeking agglomeration: A study of market entry in the lodging industry, Strategic Management Journal , 25(7), 689-699. Karaev, A., Koh, S.C. L., Szamosi, L.T., (2007) "The cluster approach and SME competitiveness: a review", Journal of Manufacturing Technology Management, Vol. 18,: 7, pp.818 - 835 Kim, S., 1995. Expansion of markets and the geographic distribution of economic activities: the trends in US regional manufacturing structure, 1860–1987. Quarterly Journal of Economics 110 (4), 881–908. Krugman, P.R. (1991a), Geography and trade. Cambridge, Mass.: MIT Press. Krugman, P.R. (1991b). Increasing returns and economic geography. Journal of political economy, 99: 483-499. Lazzeretti, L. and Capone, F. (2008), Mapping and Analysing Local Tourism Systems in Italy, 1991–2001, Tourism Geographies: An International Journal of Tourism Space, Place and Environment, 10, 2, 214-232. Lazzeretti, L. and Capone, F. (2009), Spatial Spillovers and Employment Dynamics in Local Tourist Systems in Italy (1991–2001), European Planning Studies, 17, 11, 1665-1683. Leidner, R. (2004) The European Tourism Industry: A multi-sector with dynamic markets, European Commission. Luxembourg Marcon, E., Puech, F., 2003. Evaluating the geographic concentration of industries using distance-based methods. Journal of Economic Geography 3 (4), 409–428. Marshall, A. (1890). Principles of economics, London: Mcmillan. Mazanec, J., Wöber, K., and Zins, A. (2007) Tourism destination competitiveness: From definition to explanation?, Journal of Travel Research, 46 (1), pp. 86-95. MacCarthy, B.L., Atthirawong, W. (2003) Factors affecting location decisions in international operations – a Delphi study", International Journal of Operations & Production Management, Vol. 23 Iss: 7, pp.794 - 818
#7"""
Melitz, M.J. (2003) The Impact of Trade on Intra-Industry Reallocations and Aggregate Industry Productivity, Econometrica, 71 (6), pp. 1695-1725. Michael, E.J. (2003) Tourism Micro-Clusters, Tourism Economics, 9 (2), pp. 133-145. Michaelides, Z. M., Papazian, A., 2007, Benefits of clustering for SME's suppliers in the aerospace and defence sector through the adaptation of enterprise portals, International Journal of Services and Operations Management, 3, 2, 152-171. Miller, P., Botham, R., Gibson, H., Martin, R. and Moore, B. (2001), Business Clusters in the UK – A First Assessment, Main Report, Department of Trade and Industry by a consortium led by Trends Business Research, London, UK. Miller, M.M., Henthorne, T.L., George, B.P., (2008), The Competitiveness of the Cuban Tourism Industry in the Twenty-First Century: A Strategic Re-Evaluation, Journal of Travel Research 46, 3 268-278. Miret Pastor, L., Segarra Oña, M. (2010), El papel de los Mercados Locales Laborales en la aglomeración industrial y cómo las economías externas explican la importancia del territorio, Tec Empresarial, 4, 1, 23-31. Nassimbeni, G. (2003) Local manufacturing systems and global economy: are they compatible?: The case of the Italian eyewear district, Journal of Operations Management, Volume 21, Issue 2, 151-171, Neely, A., (1999) "The performance measurement revolution: why now and what next?", International Journal of Operations & Production Management, Vol. 19 Iss: 2, pp.205 - 228 Nordin, S. (2003) Tourism clustering and innovation - Paths to economic growth and development. In Tourism clustering and Innovation, European Tourism Research Institute, Mid-Sweden University, Oestersund, Sweden. Novelli, M., Schmitz, B. and Spencer, T. (2006) Networks, clusters and innovation in tourism: A UK experience, Tourism Management, 27, pp. 1141-1152. O’Donoghue, D. and Gleave, B. (2004) A Note on Methods for Measuring Industrial Agglomeration, Regional Studies, 38 (4), pp. 419-427. O’Neill, J.W., Carlbäck, M. (2011), Do brands matter? A comparison of branded and independent hotels’ performance during a full economic cycle, International Journal of Hospitality Management, 30, 3, 515-521. Pandit, N. R., Cook, G. A. S., Beaverstock, J. V. and Ghauri, P. N. (2008) An empirical study of service sector clustering and multinational enterprises. Journal of Services Research, 7, pp. 23-39. Perrigot, R., Cliquet, G., Piot-Lepetit, I. (2009), Plural form chain and efficiency: Insights from the French hotel chains and the DEA methodology, European Management Journal,27, 268– 280. Pikkemaat, B. and Peters, M. (2005) Towards the measurement of innovation–A pilot study in the small and medium sized hotel industry. Journal of Quality Assurance in Hospitality and Tourism, 6(3-4), 89-112. Piperopoulos, P., Scase, R., 2009, Competitiveness of small and medium sized enterprises: towards a two dimensional model of innovation and business clusters International Journal of Business Innovation and Research, 3, 5, 479-499. Porter, M. (1985). Competitive advantage, The Free Press, New York, USA. Porter, M. (1998). Clusters and the new economics of competition. Harvard Business Review, Nov-Dec: 77-90. Porter M. (2003): “The economic performance of regions”, Regional Studies 37, pp 549-578. Presenza, A., Cipollina, M. (2010), Analysing tourism stakeholders networks, Tourism Review, 65(4), 17-30. Puig, F., Marques, H., Ghauri, P.N., (2009) "Globalization and its impact on operational decisions: The role of industrial districts in the textile industry", International Journal of Operations & Production Management, Vol. 29 Iss: 7, pp.692 - 719 Quarantelli, E. L. (1988). Disaster crisis management–a summary of research findings. Journal of Management Studies, 25(4), 373-385 Richards, G. and Wilson, J. (2006), Developing Creativity in Tourist Experiences: A Solution to the Serial Reproduction of Culture?, Tourism Management , 27, 1209-1220. Rüdiger, L. (2004), The European Tourism Industry. A multi-sector with dynamic markets. Structures, developments and importance for Europe’s economy, Office for Official Publications of the European Communities Luxembourg, 2004. Sapir, A., 1996. The effects of Europe's internal market program on production and trade: a first assessment. Weltwirtschaftliches Archiv 132, 457–475.
