Content uploaded by Oscar Claveria
Author content
All content in this area was uploaded by Oscar Claveria
Content may be subject to copyright.
TOURISMOS: AN INTERNATIONAL MULTIDISCIPLINARY JOURNAL OF TOURISM
Volume 4, Number 1, Spring 2009, pp. 13-28
13
TOURISM DEMAND IN CATALONIA: DETECTING
EXTERNAL ECONOMIC FACTORS
Oscar Claveria1
University of Barcelona
Jordi Datzira
Datzira Development Services, S.L. (Dds)
There is a lack of studies on tourism demand in Catalonia. To fill the gap, this
paper focuses on detecting the macroeconomic factors that determine tourism
demand in Catalonia. We also analyse the relation between these factors and
tourism demand. Despite the strong seasonal component and the outliers in the
time series of some countries, overnight stays give a better indication of tourism
demand in Catalonia than the number of tourists. The degree of linear association
between the macroeconomic variables and tourism demand is also higher when
using quarterly rather than monthly data. Finally, there are notable differences
between the results obtained for the different countries analysed. These results
indicate that the best way to model tourism demand in Catalonia is to specify a
quarterly model of overnight stays, differentiating between an aggregate demand
model for the total number of tourists and specific models for each of the
countries analysed.
Keywords: tourism demand in Catalonia, external economic factors,
expectations, descriptive analysis
INTRODUCTION
Catalonia is one of the seventeen autonomous communities in Spain.
It is located in the north-east of the Iberian Peninsula, and its capital is
Barcelona. It has a population of over seven million, which represents
16% of the total population of Spain. Catalonia is a tourist region, as
shown by the large number of tourist products and attractions and by the
increasing number of visitors each year. Tourism in Catalonia accounts
for 12% of GDP and provides employment for around 19% of the
working population in the service sector. Over 14 million foreign visitors
© University of the Aegean. Printed in Greece. Some rights reserved. ISSN: 1790-8418
Oscar Claveria & Jordi Datzira
14
come to Catalonia every year, leading to 111 million overnight stays and
an estimated tourist expenditure of around 9 billion euros.
Tourism is generated by demand. Therefore, it is important to
identify the factors that determine demand (at a tourist level). The study
of aggregate tourism demand provides in-depth information about tourist
flows. It also helps the making of business decisions and the drawing up
of tourist policies.
Although studies have been undertaken for other countries, to date,
there have been no analyses of tourism demand in Catalonia. The
literature on tourism demand has developed in two areas simultaneously.
The first area involves theoretical contributions, which are the basis of
tourism demand studies. Such studies focus on defining tourism demand
and on analysing its main characteristics. The main contributions to the
economics of tourism include papers by Figuerola (1985), Bull (1994),
Smith (1995) and De Rus and León (1997).
The second area involves different empirical studies, such as those of
Crouch (1994a,b) and Lim (1997). Crouch (1994a,b) carried out a meta-
analysis of eighty empirical studies to analyse the variables used to
explain tourism demand, and the estimated values of the corresponding
elasticity. The work of Witt (1995) is also notable in this area. Lim (1997)
extensively reviewed the empirical literature from a more methodological
perspective. In addition, pioneering contributions have been made by
O’Hagan and Harrison (1984), Uysal and Crompton (1985) and Witt and
Martin (1987). Padilla (1988) was the first to draw up indices of income
and prices to give an approximation of the relative importance of different
foreign countries in the demand for Spanish tourism services.
Other more recent contributions include papers by González and
Moral (1995) and Buisán (1997). González and Moral (1995) were the
first to use structural time series models in this environment. Buisán
(1997) carried out a cointegration analysis. Garín-Muñoz and Amaral
(2000), and Ledesma-Rodríguez et al. (2001) used panel data from Spain
and Tenerife respectively to estimate a model of tourism demand
segmented by nationality.
This brief literature review shows that there is a lack of studies on
Catalonia. Thus, the aim of this paper is to detect and analyse the factors
that determine tourism demand in Catalonia. In Section 2, we define what
we understand by tourism demand and determine how to measure it.
External factors that determine tourism demand are identified and a
database is compiled. Section 3 contains a descriptive and correlations
analysis of both tourism demand in Catalonia and the external factors that
TOURISMOS: AN INTERNATIONAL MULTIDISCIPLINARY JOURNAL OF TOURISM
Volume 4, Number 1, Spring 2009, pp. 13-28
15
could explain this demand. Finally, the main conclusions are set out in
Section 4.Start each new paragraph with indent like this.
TOURISM DEMAND IN CATALONIA
Aggregate analysis of tourism demand emerged in the sixties, from
the single equation models devised by Gerakis (1965). Since then, the
theoretical specification of these models has not altered. However, the
econometric techniques used in empirical applications have changed,
becoming increasingly sophisticated. Nevertheless, there are inherent
limitations in the analysis of aggregate data. As indicated by Morley
(1994), individual decisions are diluted in aggregated information.
Different variables are used to measure tourism demand: expenditure
at the destination (which is equivalent to tourism income); the number of
visitors, arrivals or overnight stays. As indicated by De Rus and León
(1997), the choice of one or other variable largely depends on the
availability of data and the aims of the study. Frequently, the dependent
variable is transformed into per capita terms and scaled up according to
the population of origin, as the size of the market determines the total
number of visitors.
