Tourism Management 59 (2017) 67e75

Contents lists available at ScienceDirect

Tourism Management
journal homepage: www.elsevier.com/locate/tourman

Use value of cultural events: The case of the Christmas markets

Juan Gabriel Brida a, *, Marta Meleddu b, Oksana Tokarchuk c
a

micas y de Administracio

n, Universidad de la Repblica, Eduardo Acevedo 1135,
Departamento de M

etodos Cuantitativos, Facultad de Ciencias Econo
11200, Montevideo, Uruguay
b
Department of Political Science, Communication, Engineering and Information Technologies (POLCOMING) and CRENoS, University of Sassari, Piazza
 11, 07100, Sassari, Italy
Universita
c
Competence Centre in Tourism Management and Tourism Economics (TOMTE), School of Economics and Management, Free University of Bolzano, Piazza

1, Ie39100, Bolzano, Italy
Universita

h i g h l i g h t s

A Christmas market is a cultural and retail event that creates a recreational value.

Visitors mainly attend to relax, to enjoy the Christmas atmosphere, and to spend time.

The consumer surplus of Christmas markets conveys benets for the society as a whole.

The recreational value is higher for one-day visitors.

Studies in different countries and Christmas markets are needed to compare the value for visitors.

a r t i c l e i n f o

a b s t r a c t

Article history:
Received 26 November 2015
Received in revised form
5 July 2016
Accepted 20 July 2016

The Christmas market in Meran (Italy) is one of the top 27 European Christmas markets, attracting
around 2000 visitors a day for 50 days. The Christmas market is a cultural and retail event that creates a
recreational value for visitors. This paper, using the travel cost method, estimates the recreational use
value of the Meran Christmas market for one-day and multiple-day visitors. The aim is to estimate the
economic or social value (consumer surplus) of the Christmas market to understand in what way the
event creates a benet for society as a whole. The empirical data were collected via a survey administered
at the end of a visit at the Christmas market during the 2011e2012 season. The empirical ndings
highlight that the use value differsdother than according to travel costsdaccording to the type of visit
experience and to the perceived authenticity.
2016 Elsevier Ltd. All rights reserved.

JEL:
L83
D12
C31
Keywords:
Travel cost
Economic value
Cultural event
Consumer surplus

1. Introduction
In recent years, a visit to a Christmas market has become a
Christmas tradition for many Europeans (CRR Research, 2014). In
2014, there were 157 main Christmas markets and 2634 smaller
markets in Europe, which attracted 493.7 million visitors who
spent 4450.4 billion (CRR Research, 2014). A Christmas market is a

* Corresponding author.
E-mail addresses: gbrida@ccee.edu.uy, elbrida@gmail.com, JuanGabriel.Brida@
unibz.it (J.G. Brida), mmeleddu@uniss.it (M. Meleddu), oksana.tokarchuki@unibz.
it (O. Tokarchuk).
http://dx.doi.org/10.1016/j.tourman.2016.07.012
0261-5177/ 2016 Elsevier Ltd. All rights reserved.

street market with market stalls selling Christmas decorations,

local traditional food and drinks, artisan products, toys, and other
local traditional products related to Christmas. A Christmas market
is not only a shopping occasion but also provides an experience for
the whole family with the site especially decorated for Christmas,
music, performances, and other features aiming to create a
Christmas atmosphere.
Not surprisingly, most of the research related to Christmas
markets has investigated visitors expenditures (e.g., Brida, Disegna,
& Osti, 2013a, 2013b; Brida & Tokarchuk, 2015). The main goal of
these studies is to understand what makes visitors spend and to
increase their spending to increase economic value of the event to