#8"""
Saxena, G (2005) Relationships, networks and the learning regions: case evidence from de Peak District National Park, Tourism Management, 26, pp. 277-289. Saxenian, A. (1996): Regional advantage: culture and competition in Sillicon Valley and Route 128. Cambridge, MA: Harvard University Press. Scott, A. (1991): “The aerospace-electronics industrial complex of Southern California: The formative years, 1940-1960.” Research Policy, vol. 20, n. 5, pp. 439-456. Scott, N., Laws, E., Prideaux, B., (2008) Tourism Crises and Marketing Recovery Strategies, Journal of Travel & Tourism Marketing, 23, 2-4, 1-13. Secondi, L., Meseguer-Santamaría, M., Mondéjar-Jiménez, J., Vargas-Vargas, M. (2011), Influence of tourist sector structure on motivations of heritage tourists, The Service Industries Journal, 31 (10): 1659-1668. Segarra-Oña, M., De-Miguel-Molina, B. (2009), Evaluation of the synergistic industrial concentration: methodological proposal and application to an industrial sector (in Spanish), TEC Empresarial, 3 (1-2), 65-72. Segarra-Oña, M., Miret-Pastor, L., Peiró-Signes, Á., Verma, R. (2012) The effects of localization on economic performance. Analysis of Spanish tourism clusters. European Planning Studies, Forthcoming. Sharma, A., Sneed, J. and Ravichandran, S. (2007) Spatial Analysis of Small Hotel Activity in Tanzania, International Journal of Contemporary Hospitality Management, 19 (7), pp. 589-599. Sheehan, L., Ritchie, J. R. B., Hudson, S. (2007), The Destination Promotion Triad: Understanding Asymmetric Stakeholder Interdependencies Among the City, Hotels, and DMO, Journal of Travel Research, vol. 46 no. 1 64-74 <,-(=">?="@15-3+0="A?"B#;;9C=">)*D)*'"E/0)*-00"(*F"G(3D-+)*'"<+3(+-')-0="H1/3*(I"1J"K10L)+(I)+6"M">-)0/3-"G(3D-+)*'"="$="!="47-64. Signorini, L. F. (1994) The price of Prato, or measuring the industrial district effect, Papers in Regional Science, 73 (4), pp. 369-392. Smeral, E. (2010), Impacts of the World Recession and Economic Crisis on Tourism: Forecasts and Potential Risks, Journal of Travel Research, 49 (1), 31-38. So, H.W.T., Gunasekaran, A., g, W.W.C., 2006, A development of business agility enterprise model for competitive advantage: theory and practice International Journal of Services and Operations Management, 2, 1, 42-59 Spencer, G. M., Vinodrai, T., Gertler, M. S., Wolfe, D. A. (2010), Do Clusters Make a Difference? Defining and Assessing their Economic Performance, Regional Studies, 44, 6, 697-715. Tallman, S, Jenkins, M., Henry, N. Y Pinch, S. (2004): “Knowledge, Clusters and Competitive Advantage”. Academy of Management Review, 29, pp 258-271. Tinsley, R. and Lynch, P. (2001) Small Tourism Business Networks and Destination Development, International Journal of Hospitality Management, 20 (4), pp. 367-378. UNWTO (2012). Tourism Trends and Marketing Strategies: UNWTO Tourism Highlights, 2011 Edition. www.unwto.org/facts