Crouch (1994a, b) observed that the number of visitors was the
dependent variable in 70% of the studies analysed. This is due to the fact
that most cases consisted of specific studies between the destination and
the country of origin. The remaining studies used tourism income as the
dependent variable, apart from a few that used length of stay.
When using the number of visitors as a proxy of aggregate tourism
demand, it is usual to differentiate between resident visitors and non-
resident or foreign visitors. The numbers of foreign visitors are measured
by movements at the border. However, such a measure is only feasible
when analysing tourism demand at state level. In addition, Catalonia
receives a notable number of day trippers (i.e. people who come to
Catalonia but do not stay the night), either because the region is close to
their place or residence or because they are just passing through.
In this study, we used both the number of tourists and the number of
overnight stays (first destinations)1
It is widely accepted in both theoretical and empirical literature that
income and the price of tourism services have a major influence on
tourism demand. Income acts as a restriction: until a certain level or
disaggregated by countries. Data
includes the number of tourists arriving from each visitor market from
January 1997 to August 2005; and the number of overnight stays,
according to visitor country, from January 2001 to July 2005.
Oscar Claveria & Jordi Datzira
16
threshold has been reached, tourism consumption cannot begin. Over
time, the critical level of tourism consumption that is linked to this
threshold has decreased. Nevertheless, this process is asymmetrical: once
tourism consumption has been consolidated, it becomes difficult to give it
up. The consumer begins to consume other products, changing the
destination if necessary. Thus, Figuerola (1992) indicated that income has
a greater influence on the budget allocated to tourism consumption than
on the decision to embark on a trip, as individuals adapt their activities, in
terms of time, expenditure, destination, etc. but still continue to travel.
Although income and prices are the economic variables that are most
commonly used to explain tourism demand, there are other variables:
National income of the country of origin approximated using the
gross domestic product (GDP) per capita.
Prices in the destination approximated using the consumer price
index (CPI) or a ratio between the CPI of the destination and that of
the visitor markets. In addition, the purchasing power parity (PPP) is
used to estimate relative prices.
The price of substitute and/or complementary destinations, defined as
an index that is the weighted average of prices in the other
destinations that compete for the same market.
Origin-destination cost of transport. Rus and León (1997) propose
constructing an index of the cost of transport in tourist destinations
for each mode of transport. Other studies that include this variable
use the price of fuel as a proxy for the cost of transport (Buisán,
1997; Ledesma-Rodríguez et al, 2001).
Origin-destination exchange rate. This variable could have an effect
on tourism demand that is different to that of price.
Promotion expenses. If this variable is available, the modelling
should take into account the delay effect involved in advertising
expenditure.
Fashion and preferences, which are approximated by temporary
trends. The effects of other variables that were omitted in the
specification of the model can also be included here.
Accumulated investment. This uses the series of investment in
infrastructures in general (transport, communications, etc.).
Origin-destination distance. Different kinds of distance can be used:
physical (kilometres); economic (cost of the trip); temporal (length of
the trip) and psychological or cultural. Thus, this variable can be
interpreted, to a certain extent, as a proxy for the cost of the trip.
TOURISMOS: AN INTERNATIONAL MULTIDISCIPLINARY JOURNAL OF TOURISM
Volume 4, Number 1, Spring 2009, pp. 13-28
17
Extraordinary events that have international repercussions. Dummy
variables are normally used to include these events in the
specification of models.
Delayed dependent variable. This variable includes the limitations in
the capacity of the tourism industry to meet increases in demand; the
persistence of consumption habits (Witt and Martin, 1987): and the
tourist capacity of a destination (Buisán, 1995; Garín-Muñoz and
Amaral, 2000; Ledesma-Rodríguez et al., 2001).
This list is not exhaustive or closed. It simply aims to illustrate the
main factors that determine tourism demand, as presented in the literature.
The objective of this study is to analyse which of these factors are
available for the visitor markets included in this analysis, and which
factors are relevant to tourism demand in Catalonia. Thus, a database can
be built using the homogeneous data that is available for the main visitor
countries. This database enables us to analyse the characteristics of the
selected dependent variables and their relation with tourism demand in
Catalonia.
Table 1. Distribution of the frequency of tourists and overnight
stays (expressed in thousands)
Year 2004 Tourists %
%
cumulated
Overnight
stays
%
%
cumulated
France
3566
32.29
32.29
27153
29.15
29.15
United
Kingdom
2223 20.13 52.43 18662 20.03 49.18
Belgium
and NL 1488 13.48 65.90 15388 16.52 65.70
Germany
1393
12.62
78.52
14959
16.06
81.75
Italy
986
8.93
87.45
7210
7.74
89.49
Switzerland
596
5.40
92.85
3158
3.39
92.88
US and
Japan
330
2.99
95.83
2822 3.03 95.91
Russia
326
2.95
98.79
2465
2.65
98.56
Northern
countries
134
1.21
100
1342
1.44
100
Source: Compiled by the author, using data from Turisme de Catalunya and the Statistical
Institute of Catalonia (IDESCAT), as well as Frontur data from the Institute of Tourism
Studies (IET).
An analysis of tourism demand in Catalonia for the year 2004 (first
destinations), both in terms of the number of tourists and the number of
overnight stays, shows that over three quarters of the visitors (87.45% in
Oscar Claveria & Jordi Datzira
18
the case of the number of tourists and 89.49% in the number of overnight
stays in Catalonia) come from France; the United Kingdom; Belgium and
the Netherlands; Germany; and Italy, in this order (see Table 1). Thus,
this paper focuses on macroeconomic information from these five visitor
markets2
.