68

J.G. Brida et al. / Tourism Management 59 (2017) 67e75

the hosting community. However, Christmas markets are an event

that creates value for the visitor as well. Research has shown that
tourists attend Christmas markets in order to relax, enjoy the
Christmas atmosphere, spend time with family and friends, and
sample local products (Brida, Disegna, & Osti, 2012). Shopping is
generally a secondary motivation for tourists attending Christmas
markets. The growing importance of Christmas markets in the
tourism industry has created the need to estimate the value for
visitors of the experience at a Christmas market that goes beyond a
mere shopping occasion.
Describing the value of event experiences remains a challenge
for event research. Recently, use value has been successfully applied
to event valuation, drawing on extensive research and a validated
methodology from culture economics. Accordingly, the economic
impact of cultural tourism can be analyzed from the two main
perspectives, which include macroeconomic and microeconomic
levels (Brida, Meleddu, & Pulina, 2013c, chap. 2.14). On the microeconomic level, the economic value of cultural tourism considers
the benets for society. On the macroeconomic level, the value is
reected in the stimulation of other economic branches through
direct, indirect, and induced effects.
The total economic value incorporates use value and non-use
value; non-use value is composed of the option value, existence
value, and bequest value. The use value is related to the mode of use
(or to an equivalent value in cash) and not to any intrinsic value of a
certain good or service, and accounts for the satisfaction that one
obtains by consuming the good or service.
The methodologies for assessing these values are broadly classied into two categories: stated preference methodologies (e.g.,
contingent valuation and contingent choice modeling methodologies) and revealed preference methodologies (e.g., hedonic and
travel cost methodologies). The main difference between the two
categories lies in the way consumers reveal their preferences. The
stated preference technique relies on asking people what they
would hypothetically be willing to pay or be willing to accept for a
given good or a commodity bundle. Respondents are asked to
choose the best alternative from among a set of hypothetical
scenarios, which are described by a set of attributes generated from
an experimental design. However, as pointed out by Diamond and
Hausman (1994), these surveys potentially suffer from drawbacks
such as strategic behavior, protest answers, response bias, and respondents ignoring income constraints. In contrast, revealed preference methods use observations about actual choices made by
people to measure preferences, and the primary advantage of these
methods is reliance on actual choices, which avoids the potential
problems associated with hypothetical responses such as strategic
responses or failure to properly consider behavioral constraints.
This is why in this work a revealed preference method is applied.
The paper aims at estimating the economic use value, primarily
recreational, of a particular outdoor recreation activity, the Christmas market in Meran, Italy. Valuations based on revealed preferences are derived from the prices paid for goods or services. To infer
the value derived from the direct use or interaction with cultural
services for recreation purposes, we apply a travel cost model
(TCM), which has been used to value non-market goods and outdoor events (Armbrecht, 2014; Bedate, Herrero, & Sanz, 2004; Poor
& Smith, 2004). A misspecication arises if individuals trips consist
of different trip purposes. In case of ignoring this misspecication
bias may occur. This is why the study includes visitors whose aim is
to visit the Meran Christmas market and focuses on estimating the
economic value of the event in terms of consumer surplus. The
novelty of the study is the attempt to estimate the economic or
social value (consumer surplus) of the Christmas market as distinct
from the economic impact that is a change in expenditure that
could be assessed via other methodologies (e.g., inputeoutput

analysis).
In Section 2, we discuss previous studies on Christmas markets,
and in Section 3, we provide an overview of the travel cost method.
In Section 4, we describe the Meran Christmas market and the
survey methodology and present key socio-demographic characteristics of respondents. In Section 5, we provide additional details
about the modeling process and focus on the important role that
authenticity plays, as highlighted in the literature, in the decision to
attend the Christmas market. In Section 6, we present the results
and the computation of the consumer surpluses. Finally, in Section
7, we conclude.
2. Literature review
Academic research on Christmas markets has covered several
aspects related to this special event. One of the main themes of
research related to Christmas markets is the study of the perceived

ran and Roederer (2013) found that
authenticity of the event. Caste
visitors who perceive a Christmas market as authentic are more
likely to repeat their visit to the event, thus relating the perceived
authenticity of a Christmas market to its nancial value. Brida et al.
(2013a) investigate the relationship between the authenticity of the
event and products for sale at Christmas markets and visitors'
expenditure. They demonstrate that tourists who perceive the
event and products as more authentic have higher willingness to
spend at a Christmas market. Brida et al. (2013b) further investigated visitors' expenditure at Christmas markets. In this study, the
researchers provide a list of factors that inuence visitors' propensity to spend and spending at a Christmas market. Brida and
Tokarchuk (2015) investigate spending intentions and the actual
expenditure of Christmas market visitors. This study demonstrates
that most visitors are ready to spend more than what they actually
spent at Christmas market, which suggests the management of
Christmas markets should pay closer attention not only to satisfaction but also to visitors expectations.
Brida et al. (2012) studied the categorization of visitors to
Christmas markets based on their motivations leading to the
identication of six groups of visitors. For all groups, sampling local
products was the most important factor followed by enjoying the
Christmas atmosphere, relaxing, and visiting the city. Shopping was
found to be relatively important for two of these groups while for
one group shopping had little importance. This study shows that
although academic research is concentrated on commercial aspects
of Christmas market, visitors value other things created by the
event, which are the opportunity to sample local products, enjoy
the unique Christmas atmosphere, relax, and visit the city. To the
best of our knowledge, no study has investigated consumer surplus
generated at a Christmas market. Prayaga, Rolfe, and Siden (2006)
measure the economic value for consumers of a four-day special
event Gemfest in Australia. Comparing data from 1998 to 2000,
they nd that the value of the event decreased over the period.
Investigating the reason for this decrease, they found that in 2000
the event attracted fewer participants and visitors came in larger
groups that led to the distribution of travel costs among a larger
group of people. This study showed the managerial importance of
measuring consumer surplus for events.