Urtasun, A., Gutiérrez, I. (2006), Hotel location in tourism cities: Madrid 1936–1998, Annals of Tourism Research, 33, 2, 382-402 Van Den Berg, L., Braum, E. and Van Winden, W. (2001) Growth cluster in European Metropolitan cities, Urban Studies, 38,1, 185-205. Vargas-Vargas, M., Meseguer-Santamaría, M.L., Mondéjar-Jiménez, J. Mondéjar-Jiménez, J.A. (2010), Environmental protection expenditure for companies: A Spanish regional analysis, International Journal of Environmental Research, 4 (3), pp. 373-378. Von Friedrichs Y., Gummesson, E. (2006) "Hotel networks and social capital in destination marketing", International Journal of Service Industry Management, Vol. 17 Iss: 1, pp.58 - 75 Voss, C., Roth, A. V. and Chase, R. B. (2008), Experience, Service Operations Strategy, and Services as Destinations: Foundations and Exploratory Investigation. Production and Operations Management, 17: 247–266. Weindenfeld, A. Butler, R. and Williams, A. (2010): “Clustering and Compatibility between Tourism Attractions” International Journal of Tourism Research. Vol 12. Iss 1. Pp 1-16 Weidenfeld, A., Butler, R. and Williams, A. W. (2011), The role of clustering, cooperation and complementarities in the visitor attraction sector, Current Issues in Tourism, 14, 7, 595-629 Wilhelm, M. M., 2011, Managing coopetition through horizontal supply chain relations: Linking dyadic and network levels of analysis, Journal of Operations Management, 29, 7–8, 663-676. Wright, D.T., Burns, N.D., (1998), New organisation structures for global business: an empirical study, International Journal of Operations & Production Management, 18, 9/10, 896-923.
#9"""
Yang Y., Wong, K.F., Wang, T. (2012), How do hotels choose their location? Evidence from hotels in Beijing, International Journal of Hospitality Management, http://dx.doi.org/10.1016/j.ijhm.2011.09.003, (http://www.sciencedirect.com/science/article/pii/S0278431911001472) Yin, Y., Liu, C., Kaku, I. (2011), Cooperation and leadership policies in a serial supply chain, Journal of Manufacturing Systems, 30, 1, 1-7.
!"#$%&'(&)#*+,-"$&./"*0%1&2*&3/%&4%,5+,-"*.%&,%16$31&5+,&3/%&3+3"$&1"-4$%(&!!
"#$%&'$!()!*+)(,-*&!./*)0#$1!23$'*'4$'4.$!5(,!'/#!-#*)6!74&.(8()!$40)#93,*):!'#$'!;3$'*'4$'4.!5(,!'/#!-#94*)6!*)9!+4)(-4*&!$40)!'#$'!;3$'*'4$'4.!5(,!'/#!<#,.#)'!<($4'4=#!*,#!,#<(,'#9!4)!<*,#)'/#$#$1!*!>40)454.*)'&?!9455#,#)'!5,(-!@#,(!ABCD!4)!'/#!.*$#!(5!<#,.#)'!<($4'4=#E!*'!'/#!FD!&#=#&!5(,!()#3'*4	!'#$'1!!+!>40)454.*)'&?!9455#,#)'!5,(-!@#,(!ABCD!4)!'/#!.*$#!(5!<#,.#)'!<($4'4=#E!*'!'/#!G1BD!&#=#&!5(,!()#3'*4	!'#$'1!!.!>40)454.*)'&?!9455#,#)'!5,(-!@#,(!ABCD!4)!'/#!.*$#!(5!<#,.#)'!<($4'4=#E!*'!'/#!BD!&#=#&!5(,!()#3'*4	!'#$'1!!!!!!!
2007-2008 2007-2009 2007-2010 2007-2011 Cum.2009 Cum.2010 Cum.2011
Mean Median % pos. Mean Median % pos. Mean Median % pos. Mean Median % pos. Mean Median % pos. Mean Median % pos. Mean Median % pos.