EMPIRICAL ANALYSIS
In this section we analysed the development of tourism demand, both
in terms of the number of tourists arriving in Catalonia and the number of
overnight stays in the Catalan region. First, a descriptive study of the
different variables that determine tourism demand in Catalonia was
carried out. From the available homogeneous data, 45 exogenous monthly
variables and 29 quarterly variables were included for each one of the five
main visitor markets (see Tables 6 and 8). In addition, an analysis of
correlations was carried out. This enabled us to study the degree of linear
association between different variables and tourism demand in Catalonia
for each visitor country3
This methodology enabled us to detect and analyse the characteristics
of the factors determining tourism demand in Catalonia. Thus, we could
fill the existing gap in studies on tourism demand in this region, and
verify the hypotheses that are often taken as valid with no empirical
evidence.
.
Descriptive statistical analysis of tourism demand in Catalonia
This subsection contains a descriptive statistical analysis of the
tourists arriving in Catalonia and the number of overnight stays,
according to the country of origin (Tables 2 to 5). This analysis is limited
to the main five visitor markets. The results of the descriptive statistical
analysis are presented both for the original monthly series and for the
quarterly series4. The year-on-year rate of the series, adjusted for
seasonality, was used for both the number of tourists and the overnight
stays5
The following summary measures were obtained: the arithmetic
mean (
.
x), the median (
Me
), the first (
1
Q
) and third quartile s (
3
Q
) as
measurements of position, and the minimum (
.Min
), maximum (
.Max
),
the range (
.MinMaxR −=
), the standard deviation (
σ
) and the variation
coefficient (
CV
) as dispersion measures.
TOURISMOS: AN INTERNATIONAL MULTIDISCIPLINARY JOURNAL OF TOURISM
Volume 4, Number 1, Spring 2009, pp. 13-28
19
Table 2. Descriptive analysis of numbers of tourists
(1:1998-7:2005)
x
Me
1
Q
3
Q
.Min
.Max
R
σ
VC
Total 10.11 9.59 5.09 15.54 -12.61 28.05 40.67 8.01 79.18
France 14.86 12.08 -1.92 26.25 -32.55 87.74 120.29 22.94 154.33
UK 15.36 14.17 4.67 23.10 -16.73 76.62 93.35 15.40 100.29
Belgium 9.47 8.02 2.56 15.35 -48.86 87.58 136.44 15.78 166.53
NL 22.29 17.15 -6.99 40.04 -66.99 186.39 253.38 48.43 217.26
Germany 5.49 4.45 -2.44 12.10 -42.06 99.03 141.09 15.61 284.41
Italy 13.46 7.76 -0.58 20.66 -48.53 159.74 208.27 27.14 201.67
Source: Compiled by the author.
Table 3. Descriptive analysis of numbers of tourists
(I:1998-II:2005)
x
Me
1
Q
3
Q
.Min
.Max
R
σ
VC
Total 9.94 9.91 6.89 13.40 -2.58 21.00 23.58 5.87 59.05
France 13.73 12.71 0.39 25.58 -12.97 45.56 58.53 16.23 118.19
UK 15.06 12.35 8.71 19.69 -3.31 43.35 46.66 10.85 72.02
Belgium 8.76 7.81 2.56 16.10 -9.85 29.87 39.72 9.48 108.21
NL 20.57 19.60 -2.04 39.70 -61.66 167.26 228.92 43.81 213.01
Germany 5.15 4.87 0.55 10.94 -23.22 38.34 61.56 12.26 238.19
Italy 12.33 5.97 1.38 21.82 -17.79 71.03 88.82 18.74 152.02
Source: Compiled by the author.
Table 4. Descriptive analysis of overnight stays (1:2002-8:2005)
x
Me
1
Q
3
Q
.Min
.Max
R
σ
VC
Total 10.13 10.12 -0.54 19.23 -15.75 60.50 76.24 14.90 147.11
France 14.11 9.66 -1.35 27.06 -39.52 93.52 133.04 24.08 170.66
UK 18.16 13.33 -0.34 24.76 -51.16 214.26 265.42 37.03 203.88
Belgium 4.63 5.59 -9.16 9.58 -27.54 76.76 104.30 20.60 445.07
NL 10.89 10.30
-
10.74 31.55 -30.17 59.56 89.72 25.05 229.90
Germany 6.13 4.37 -3.31 15.84 -28.98 61.20 90.18 16.99 277.17
Italy 19.60 8.93 -6.65 30.58 -34.08 161.92 196.00 43.51 221.99
Source: Compiled by the author.
Table 5. Descriptive analysis of overnight stays (I:2002-II:2005)
Oscar Claveria & Jordi Datzira
20
x
Me
1
Q
3
Q
.Min
.Max
R
σ
VC
Total 9.71 10.71 -0.65 17.23 -6.37 35.41 41.78 11.72 120.77
France 12.81 11.10 -2.88 23.16 -11.00 54.22 65.22 19.41 151.47
UK 17.66 12.97 3.11 25.22 -27.67 99.83 127.51 29.39 166.44
Belgium 4.23 -0.68 -6.10 9.81 -20.36 48.90 69.26 17.64 416.72
NL 8.57 7.06 -0.72 17.94 -18.39 37.26 55.65 14.58 170.22
Germany 5.97 6.47 -6.98 18.02 -13.96 25.31 39.27 14.40 241.38
Italy 17.03 8.29 -2.38 46.10 -20.70 61.47 82.17 28.09 164.99
Source: Compiled by the author.