rez de Vivero, and Agardy (2015),
As pointed out by Smith, Sua
measuring social value is difcult because there is no consensus on
its denition yet. We can rely only on the Marshallian concept of
consumer surplus rehabilitated by Hicks and made measurable
during the 1970s. It was the result of the development of what is
now known as duality theory, including the demonstration by
Hurwicz and Uzawa (1971), chap. 6 of a theoretically rigorous yet
practical numerical procedure. This approach identies the specic
utility function underlying any given system of demand equations

J.G. Brida et al. / Tourism Management 59 (2017) 67e75

that satises the formal requirements of modern ordinal utility

theory. This is now possible with an econometric estimate of a
suitably specied demand equation for a marketed commodity or a
system of demand equations for a set of commodities, and derive a
theoretically consistent and rigorous estimate of the DupuitMarshall measure of the economic value.
Originally developed by Hotelling (1947) in order to measure the
economic value of national parks, today, the travel cost method is
widely applied to measure the economic value of cultural goods.
Most cultural heritage sites receive government subsidies or
nancial support from other organizations. Due to the scarcity of
resources, the question of which site should receive support and in
which entity is very important for decision makers. Measuring
consumer surplus can provide an answer to this question and help
estimate the amount consumers are willing to allocate to a
particular site.
Tourism, cultural events, and the heritage literature account for
numerous empirical applications of the travel cost method. Various
cultural products and events have been measured, including museums (Armbrecht, 2014; Bedate et al., 2004), theatre (Willis,
Snowball, Wymer, & Grisola, 2012), a concert hall (Armbrecht,
2014), a special exhibition (Vicente & De Frutos, 2011), a historic
city (Poor & Smith, 2004), a historic village and a cathedral (Bedate
et al., 2004), and an ancient Greek temple (Tourkolias, Skiada,
Mirasgedis, & Diakoulaki, 2015). This strand of research demonstrated the validity of the travel cost method applied to the eld of
cultural events and heritage sites. Comparison of the consumer
welfare values generated by different sites helps to create relative
ranking that is helpful for the study of tourists individual preferences and social decisions (Bedate et al., 2004). Measuring consumer surplus is helpful for decisions about allocating subsidies
(Willis et al., 2012). Measures of consumer surplus vary signicantly based on the functional form of the TCM.
Christmas markets present a combined cultural and retail event.
The goal of the event is to create tourist demand in the off-season

ran & Roederer, 2013) and to provide
for some destinations (Caste
an attraction during the evening hours at winter destinations
(Brida et al., 2013a, 2012). Although the retail features of the event
have been studied extensively in the literature, the cultural benets
the event bestows on tourists have not been evaluated. The present
paper aims at lling this gap by measuring consumer surplus with
the help of the TCM approach.

69

visited the Christmas market in Meran. We also control for whether

visitors have visited a Christmas market. The latter determinant
helps identify to what extent growing familiarity with the event
increases tourist retention. Thus, the objective of the quantitative
analysis is to estimate the factors that inuence visits per capita and
how the factors are related to travel costs and other important
variables. In this simple model, the analysis includes demographic
variables, such as age, income, gender, and education levels. The
model assumes that individuals respond to the travel cost in the
same way as to a change in access cost: The time spent at the
recreational site is exogenous and xed; the wage rate represents
the opportunity cost of time. Therefore, the estimate of the consumer surplus is based on a function that expresses the number of
times at the event in years as a function of travel costs. Generally, it
corresponds to the difference between the willingness to pay for a
good or service or to an event and the actual price paid (Hicks,
1946), but according to Andersson, Armbrecht, and Lundberg
(2008), it can be employed as a satisfaction indicator. This occurs
when the individual's value of attending an event is greater than
the actual costs. Graphically, the consumer surplus corresponds to
the area under the demand curve but above the average travel costs
(Fig. 1).
Computationally, the consumer surplus is calculated as the
negative reciprocal of the rate of change in the number of times at
the event in years regarding travel cost (Haab & McConnell, 2002).