N 4313 4311 4321 4341 4311 4311 4311 Abnormal
change in the % of demand
0,02 -1,04 46,49 0,72 0,03 50,03 0,77 -0,32 49,22 0,37 -0,48 48,79 0,28 -0,40 48,43 0,41 -0,33 48,55 0,40 -0,55 48,20
(0,02) (-4,53)a (-4,6)a (0,92) (-0,57) (-0,03) (0,9) (-0,91) (-1) (0,4) (-1,61) (-1,58) (0,39) (-1,82)c (-2,04)b (0,54) (-1,88)c (-1,89)c (0,51) (-2,17)b (-2,35)a
Abnormal change in the % of revenue
-0,07 -1,50 44,77 0,54 -0,43 49,06 1,49 -0,32 49,41 2,94 0,57 51,05 0,11 -0,83 47,74 0,51 -0,88 47,78 1,09 -0,42 48,78
(-0,08) (-6,44)a (-6,85)a (0,59) (-0,81) (-1,22) (1,46) (-0,22) (-0,76) (2,62)a (-2,58)a (-1,37) (0,13) (-3,64)a (-2,95)a (0,57) (-2,35)a (-2,89)a (1,16) (-0,84) (-1,58)
Abnormal change in the
occ
-0,59 -0,59 46,64 0,09 0,08 50,28 -0,10 -0,17 49,17 -0,35 -0,09 49,61 -0,25 -0,34 48,68 -0,21 -0,23 48,84 -0,24 -0,31 48,56
(-4,36)a (-5,31)a (-4,4)a (0,55) (-1,02) (-0,35) (-0,53) (-0,48) (-1,08) (-1,69)c (-1,27) (-0,5) (-1,84)c (-1,96)b (-1,72)c (-1,42) (-1,7)c (-1,51) (-1,59) (-1,84)c (-1,87)c
Abnormal change in the
ADR
-0,07 -0,55 45,41 -0,10 -0,58 47,01 0,15 0,04 50,21 1,53 0,81 52,70 -0,11 -0,57 46,48 -0,04 -0,32 48,17 0,35 -0,01 49,93
(-0,22) (-3,97)a (-6,01)a (-0,3) (-3,4)a (-3,91)a (0,41) (-0,12) (-0,26) (3,94)a (-5,76)a (-3,54)a (-0,36) (-3,69)a (-4,61)a (-0,11) (-2,21)b (-2,39)a (1,07) (-0,53) (-0,08)
Abnormal change in the
revPAR
-0,70 -0,75 45,25 -0,13 -0,16 49,30 0,28 0,04 50,23 1,46 0,27 51,04 -0,43 -0,47 47,45 -0,19 -0,34 48,26 0,22 -0,21 48,82
(-3,05)a (-6,17)a (-6,23)a (-0,55) (-0,03) (-0,9) (1,05) (-1,21) (-0,29) (4,54)a (-3,86)a (-1,35) (-1,95)c (-2,75)a (-3,33)a (-0,86) (-1,33) (-2,27)b (0,94) (-0,64) (-1,54)
!"#$%&7(&)#*+,-"$&./"*0%1&2*&3/%&4%,5+,-"*.%&,%16$31&1%0-%*3%8	&/+3%$&$%:%$(&!
!2007-2008 2007-2009 2007-2010 2007-2011 Cum.2009 Cum.2010 Cum.2011
!
Mean Median % pos. Mean Median % pos. Mean Median % pos. Mean Median % pos. Mean Median % pos. Mean Median % pos. Mean Median % pos.
"#$#%&! !! '()! !! !! '()! !! !! '(*! !! !! '(+! !! !! '()! !! !! '()! !! !! '()! !!
!! ,-./! ,-01! .1-0+! !"#$% !"$&% '(")'% $"#$% $"$!% ')"*+% )"#'% '"#+% (&"#$% 1-01! #"+&% ''"!'% #",(% !"&,% '(")'% $"!,% !"+)% '+"&$%
!! 2,-.+3! 24,-0)3! 24,-.(3! -#",#./% -0!").1% -0!"+).1% -!"'*.1% -0$"!*.1% -0$"#,.1% -*"++.1% -0*",,.1% -0*"$*.1% 21-'03! -0!"&).2% -0!"!!.2% -!"&*.2% -0!").1% -0!"+).1% -$"#,.1% -0$"($.1% -0$"*!.1%
56789:;! !! 1)()! !! !! 1)((! !! !! 1)(+! !! !! 1)).! !! !! 1)((! !! !! 1)((! !! !! 1)((! !!
!! 4,-(.! !"#$%& $%#'$& 4,-1,! ,-.,! '(#"$& ,-0+! "#)*& '+#$(& #")*% #"*,% '*"*!% 4,-',! 4,-,)! '+-*,! 4,-1'! ,-0,! .1-(1! ,-00! &")(% '!"')%
!! 241-'3! ,!(#*+-.& ,!(#"'-.& 24,-/'3! 241-'03! ,!*#/)-0& 2,-*13! ,!(#$*-1& ,!(#2'-1& -$"&$.1% -0*"$#.1% -0$").1% 24,-+.3! 24,-,*3! 24,-1'3! 24,-0'3! 24,-+3! 241-003! 2,-)03! -0!"$(.1% -0!"#*.2%
<=>6%=8;! !! 10'0! !! !! 100*! !! !! 10'0! !! !! 10'+! !! !! 10'0! !! !! 10'0! !! !! 10'0! !!
!! !"#)+& !*#**& $(#*$& 4,-0*! !"#//& $'#)/& 4,-'1! !*#"*& $'#*(& ,-,.! 4,-+,! $/#"+& !"#//& !"#2(& $$#%'& !"#'2& !*#"*& $+#/+& 4,-'/! !*#"2& $$#"*&
!! ,!$#"$-1& ,!'#2+-1& ,!'#%+-1& 241-/+3! ,!(#('-.& ,!(#)+-1& 241-/'3! ,!(#/)-1& ,!+#''-1& 2,-1/3! 241-13! ,!(#2)-1& ,!(#/)-1& ,!$#"/-1& ,!+#2(-1& ,!(#(*-.& ,!+#2-1& ,!$#/$-1& 241-')3! ,!+#"2-1& ,!$#+%-1&
?8@A@B&! !! ''*! !! !! './! !! !! '.0! !! !! '..! !! !! '')! !! !! '')! !! !! '')! !!