The difference between the arithmetic mean and the median in
countries like Italy—for the rate of variation in both the number of
tourists and the number of overnight stays—shows that there was strong
asymmetry in the distribution of variation rates, in this case towards the
lower values of the distribution. These results could be due to outliers6
We found less relative dispersion for the monthly data than for the
quarterly data. However, the high degree of dispersion in the series of
some countries affected the strength of the linear relation between the
exogenous variables (external economic factors) and the endogenous
variables (rates of tourists and overnight stays). Therefore, an intervention
analysis could be carried out on the outliers in the monthly series of
tourists and overnight stays in Catalonia.
.
External macroeconomic factors that determine tourism
demand in Catalonia
The homogeneous information available for the main visitor
countries and the main macroeconomic factors that determine tourism
demand in Catalonia (see Tables 6 and 8) were used to analyse the degree
of linear association between the selected dependent variables and
tourism demand in Catalonia, by means of a correlations analysis. Tables
7 and 9 show the correlation coefficients for the relationship between
each one of the selected exogenous variables and the corresponding
endogenous variable (tourists or overnight stays) for each country.
Significant correlation coefficients at 5% are indicated in bold.
TOURISMOS: AN INTERNATIONAL MULTIDISCIPLINARY JOURNAL OF TOURISM
Volume 4, Number 1, Spring 2009, pp. 13-28
21
Table 6. Coding of variables (monthly data)7
Code
Variable
Type
1
Crude oil domestic first purchase price
$/barrel
2
Harmonized consumer price index (HICP) for all products
T(1,12)
3
HICP – Transport
T(1,12)
4
HICP - Hotels, cafes and restaurants
T(1,12)
5
Harmonised unemployment (thousands of people)
T(1,12)
6
Harmonised unemployment (rate %)
%
7
Industrial Confidence Indicator
Balance
8
Service Confidence Indicator
Balance
9
Consumer Confidence Indicator
Balance
10
Confidence Indicator for the retail trade sector
Balance
11
Construction Confidence Indicator
Balance
12
Economic Sentiment Indicator
Balance
13
Production trends in recent past
(industry)
Balance
14
Order books (industry)
Balance
15
Export order books (industry)
Balance
16 Stocks of finished products
(industry)
Balance
17
Production expectations for the months ahead (industry)
Balance
18
Selling price expectations for the months ahead (industry)
Balance
19
Employment expectations for the months ahead (industry)
Balance
20
Financial situation of household compared with 12 months ago (consumer)
Balance
21
Financial position of household over the next 12 months (consumer)
Balance
22
General economic situation over the last 12 months (consumer))
Balance
23
General economic situation over the next 12 months (consumer))
Balance
24
Cost of living compared with 12 months ago (consumer)
Balance
25
Price expectations for the next 12 months (consumer)
Balance
26
Unemployment expectations for the next 12 months
(consumer)
Balance
27
Major purchases at the present time (consumer)
Balance
28
Major purchases for the next 12 months (consumer)
Balance
29
Savings at the present time (consumer)
Balance
30
Savings for the next 12 months (consumer)
Balance
31
Present financial situation of household (consumer)
Balance
32
Assessment of business climate (services)
Balance
33
Evolution of demand in recent months (services)
Balance
34
Evolution of demand expected in the months ahead (services)
Balance
35
Evolution of employment in recent months (services)
Balance
36
Evolution of employment expected in the months ahead (services)
Balance
37
Present business (sales) position (retail trade)
Balance
38
Present stock (retail trade)
Balance
39
Expected orders placed with suppliers during the next 3 months (retail trade)
Balance
40
Business trend over the next 6 months (retail trade)
Balance
41
Employment expectations over the next 3 months (retail trade)
Balance
42
Development of activity compared with the preceding month (construction)
Balance
43
Evaluation of order books or production schedules (construction)
Balance
44
Employment expectations over the next 3 or 4 months (construction)
Balance
45
Price expectations over the next 3 or 4 months (construction)
Balance
Source: Compiled by the author.