4. The case study and the empirical data

4.1. Meran Christmas market
The Christmas market tradition goes back to German-speaking
countries in the Middle Ages. Today, Christmas markets are held
all over Europe. In 2014, Germany dominated the industry with
2,234 Christmas markets, followed by France with 273 Christmas
markets (CRR Research, 2014). Italy is also an important player in
this industry.
The Christmas market in Meran is one of the nine main
Christmas markets in Italy. The rst Christmas market in Meran was
held in 1992 and was one of the rst Christmas markets to be
established in Italy. Meran is a small tourist town in northeast Italy
with around 39,000 inhabitants. The town is located close to the

3. The travel cost method

The TCM is traditionally employed to estimate economic use
values associated with environmental goods and services that are
used for recreation. This method has three approaches: the zonal
travel cost approach (which uses mostly secondary data, with some
simple data collected from visitors), the individual travel cost (ITC)
approach (which exploits a more detailed survey of visitors), and
the random utility approach (which requires information on all
possible alternatives that a visitor might consider to visit). As the
TCM is based on observed behavior, this approach is used to estimate use values only. The underlying assumption of the TCM is that
the time and travel cost expenses that people incur visiting a site
represent the price of access to the site or, as in the present case,
to an event. Thus, people's willingness to pay to visit a site can be
estimated based on the number of trips that the individuals make
at different travel costs. This is analogous to estimating people's
willingness to pay for a marketed good based on the quantity
demanded at different prices. The number of trips complements
the site or event quality.
In the present study, the ITC is assessed in which one takes into
account the number of times or years individuals have previously

Fig. 1. Consumer surplus and travel costs (from Sohngen, Lichtkoppler, & Bielen, 1999).

70

J.G. Brida et al. / Tourism Management 59 (2017) 67e75

border with Austria, and the town's historical roots are closely
related to German-speaking culture. Until the end of World War I,
Meran was part of the Austro-Hungarian Empire. After the war,
Meran became part of the Kingdom of Italy.
The town of Meran is generally known for its spa resorts, and the
main tourist season is the summer. During the winter, Meran is a
tourist attraction for skiers visiting surrounding hills. The Christmas market in Meran is organized to attract tourists to the town
during its low season. The Christmas market takes place during
Advent and is extended over the duration of the Christmas holidays, from 25 November to 6 January. The Christmas market is
composed of stalls featuring local artisans wood, felt, ceramics,
glass, and traditional clothes products. Local food and beverages
dominate, for consumption on-site and for purchase as gifts and
souvenirs. The stands are located along the main shopping street in
Meran allowing visitors to experience the whole town. The town of
Meran is adorned with Christmas decorations everywhere with
Christmas trees, string lights, torchlights, and candles. To complete
the Christmas atmosphere, there are background Christmas music
and live musical and drama performances.
Due to the German-speaking history of the town's roots and
more than 20 year tradition, the Meran Christmas market is one of
the most popular Christmas markets in Italy, and the town's hotels
are full during its low tourist season.
4.2. Empirical data
Data for the present study were collected during the four weeks
of Advent (26 November to 24 December) in 2011. A total of 886
visitors to the Meran Christmas market were interviewed. Data
were collected during all weeks, weekdays, and weekends equally.
A trained representative of the research team conducted interviews
based on a questionnaire. Respondents were selected according to
an accidental sampling technique. Due to the lack of prior information about the visitors proles, it was not possible to implement
a probabilistic sampling technique. Only one member per travel
party was interviewed.
The questionnaire was divided into three parts to collect information about the visit to the market, the visit to Meran, and the
visitor's socio- and demographic information. To apply the travel
cost method to capture the consumer surplus, respondents were
asked to provide an estimate of the travel costs per travel party that
they incurred to get to the Christmas market. This estimate was
transformed into individual travel costs based on the information
about the travel party.
The assumptions of the travel cost method require that the
Christmas market was the visitor's only travel goal. To accommodate this assumption, we excluded from the analysis respondents
whose main motivation for their journey were different from
visiting the Christmas market in Meran. Some observations were
dropped during the analysis due to missing data. The overall
sample for the present study was composed of 436 visitors to the
Christmas market in Meran.
In what follows we report socio-demographic characteristics of
visitors included in the sample. Fifty-six percent of the respondents
were female. The average age of the interviewed visitors was 37
years old; the youngest respondent was 19 years old and the oldest
65 years old. Ninety-nine percent of the sample visited the
Christmas market in a group, 34% visited with children, and 13%
came as part of an organized group while the rest of the sample
attended the Christmas market with relatives or friends. On
average, a travel party consisted of four people. Eighty-nine percent
of the sample were employed, 64% of the sample declared an
annual family household income of less than V50,000, and 29% of
interviewed visitors had a university degree or higher.