!! !*#$$& !*#$$& $"#/+& !*#*(& !"#%"& $$#"+& 4,-'/! 4,-00! ')-,/! ,-/.! 4,-*1! '(-.+! !*#(2& !*#('& $"#$"& !"#))& !"#2$& $(#/+& !"#/2& !"#)(& $(#2/&
!! ,!$#$/-1& ,!$#2)-1& ,!+#)(-1& ,!(#%$-1& ,!(#$2-1& ,!(#$)-1& 24,-**3! 241-103! 241-//3! 2,-'3! 24,-(*3! 241-'13! ,!+#2'-1& ,!+#)(-1& ,!$#"(-1& ,!(#2*-1& ,!(#)-1& ,!+#"%-1& ,!*#%$-0& ,!(#"2-.& ,!(#)2-1&
C#>D;E! !! /**! !! !! /**! !! !! /*(! !! !! /+1! !! !! /**! !! !! /**! !! !! /**! !!
!! 4,-),! !"#/"& $+#(*& !*#()& !"#%)& $$#/"& 4,-)/! 4,-0+! ')-0)! 41-,,! !"#)%& '.-./! !*#""& !"#'/& $$#('& !"#)(& !"#$)& $$#/"& !"#)/& !"#''& $(#'*&
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	!'#$'1!!+!>40)454.*)'&?!9455#,#)'!5,(-!@#,(!ABCD!4)!'/#!.*$#!(5!<#,.#)'!<($4'4=#E!*'!'/#!G1BD!&#=#&!5(,!()#3'*4	!'#$'1!!.!>40)454.*)'&?!9455#,#)'!5,(-!@#,(!ABCD!4)!'/#!.*$#!(5!<#,.#)'!<($4'4=#E!*'!'/#!BD!&#=#&!5(,!()#3'*4	!'#$'1!H4,$'!=*&%#!4)94.*'#9!,#5#,$!'(!'/#!)%-+#,!(5!<,(<#,'4#$!4)!'/#!$#0-#)'1!!!!!!!!
!"#$%&;(&)#*+,-"$&./"*0%1&2*&3/%&4%,5+,-"*.%&,%16$31&1%0-%*3%8	&"552$2"32+*(&!
!2007-2008 2007-2009 2007-2010 2007-2011 Cum.2009 Cum.2010 Cum.2011
!
Mean Median % pos. Mean Median % pos. Mean Median % pos. Mean Median % pos. Mean Median % pos. Mean Median % pos. Mean Median % pos.
FG9=A!<9A9D;B;AE! !! .1)! !! !! .1)! !! !! .1)! !! !! .1)! !! !! .1)! !! !! .1)! !! !! .1)! !!
!! ,-,'! 4,-(/! ')-,,! 4,-1.! 4,-10! '+-10! ,-.'! ,-/1! .,-(*! $"#*% #"$$% .0-,,! 4,-,(! 4,-0'! '*-1(! ,-1'! 4,-/.! '*-)'! ,-*+! ,-1)! .,-'*!
!! 2,-,(3! 24,-//3! 241-0/3! 24,-1*3! 24,-/3! 24,-0.3! 2,-.*3! 24,-+03! 24,-/(3! -!"),.1% -0$"&#.1% 241-0/3! 24,-,*3! 24,-/(3! 24,-)+3! 2,-1+3! 24,-/+3! 24,-.03! 21-1/3! 241-/03! 24,-1*3!
H%9A8G=7;! !! 0/1*! !! !! 0/1(! !! !! 0///! !! !! 0/',! !! !! 0/1(! !! !! 0/1(! !! !! 0/1(! !!
!! !"#2(& !"#22& $+#2/& 4,-,)! 4,-0,! '*-))! ,-,*! 4,-1)! '+-1.! &"+'% 4,-10! '+-.*! !"#$'& !"#'/& $/#/"& 4,-/+! !"#$/& $%#")& ,-,,! 4,-',! $%#$$&
!! ,!$#)'-1& ,!%#/+-1& ,!/#)'-1& 24,-0.3! 24,-.)3! 241-0*3! 2,-0(3! 24,-/'3! 24,-+.3! -$"#.1% 241-0/3! 24,-'(3! ,!(#/(-1& ,!+#2$-1& ,!+#2$-1& 241-.(3! ,!(#/*-1& ,!+#(2-1& 24,-,13! 241-0'3! ,!(#2)-1&
IA>;6;A>;AE! !! .)*! !! !! .)*! !! !! .*,! !! !! .*/! !! !! .)*! !! !! .)*! !! !! .)*! !!