Oscar Claveria & Jordi Datzira
22
Table 7. Analysis of correlations for monthly variables
Code
France
UK
Belgium
NL
Germany
Italy
touri
sts stays touri
sts stays touri
sts stays touri
sts stays touri
sts stays touri
sts stays
1
-0.35
-0.01
-0.01
-0.14
-0.19
-0.04
0.02
0.05
-0.19
-0.41
0.10
-0.04
2 -0.15 -0.13 0.09 -0.21 -0.16 -0.22 0.18 0.21 -0.22 -0.12 -0.30 -0.15
3
-0.34
-0.12
0.17
0.07
-0.06
-0.29
-0.08
0.04
-0.12
-0.24
-0.05
-0.22
4
0.31
0.46
-0.28
-0.31
-0.21
-0.14
0.30
0.32
0.18
0.28
-0.30
-0.08
5
0.08
-0.22
-0.02
-0.41
-0.13
-0.51
-0.16
-0.28
0.02
0.47
-0.02
-0.22
6
-0.15
-0.49
0.07
-0.15
0.12
-0.14
-0.02
-0.12
0.13
-0.30
-0.04
0.00
7
-0.28
-0.22 0.12 0.13 0.07 0.27 0.18 0.28 -0.22 -0.15 -0.20 0.09
8
-0.16
0.06
0.04
0.20
0.21
0.23
0.06
0.26
-0.16
-0.06
0.12
0.21
9
-0.02
0.41
0.10
-0.06
0.00
0.24
0.12
0.43
-0.09
0.33
-0.27
-0.01
10
-0.42
-0.25
0.07
0.35
-0.04
-0.06
0.12
0.31
-0.28
-0.22
-0.16
-0.25
11
-0.04
0.58
0.07
0.36
0.15
0.04
0.18
0.38
0.09
0.11
-0.20
-0.26
12
-0.21
0.05
0.11
0.21
0.12
0.27
0.16
0.44
-0.17
-0.02
-0.14
0.10
13
-0.26
-0.07
0.10
0.17
0.00
0.27
0.23
-0.23
-0.08
-0.04
-0.22
0.01
14
-0.21
0.05
0.01
0.02
0.07
0.12
0.20
0.21
-0.28
-0.34
-0.18
0.08
15
-0.30
-0.21
-0.10
-0.06
0.06
0.21
0.20
0.22
-0.28
-0.31
-0.11
0.09
16
0.42
0.50
-0.09
-0.20
0.02
-0.26
-0.11
-0.31
0.18
0.13
0.09
-0.09
17 -0.24 -0.12 0.21 0.20 0.11 0.40 0.14 0.24 -0.10 0.13 -0.21 0.05
18
-0.25
0.06
0.22
-0.05
-0.07
0.14
0.12
0.10
-0.16
-0.18
-0.14
0.18
19
0.02
0.52
0.05
0.14
0.05
0.26
0.13
0.06
-0.24
-0.25
-0.26
0.03
20
0.23
0.34
-0.24
-0.35
0.03
0.18
0.16
0.30
-0.17
-0.18
-0.20
-0.04
21
0.16
0.46
-0.23
-0.33
-0.06
0.38
0.15
0.33
-0.04
0.09
-0.16
-0.01
22
0.18
0.63
0.03 -0.20 0.05 0.19 0.12 0.37 -0.12 -0.03 -0.15 0.03
23
0.04
0.48
0.12
-0.07
-0.07
0.45
0.12
0.29
-0.06
0.34
-0.14
0.09
24
-0.04
-0.24
0.08
0.31
-0.21
-0.04
-0.27
-0.15
0.06
0.35
-0.11
-0.05
25
-0.02
-0.19
-0.26
-0.03
0.08
0.10
0.22
0.26
-0.14
0.28
-0.21
-0.14
26
0.07
-0.35
-0.01
0.24
-0.01
-0.08
-0.12
-0.42
0.10
-0.34
0.18
-0.16
27
0.04
0.28
-0.30
-0.12
0.01
0.07
0.09
0.40
-0.09
-0.24
-0.05
0.10
28
0.12
-0.02
-0.22
-0.13
-0.11
0.27
0.17
0.28
-0.09
-0.09
0.17
0.03
29
0.18
-0.01
-0.06
-0.02
-0.05
-0.34
0.09
-0.05
0.13
0.31
-0.28
-0.29
30
0.00
0.14
0.12
0.15
0.15
-0.17
0.06
0.05
-0.13
0.24
-0.38
-0.20
31
-0.26
-0.12
0.04
-0.39
-0.27
0.06
0.14
0.10
-0.05
-0.31
0.19
0.07
32 -0.14 0.11 -0.04 0.09 0.24 0.18 0.16 0.30 -0.19 -0.03 0.03 0.23
33
-0.20
-0.14
0.04
0.18
0.25
0.19
-0.22
-0.24
-0.06
-0.13
0.12
0.01
34
-0.09
0.22
0.11
0.27
0.06
0.19
0.18
0.32
-0.18
0.05
0.13
0.10
35
-0.01
0.07
-0.06
0.04
0.23
0.02
0.05
0.26
-0.08
-0.07
0.09
0.27
36
0.04
0.38
0.03
0.06
0.18
0.18
0.10
0.33
-0.32
-0.09
-0.07
0.06
37
-0.02 -0.01 0.06
0.35
0.01 -0.14 0.11 0.25 -0.21 -0.25 -0.20 -0.10
38
0.04
-0.15
0.08
-0.13
0.10
0.28
-0.04
-0.16
0.17
0.24
0.06
0.45
39
-0.11
0.07
0.04
0.39
-0.08
0.10
0.18
0.40
-0.33
-0.10
0.05
-0.02
40
-0.50
-0.37
0.13
0.34
-0.05
0.23
0.16
0.43
-0.32
-0.04
-0.03
-0.14
41
0.04
0.29
-0.24
-0.12
0.09
0.04
0.14
0.40
-0.14
-0.10
-0.17
-0.10
42
-0.11
0.22
0.26
0.38
0.19
0.18
0.20
0.17
0.02
0.05
-0.20
-0.18
43
0.01
0.68
0.08
0.29
0.15
0.03
0.16
0.38
0.11
0.14
-0.23
-0.32
44
-0.12
0.38
0.05
0.37
0.14
0.04
0.18
0.36
0.08
0.07
-0.14
-0.06
45
-0.04
0.28
0.09
0.22
0.11
0.10
0.19
0.43
0.03
0.34
-0.20
0.03
Source: Compiled by the author.