On average, visitors to the Christmas market traveled 335 km,

and the travel distance ranged from 31 to 1341 km. Eighty percent
of the sample came from northeast and northwest regions of Italy,
the closest to the Trentino-Alto Adige region where Meran is
located. Ninety-eight percent of the interviewed visitors had visited
at least one Christmas market in their life; however, for 70% the
present visit to the Meran Christmas market was their rst visit.
Loyal visitors to the Meran Christmas market had visited twice, on
average, with the number of visits ranging from two to nine.
For 63% of the sample (276 respondents), the Christmas market
visit was a one-day trip. The subsamples of one-day and multipleday visitors present similar characteristics in terms of demographic
and economic variables. The origins of the visitors differed. The
majority of one-day visitors (54%) come from the northeast region
of Italy, while the multiple-day visitors came mainly from the
northwest region. The former region is closer to Meran. On average,
one-day visitors traveled 306 km to reach the Christmas market,
while multiple-day visitors traveled 386 km, on average. Twenty
percent of the one-day visitors came to the Christmas market as
part of an organized group while less than 2% of the multiple-day
visitors came with an organized group. In contrast, a slightly
higher percentage of multiple-day visitors came with children, 43%
compared to 30% of the one-day visitors.

5. Empirical application of the travel cost method

5.1. The econometric specication
This hypothetical framework was made operational by running
a regression analysis to infer the demand function for the average
visitor to the site where the area below this demand curve gives the
average consumer surplus. As the dependent variable can come
only in the form of positive gures (i.e., count variable), the most
commonly used estimation technique is Poisson regression
(Cameron & Trivedi, 2013).
The standard procedure for the count model usually begins by
running a standard Poisson model, where the distribution is given
by the following expression:

ProbYi yi jwi

el l
yi !

yi

yi 0; 1; 2; ::

Eyi jxi Varyi jxi

l exi b :
(1)
The parameter l represents the average and the variance, as
assumed by the Poisson distribution, and is greater than zero. The
Poisson model is non-linear; however, it can be easily estimated by
the maximum likelihood technique. Several extensions of the
Poisson model may be employed according to the characteristics of
the empirical data and of the dispersion feature (differences between the mean and the variance). An alternative to Poisson
regression is a negative binomial regression, which is generally
used when the values of the mean and variance of the dependent
variable (in this case, the number of visits to the Christmas market
in years) diverge considerably (Haab & McConnell, 2002).
In Christmas market studies, the survey is run at the site;
therefore, the count data are truncated, as the data are observed
only over part of the range of the response variable. Therefore, zero
counts are not observed, and the data are zero-truncated, or more
generally left-truncated. Right-truncation results from the loss of
observations greater than a specic value.
A zero-truncated Poisson model is specied with the following
equation:

J.G. Brida et al. / Tourism Management 59 (2017) 67e75

el l i
1

yi 0; 1; 2; ::::::
$
yi !
1  el
y

ProbYi yi jwi > 0

(2)

where wi denotes the controls other than travel costs. The dependent variable assumes values that range from one (i.e., rst time at
the cultural event) to N. Thus, visit is zero-truncated, and a zerotruncated Poisson (or negative binomial) regression allows one to
model visits with this specic restriction.
The dependent variable was constructed as the number of years
the respondents visited the Meran Christmas market in the past.
The hypothesis postulates that the number of visits of individual i,
Vi, when controlling for socioeconomic characteristics, is a function
of the round-trip costs (TCi). Therefore, we started with the basic
specication that includes, on the right-hand side, other than travel
cost, a set of socio-economics controls, such as age, income (in the
form of dummy variables that take value 1 if: Income1: up to
V20,000; Income2: from V20,001 to V40,000; Income3: from
V40,001 to V70,000), gender, and education level (in the form of a
dummy variable that takes the value of one for those who achieved
an average level of education):