!! 4,-(.! ,-/0! .1-.(! 4,-'.! ,-..! ./-(,! 1-10! !"!$% ''"(,% $"$'% $",(% ')"!!% 4,-.+! ,-.'! .1-)0! ,-,0! 1-1*! '*"'&% ,-*+! !"&(% ''"&!%
!! 24,-.13! 24,-''3! 24,-)13! 24,-0+3! 24,-)13! 241-/13! 2,-*+3! -0!"$.2% -0!").1% -!"#(.2% -0$"($.1% -0$"**.1% 24,-./3! 24,-(/3! 24,-)+3! 2,-,03! 241-'03! -0!"#!.2% 2,-)03! -0!"($.1% -0!"$).1%"#$%&'$!()!*+)(,-*&!./*)0#$1!23$'*'4$'4.$!5(,!'/#!-#*)6!74&.(8()!$40)#93,*):!'#$'!;3$'*'4$'4.!5(,!'/#!-#94*)6!*)9!+4)(-4*&!$40)!'#$'!;3$'*'4$'4.!5(,!'/#!<#,.#)'!<($4'4=#!*,#!,#<(,'#9!4)!<*,#)'/#$#$1!*!>40)454.*)'&?!9455#,#)'!5,(-!@#,(!ABCD!4)!'/#!.*$#!(5!<#,.#)'!<($4'4=#E!*'!'/#!FD!&#=#&!5(,!()#3'*4	!'#$'1!!+!>40)454.*)'&?!9455#,#)'!5,(-!@#,(!ABCD!4)!'/#!.*$#!(5!<#,.#)'!<($4'4=#E!*'!'/#!G1BD!&#=#&!5(,!()#3'*4	!'#$'1!!.!>40)454.*)'&?!9455#,#)'!5,(-!@#,(!ABCD!4)!'/#!.*$#!(5!<#,.#)'!<($4'4=#E!*'!'/#!BD!&#=#&!5(,!()#3'*4	!'#$'1!H4,$'!=*&%#!4)94.*'#9!,#5#,$!'(!'/#!)%-+#,!(5!<,(<#,'4#$!4)!'/#!$#0-#)'1!!
!!!!!!!!!!!!
!"#$%&<(&)#*+,-"$&./"*0%1&2*&3/%&4%,5+,-"*.%&,%16$31&1%0-%*3%8	&$+."32+*(&!
!2007-2008 2007-2009 2007-2010 2007-2011 Cum.2009 Cum.2010 Cum.2011
!
Mean Median % pos. Mean Median % pos. Mean Median % pos. Mean Median % pos. Mean Median % pos. Mean Median % pos. Mean Median % pos.
5%J9A! !! 0,*! !! !! 010! !! !! 01'! !! !! 01'! !! !! 0,*! !! !! 0,*! !! !! 0,*! !!
!! !"+#% *"#*% ('",#% *"($% '"(*% (*"'*% )"&&% (",*% (,"*$% #!"&$% #$"$&% )*"+*% $"(,% *"'(% (("++% *"+!% '"*'% (("++% ("($% ("+(% )#")'%
!! -$"(.1% -0'"#).1% -0'"'$.1% -'"''.1% -0'",!.1% -0'"&,.1% -(",).1% -0)"*.1% -0("+$.1% -#&"$,.1% -0,").1% -0+")'.1% -'"&'.1% -0'")+.1% -0'"+).1% -("#'.1% -0("($.1% -0'"+).1% -+"&#.1% -0+"&!.1% -0)"'+.1%
K#J#%J9A! !! ++.! !! !! ++0! !! !! ++.! !! !! ++*! !! !! ++.! !! !! ++.! !! !! ++.! !!
!! !"#2'& !*#$/& +2#/)& !"#%/& !"#)"& $$#**& !*#+)& !*#+'& $*#)*& !"#2%& !*#(/& $+#))& !"#2*& !*#(/& $"#+"& !*#"*& !*#('& $"#)"& !"#)/& !*#+$& $"#'"&
!! ,!+#*-1& ,!/#2*-1& ,!%#*-1& ,!(#$(-1& ,!+#''-1& ,!+#/2-1& ,!+#/%-1& ,!'#"/-1& ,!'#"%-1& ,!*#)$-0& ,!+#'/-1& ,!+#%%-1& ,!+#"$-1& ,!'#+$-1& ,!/#")-1& ,!+#$)-1& ,!'#'-1& ,!'#%*-1& ,!+#"+-1& ,!$#))-1& ,!'#)/-1&
L=%6@%E! !! 10)! !! !! 10(! !! !! 10)! !! !! 10)! !! !! 10)! !! !! 10)! !! !! 10)! !!