TOURISMOS: AN INTERNATIONAL MULTIDISCIPLINARY JOURNAL OF TOURISM
Volume 4, Number 1, Spring 2009, pp. 13-28
23
Table 8. Coding of variables (quarterly data)8
Code
Variable
Type
1
Gross domestic product at market prices
T(1,4)
2
Final consumption expenditure
T(1,4)
3
Final consumption expenditure: households
T(1,4)
4
Gross fixed capital formation – total
T(1,4)
5
Changes in inventories
6
Net national income
T(1,4)
7
Net disposable income
T(1,4)
8
Net saving
T(1,4)
9
Final consumption expenditure
T(1,4)
10
Gross capital formation
T(1,4)
11
Gross domestic product at market prices
T(1,4)
12
Final consumption expenditure
T(1,4)
13 Final consumption expenditure – household and NPISH (Net profit
institutions serving households) T(1,4)
14
Gross capital formation
T(1,4)
15
Exports of goods and services
T(1,4)
16
Imports of goods and services
T(1,4)
17
Labour cost index - Total labour cost (Industry and services excluding
public administration) Index
18
Labour cost index - Wages and salaries (Industry and services
excluding public administration) Index
19
Production capacity (industry)
Balance
20
Duration of assured production in months (industry)
Months
21
New orders in recent past (industry)
Balance
22
Export expectations for the months ahead (industry)
Balance
23
Capacity utilization (industry)
%
24
Competitive position in recent past on the domestic market (industry)
Balance
25
Competitive position in recent past on the foreign market inside the
EU (industry) Balance
26
Competitive position in recent past outside the EU (industry)
Balance
27
Likelihood of buying a car within the next two years (consumer)
Balance
28
Plans to purchase or build a home within the next two years
(consumer) Balance
29
Likelihood of spending any large sum of money on home
improvements (consumer) Balance
Source: Compiled by the author.
Oscar Claveria & Jordi Datzira
24
Table 9. Analysis of correlations for quarterly variables
Code
France
UK
Belgium
NL
Germany
Italy
tourists stays tourists stays tourists stays tourists stays tourists stays tourists
1
-0.18
0.03
0.09
0.78
0.39
0.17
0.17
0.16
-0.07
0.21
-0.07
0.34
2 0.10 0.39 0.05 0.46 0.29 0.10 -
0.09 0.38 -0.07 -0.29 -0.31 -0.06
3
0.04
0.46
0.06
0.27
0.23
0.01
0.00
0.54
-0.12
-0.56
-0.21
-0.05
4
-0.42
-0.66
-0.37
-0.17
0.17
-0.19
0.23
0.05
-0.08
-0.38
0.04
0.17
5
0.05
-0.29
-0.03
0.24
0.52
0.29
-
-
0.05
-0.02
0.05
-0.23
6
-0.33
-0.31
-0.25
-0.05
0.25
0.12
-
-
0.38
0.24
0.03
0.48
7
-0.30
-0.26
-0.24
-0.04
0.24
0.06
-
-
0.35
0.21
0.10
0.52
8
-0.24
-0.42
-0.46
-0.26
0.47
0.14
-
-
0.31
0.32
0.49
0.25
9
-0.08
0.42
-0.14
0.02
-0.05
0.05
0.01
0.39
-0.07
-0.40
-0.33
0.22
10
-0.52
-0.25
-0.15
0.14
0.36
-0.17
0.16
-0.15
-0.09
-0.27
0.02
-0.20
11
0.09
0.61
0.09
-0.40
-0.50
0.20
0.00
0.28
0.18
0.28
-0.02
0.38
12
-0.24
0.46
0.01
-0.40
-0.46
-0.06
0.05
0.36
-0.23
-0.20
-0.22
0.40
13
-0.31
-0.11
0.08
0.13
-0.31
0.03
0.06
0.19
-0.39
-0.02
-0.38
0.08
14
-0.14
0.23
0.32
0.04
0.23
-0.19
0.13
0.60
-0.45
-0.46
-0.10
-0.13
15
-0.49
-0.11
-0.11
-0.54
-0.05
-0.20
0.04
-0.18
-0.27
-0.41
-0.01
0.00
16
-0.52
-0.23
-0.09
-0.31
-0.01
-0.26
0.09
-0.11
-0.36
-0.54
-0.12
-0.07
17
-0.19
-0.23
0.13
0.75
-0.29
-0.79
0.09
-0.19
-0.08
-0.16
0.02
0.16
18
-0.17
-0.23
-
-
-0.35
-0.68
0.09
-0.18
-0.06
-0.16
0.01
0.16
19
0.21
-0.22
0.07
0.12
-0.15
-0.11
-
0.16
-0.25
0.19
0.17
0.12
-0.32
20
0.09
0.07
0.15
-0.01
0.05
-0.11
0.23
0.03
-0.48
0.48
-0.31
-0.18
21
-0.36
-0.05
-0.07
0.36
0.20
0.39
0.30
0.27
0.02
0.09
0.10
0.13
22
-0.49
-0.35
0.27
0.10
0.24
0.44
0.24
0.50
-0.02
-0.10
-0.10
-0.05
23
0.08
0.43
-0.30
-0.34
0.10
0.04
0.24
0.70
-0.15
0.01
-0.32
0.06
24
-0.01
0.47
0.05
0.10
-0.03
0.06
0.45
0.45
-0.31
-0.29
-0.15
0.41
25
-0.04
0.01
0.12
0.19
-0.03
0.06
0.28
0.38
-0.19
0.04
-0.43
-0.26
26
-0.02
0.46
-0.37
-0.11
-0.15
-0.35
0.13
0.42
-0.18
0.01
-0.44
-0.29
27
0.28
0.40
-0.17
-0.28
-0.08
-0.38
0.17
0.75
0.13
0.27
-0.31
0.00
28
-0.49
-0.43
-0.17
0.05
0.24
-0.36
0.11
0.55
-0.16
0.22
-0.14
0.14
29
-0.27
0.08
0.06
0.26
-0.28
-0.27
0.21
0.47
0.00
-0.16
-0.31
-0.08
Source: Compiled by the author.