!! !(#+(& !(#/)& +'#%%& 41-/1! 41-0.! ''-*.! !(#"+& !*#'%& $"#22& !(#*'& !*#2/& '0-,)! !*#%)& !(#$'& +)#$(& !*#2%& !(#'*& +%#(+& !*#)$& !*#)"& +2#/)&
!! ,!+#$*-1& ,!+#2-1& ,!+#('-1& 241-'(3! 241-(/3! 241-113! ,!(#(-.& ,!(#$/-1& ,!(#"'-.& ,!(#*)-.& ,!(#"$-.& 241-.'3! ,!(#'$-1& ,!(#))-1& ,!(#+)-1& ,!(#'+-1& ,!(#)2-1& ,!(#)-1& ,!(#'+-1& ,!(#%/-1& ,!(#'/-1&
IAE;%7E9E;! !! )/+! !! !! )/+! !! !! )01! !! !! )0.! !! !! )/*! !! !! )/*! !! !! )/*! !!
!! !"#)%& !"#//& $$#+*& !"#2+& !"#/"& $'#2(& 4,-'*! 4,-/.! '*-.(! 4,-/'! 4,-'0! '*-0,! !"#2$& !"#'/& $'#*+& !"#%(& !"#$$& ')-,.! !"#')& 4,-./! $/#+/&
!! ,!+#)'-1& ,!$#"$-1& ,!+#"$-1& ,!(#$'-1& ,!(#"/-.& ,!(#((-.& 241-//3! 24,-)13! 24,-)'3! 24,-.3! 24,-0+3! 24,-*+3! ,!+#*$-1& ,!(#))-1& ,!(#')-1& ,!(#$$-1& ,!(#*(-.& 241-.(3! ,!*#2+-0& 241-(/3! ,!*#)+-0&
M;7@%E! !! +0(! !! !! +0.! !! !! +0.! !! !! +'1! !! !! +0(! !! !! +0(! !! !! +0(! !!
!! !*#2/& !"#2)& $'#/(& 4,-*,! ,-)(! ./-0,! 4,-1(! ,-),! ./-0,! 1-,.! 4,-''! '+-/,! 41-0'! 4,-/,! '+-(*! 4,-+0! ,-1*! .,-*.! 4,-'1! ,-11! .,-'0!
!! ,!(#*-.& ,!(#)2-1& ,!(#/'-1& 24,-+03! 24,-,)3! 241-0)3! 24,-1)3! 24,-()3! 241-0)3! 2,-+.3! 24,-'3! 24,-'(3! 241-(/3! 2413! 24,-1(3! 241-1/3! 24,-/+3! 24,-'+3! 24,-'*3! 24,-1.3! 24,-/03!
KB9::!<;E%@NO@PA! !! 1/,*! !! !! 1/,.! !! !! 1/,)! !! !! 1/1'! !! !! 1/,)! !! !! 1/,)! !! !! 1/,)! !!
!! 4,-//! !"#$)& $/#2*& ,-/1! ,-,1! .,-,,! &",)% &"'#% '!"*,% !"$,% #"*'% '*"'+% 4,-,'! 4,-/.! '+-,.! ,-/*! 4,-,*! '+-(0! &"+&% &"*#% .1-/,!
!! 24,-*13! ,!(#/%-1& ,!(#*)-.& 2,-.+3! 24,-(*3! ,-,,! -!"'*.1% -0!"$!.2% -0#")./% -*"+.1% -0*")$.1% -0$"#(.1% 24,-1(3! 24,-)*3! 24,-(03! 2,-+.3! 24,-0)3! 24,-/03! -!"*$.1% -0#",$./% 24,-*13!"#$%&'$!()!*+)(,-*&!./*)0#$1!23$'*'4$'4.$!5(,!'/#!-#*)6!74&.(8()!$40)#93,*):!'#$'!;3$'*'4$'4.!5(,!'/#!-#94*)6!*)9!+4)(-4*&!$40)!'#$'!;3$'*'4$'4.!5(,!'/#!<#,.#)'!<($4'4=#!*,#!,#<(,'#9!4)!<*,#)'/#$#$1!*!>40)454.*)'&?!9455#,#)'!5,(-!@#,(!ABCD!4)!'/#!.*$#!(5!<#,.#)'!<($4'4=#E!*'!'/#!FD!&#=#&!5(,!()#3'*4	!'#$'1!!+!>40)454.*)'&?!9455#,#)'!5,(-!@#,(!ABCD!4)!'/#!.*$#!(5!<#,.#)'!<($4'4=#E!*'!'/#!G1BD!&#=#&!5(,!()#3'*4	!'#$'1!!.!>40)454.*)'&?!9455#,#)'!5,(-!@#,(!ABCD!4)!'/#!.*$#!(5!<#,.#)'!<($4'4=#E!*'!'/#!BD!&#=#&!5(,!()#3'*4	!'#$'1!H4,$'!=*&%#!4)94.*'#9!,#5#,$!'(!'/#!)%-+#,!(5!<,(<#,'4#$!4)!'/#!$#0-#)'1!!!!!
!!"#$%&=>(&)#*+,-"$&./"*0%1&2*&3/%&4%,5+,-"*.%&,%16$31&1%0-%*3%8&.$613%,&.+*.%*3,"32+*(&!!
!2007-2008 2007-2009 2007-2010 2007-2011 Cum.2009 Cum.2010 Cum.2011
!Mean Median % pos. Mean Median % pos. Mean Median % pos. Mean Median % pos. Mean Median % pos. Mean Median % pos. Mean Median % pos.
"@P!8@A8;AE%9E;>! !! 0,'0! !! !! 0,'(! !! !! 0,.,! !! !! 0,('! !! !! 0,'1! !! !! 0,'1! !! !! 0,'1! !!
!! 4,-/(! !"#//& $'#/"& ,-'1! ,-,1! .,-,.! "#%+& ,-,(! .,-'*! (#($& "#/%& '(#+%& ,-,)! 4,-/*! $2#*+& ,-/+! 4,-/'! '*-(.! "#%2& 4,-,)! '+-.)!
!! 241-'/3! ,!$#"%-1& ,!$#2$-1& 21-*/38! 241-,.3! 24,-,'3! ,(#2$-1& 241-'13! 24,-.13! ,/#2$-1& ,!'#(+-1& ,!(#/-1& 2,-0*3! 241-/*3! ,!(#"'-.& 21-''3! 24,-0(3! 241-')3! ,+#'$-1& 241-(3! 24,-'.3!
<;>=#B!F@A8;AE%9E;>! !! *+,! !! !! **+! !! !! *+,! !! !! *+.! !! !! *+,! !! !! *+,! !! !! *+,! !!
!! !*#$$& !*#"2& $+#%*& !*#$(& !"#/*& $/#'%& 4,-.0! 4,-/,! '+-)*! 4,-1.! 4,-,(! '+-*0! !*#$'& !"#)/& $$#*/& !*#*'& !"#%$& $/#'(& !"#)*& 4,-(0! $/#%$&
!! ,!+#%+-1& ,!$#%$-1& ,!+#%(-1& ,!(#)'-1& ,!*#)+-0& ,!(#"*-.& 241-,13! 24,-1(3! 24,-13! 24,-/03! 24,-1.3! 24,-,)3! ,!+#'%-1& ,!+#((-1& ,!+#$'-1& ,!(#%$-1& ,!(#")-.& ,!(#"$-.& ,!*#))-.& 241-/13! ,!*#)*-0&
Q=DG!F@A8;AE%9E;>! !! 0*,! !! !! 0)(! !! !! 0)+! !! !! 0*,! !! !! 0*,! !! !! 0*,! !! !! 0*,! !!
!! 4/-''! !"#%/& '(-0/! 41-'+! 4,-,'! .,-,,! 41-'0! 4,-1)! '+-,*! 41-,(! 4/-.)! $(#2)& 4/-,,! 4,-,/! .,-,,! 41-)*! 4,-1'! '+-')! 41-.*! 4,-.*! ')-(0!
!! 241-/)3! ,!*#)'-0& 241-0+3! 24,-**3! 24,-//3! ,-,,! 24,-)*3! 24,-/3! 24,-013! 24,-./3! 241-,/3! ,!(#%(-1& 241-1(3! 24,-*03! ,-,,! 241-,03! 24,-0*3! 24,-1.3! 24,-+3! 24,-'.3! 24,-*)3!"#$%&'$!()!*+)(,-*&!./*)0#$1!23$'*'4$'4.$!5(,!'/#!-#*)6!74&.(8()!$40)#93,*):!'#$'!;3$'*'4$'4.!5(,!'/#!-#94*)6!*)9!+4)(-4*&!$40)!'#$'!;3$'*'4$'4.!5(,!'/#!<#,.#)'!<($4'4=#!*,#!,#<(,'#9!4)!<*,#)'/#$#$1!*!>40)454.*)'&?!9455#,#)'!5,(-!@#,(!ABCD!4)!'/#!.*$#!(5!<#,.#)'!<($4'4=#E!*'!'/#!FD!&#=#&!5(,!()#3'*4	!'#$'1!!+!>40)454.*)'&?!9455#,#)'!5,(-!@#,(!ABCD!4)!'/#!.*$#!(5!<#,.#)'!<($4'4=#E!*'!'/#!G1BD!&#=#&!5(,!()#3'*4	!'#$'1!!.!>40)454.*)'&?!9455#,#)'!5,(-!@#,(!ABCD!4)!'/#!.*$#!(5!<#,.#)'!<($4'4=#E!*'!'/#!BD!&#=#&!5(,!()#3'*4	!'#$'1!H4,$'!=*&%#!4)94.*'#9!,#5#,$!'(!'/#!)%-+#,!(5!<,(<#,'4#$!4)!'/#!$#0-#)'1!!!