As expected, the sign of the correlations between the overnight stays
and the corresponding exogenous variables usually coincides with that of
the correlations obtained using the number of tourists. When the signs do
not coincide there usually is a weak linear association. Tables 7 and 9
show how the correlations between overnight stays and the corresponding
exogenous variables are higher than those obtained using the variation
rate for the number of tourists. In addition, the correlations obtained using
quarterly data are higher overall than those obtained with the monthly
data. These results indicate that quarterly data on overnight stays should
be used as a proxy of tourism demand when specifying and estimating a
model of aggregate tourism demand for Catalonia.
TOURISMOS: AN INTERNATIONAL MULTIDISCIPLINARY JOURNAL OF TOURISM
Volume 4, Number 1, Spring 2009, pp. 13-28
25
CONCLUSIONS
The aim of this study was to detect the external macroeconomic
factors that determine tourism demand in Catalonia and analyse their
relationship, given that there is a lack of applied studies on Catalonia in
the literature. The study focuses on first destinations and uses the number
of tourists and overnight stays as proxies of tourism demand in Catalonia.
However, from a conceptual perspective and from the results of the
empirical analysis, the variable of overnight stays was shown to be more
appropriate for approximating tourism demand in Catalonia.
First we collected the available homogenous statistical information
for the main five visitor markets in order to carry out an analysis of
correlations between the group of factors that explain tourism demand in
Catalonia and the year-on-year variation rates in the seasonally adjusted
series of tourists and overnight stays. We found that final consumption
expenditure and expected activity in the retail sector have a notable effect
on aggregate demand. However, we can state that tourism demand in
Catalonia varies according to the visitor market. Therefore, it is important
to differentiate between an aggregate model of demand for the total
number of tourists, and specific models for each one of the countries
analysed.
The overnight stays of foreign tourists in Catalonia mainly had a
higher degree of linear association with the macroeconomic factors.
Therefore, we confirmed that the variable overnight stays gives a better
approximation of tourism demand in Catalonia than the number of
tourists arriving from the visitor markets.
To model tourism demand, we also suggest including dichotomous
variables or dummies, as they enable us to incorporate events that have
historical series with an insufficient number of observations. This is the
case of series related to the influence of low cost airlines on tourism
demand in Catalonia; the effect of major events; trade fairs, etc.
In accordance with Buisán (1997), we consider that variations in
demand should be studied by means of a function that includes all of the
factors that have an effect on demand. Such factors are not only
economic. Although data are not always available for many of these
factors, particularly at regional level, deeper knowledge of tourism
demand could help to detect and act effectively on the different
requirements for improving existing offerings and adapting them to the
demands of current and potential clients who are increasing well-
informed and demanding.
Oscar Claveria & Jordi Datzira
26
REFERENCES
Anderson, O. (1952). The business test of the IFO-Institute for economic research,
Munich, and its theoretical model, Revue de l’Institut International de
Statistique, Vol. 20, pp.1-17.
Buisán, A. (1997). Exportaciones de turismo y competitividad, Revista de
Economía Aplicada, Vol. 5, pp.65-81.
Bull, A. (1994). La economía del sector turístico. Madrid, Alianza Editorial.
Claveria, O., Pons, E. and Suriñach, J. (2006). Quantification of expectations. Are
they useful for forecasting inflation? Economic Issues, Vol. 11, pp.19-38.
Claveria, O., Pons, E. & Ramos, R. (2007). Business and consumer expectations
and macroeconomic forecasts. International Journal of Forecasting, Vol.
23, pp.47-69.
Crouch, G.I. (1994b). The Study of International Tourism Demand: A Review of
Findings. Journal of Travel Research, Vol. 33, pp.12-23.
De Rus, G. and León, C. (1997). Economía del turismo. Un panorama. Revista de
Economía Aplicada, Vol. 5, pp.71-109.
Figuerola, M. (1992). Teoría económica del turismo. Madrid, Alianza Editorial.
Figuerola, M. (1996). Experiencias cuantitativas en el análisis económico del
turismo. Estudios Turísticos, Vol. 129, pp.37-58.
Garín-Muñoz, T. & Amaral, T.P. (2000). An econometric model for international
tourism flows to Spain. Applied Economics Letters, Vol. 7, pp.525-29.
Gerakis, A.S. (1965). Effects of exchange-rate devaluations and revaluations on
receipts from tourism. International Monetary Fund Staff Papers, Vol. 12,
pp.365-84.
González P. and Moral, P. (1995). An analysis of the international tourism
demand in Spain. International Journal of Forecasting, Vol. 11, pp.233-
251.
Ledesma-Rodríguez, F.J., Navarro-Ibáñez, M. & Pérez-Rodríguez, J.V. (2001).
Panel data and tourism: a case study of Tenerife. Tourism Economics, Vol.
7, pp.75-88.
Lim, C. (1997). Review of international tourism demand models. Annals of
Tourism Research, Vol. 24, pp.835-849.
Morley, C.L. (1994). Experimental destination choice analysis. Annals of Tourism
Research, Vol. 21, pp.315-328.
O'Hagan, J.W. & Harrison, M.J. (1984). Market shares of US tourist expenditure
in Europe: an econometric analysis. Applied Economics, Vol. 16, pp.919-
931.
Padilla, R. (1988). La demanda de servicios turísticos en España. Investigaciones
Económicas, Vol. 12, pp.133-157.
Smith, S. (1995). Tourism Analysis. London, Longman.
Uysal M. & Crompton J.L. (1985). An overview of approaches used to forecast
tourism demand. Journal of Travel Research, Vol. 23, pp.7-15.
Witt. S.F. & Martin, C.A. (1987). Econometric models for forecasting
international tourism demand. Journal of Travel Research, Vol. 25, pp.23-
30.
TOURISMOS: AN INTERNATIONAL MULTIDISCIPLINARY JOURNAL OF TOURISM
Volume 4, Number 1, Spring 2009, pp. 13-28
27
Witt, S.F. & Witt, C.A. (1995). Forecasting tourism demand: A review of
empirical research. International Journal of Forecasting, Vol. 11, pp.447-
475.
ENDNOTES
1. Source: Institute of Tourism Studies (IET).
2. Belgium and the Netherlands are treated as one visitor market.
3. We estimated a regression model for each of the countries included in the
study, obtaining very poor results except for Germany. Models were
specified for quarterly overnight stays using the significant correlated
economic factors as dependent variables.
4. Monthly data was converted to quarterly data, as many of the economic
variables are only available for quarterly periods.
5. The strong seasonal component observed in the graphic analysis indicated
that it was essential to seasonally adjust the series. Overall, the series of
tourists and overnight stays by country had different evolutions for France,
the United Kingdom and the other visitor countries. The evolution of the
series for countries such as France, Germany and Belgium was similar to that
of the total number of tourists and overnight stays in Catalonia. In contrast,
the evolution of series for countries such as the United Kingdom, Ireland and
the Netherlands was different to the total, and showed a high degree of
variability. Despite the different evolutions observed between the countries,
there was a strong seasonal component in the series for all of the visitor
markets. This marked behaviour led us to work with seasonally adjusted
series. The series were deseasonalised by means of the Tramo-Seats
programme. Adjusted series could then be used to obtain the corresponding
year-on-year variation rates.
6. Outliers were observed for some countries Particularly in Germany, Ireland,
Italy, the Netherlands and the Nordic Countries in terms of the number of
tourists arriving in Catalonia; and Ireland and the United Kingdom, in terms
of the overnight stays in Catalonia and notably increased the variability of
the series.
7. All of the variables, except 1, 2 3 and 4, were seasonally adjusted from their
source of origin. The variables 1 to 4 were not deseasonalised, either because
they had no seasonal components or because seasonality was corrected when
the year-on-year variation rate was applied. The variables 2 to 4 are indices
(1996=100). The variable balance was taken as the percentage of surveyed
individuals who expected an increase in the variable minus the percentage of
surveyed individuals who expected a decrease. As shown by Anderson
(1952), the balance is comparable to a variation rate. See Claveria et al.
(2006) for a description of a conversion method of balances into a
quantitative measure of agents’ expectations and Claveria et al. (2007).for
their usefulness with forecasting purposes. The quantitative information
comes from the Department of Energy of the US and the European Central
Oscar Claveria & Jordi Datzira
28
Bank and the balances from the harmonised opinion polls published by the
European Commission, including the following: industrial survey; consumer
survey; services survey; retail trade survey and construction survey.
8. All of the variables except 17 and 18 have been seasonally adjusted. The
variables 1 to 1 were measured in millions of € from 1.1.1999 (ECUs until
31.12.1998). The variables 11 to 16 are indices (1995=100), based on
national currency. The variables 17 and 18 are indices (2000=100). The
variables 1 to 5 were measured in constant prices of 1995. The quantitative
information comes from the European Central Bank and the balances from
the harmonised opinion polls published by the European Commission,
including the industrial survey and the consumer survey.
SUBMITTED: FEBRUARY 2008
REVISION SUBMITTED: APRIL 2008
ACCEPTED: MAY 2008
REFEREED ANONYMOUSLY
Oscar Claveria (oclaveria@ub.edu) is a research fellow at the Research
Institute of Applied Economics (IREA) and assistant professor in the
Department of Econometrics at the University of Barcelona. Diagonal,
690. E-08034 Barcelona (Spain).
Jordi Datzira (jordi.datzira@ddservices.eu) is a external lecturer at
several universities in Europe and a senior consultant and director of
Datzira Development Services (DDS). Princesa, 31, 3er. E-08003
Barcelona (Spain).