Vi b1 Agei b2 Age2i b3 Malei b4 Marriedi

b5 Upper Secondaryi b6 Income1i b7 Income2i
b8 Income3i b9 Employedi b10 Travel Costi :

(3)

According to the TCM hypothesis, we considered only visitors

whose trip was attributable to the cultural experience of the
Christmas market only and not to other reasons. We perform the
analysis rst considering all visitors, and to have a clear-cut image
of the phenomenon, we split the sample into one-day visitors and
multiple-day visitors. In the present paper, for each subsample, we
ran a set of four regressions to better understand the relationship
between the factors that inuence visits per capita and how they
are related to travel costs and other important variables but also to
gradually check for coefcient robustness.
5.2. Authenticity and Christmas markets
Empirical research on Christmas markets allocates an important
role to the perceived authenticity of a Christmas market that in
ran &
uences decisions about attending the markets (Caste
Roederer, 2013) and shopping at a Christmas market (Brida et al.,
2013b, 2013a). To account for the authenticity dimension, we performed principal component analysis (PCA) to reduce the number
of variables to be introduced in the models without a signicant
loss of information (Hair, Anderson, & Tatham, 1989).
PCA is a mathematical procedure that seeks to convert a set of
correlated variables into a set of values of linearly uncorrelated
variables called principal components. This technique is widely
used in data processing and dimensionality reduction. PCA redistributes all variance into orthogonal components where each
principal component is a linear combination of all the original
variables. Specically, PCA reorganizes the variance into a new set
of components equal to the number of original variables. The rst
component is a linear combination of variables that maximizes the
component score variance, and therefore, this component captures
most of the variance. In this paper, PCA is employed using a varimax
rotation method.
To account for the authenticity, we considered the following
variables: the importance of sampling local products, the importance of staying in a unique Christmas atmosphere, visitors
perception of the authenticity of the market products, the
perception of the uniqueness of the Meran Christmas market, the

71

perception that through the market one can learn about local culture, and the perception that the market represents the local
tradition.
We employed Cronbach's alpha, a measure of internal consistency, to test how closely related a set of items as a group are.
Usually, alpha coefcients higher than 0.7 are considered acceptable. This rule should be applied with caution when Cronbach's
alpha has been computed from items that are not correlated
(Nunnally, 1978). In this case, the coefcient was above 0.5
(alpha 0.56) meaning that the items have relatively low internal
consistency. Cronbach's alpha generally increases when the correlations between the items increase, and in this case, it was low
(Table 1).
Therefore, we performed PCA over these six variables together.
Two components were identied (Table 2). Specically, only variables with a factor loading higher than 0.30 were considered, which
indicates an important correlation between the factor and individual items. The rst component considers the last four items and
represents the experienced authenticity while the rst two items,
which represent the expected authenticity, drive the second
component. The scores for each component were used as independent variables in the regression analysis as they summarized
the underlying structure of the most important information for
respondents in relation to the authenticity measures. We calculated
the KaisereMeyereOlkin (KMO) measure of sampling adequacy to
establish the adequacy of the extracted components. The KMO is an
index that lies between 0 and 1 of the proportion of variance among
the variables that might have common variance. A value between
0.50 and 1.00 implies that the analysis is satisfactory (Kaiser, 1974).
The computed index indicates that the analysis was satisfactory
with a KMO value of 0.61. Therefore, the two components were
employed as independent variables in the regression analysis.
6. The results
6.1. Travel cost model
To choose the right estimation model, we conducted a
goodness-of-t test that allowed us to select the best model between the Poisson model and the negative binomial described in
Section 4. The results of this test (goodness-of-t chi2 356.03e
Prob > chi2 0.9935) suggested that the Poisson model empirically
t the data better and thus could be accepted for all models proposed here. As a further assessment, we ran the likelihood ratio
test. It is a test of the over-dispersion parameter alpha. When the
over-dispersion parameter is zero, the negative binomial distribution is equivalent to a Poisson distribution. In this case, the alpha
was not statistically signicantly different from zero (likelihoodratio test of alpha 0: chibar2 0.00 Prob  chibar2 1.000) and
thus reinforced one last time that the Poisson distribution is
appropriate.
In what follows, we proceed with the zero truncated Poisson
estimation. The Wald test indicated that the models were well
specied at the 1% level of signicance.
In the rst model (results are included in Table 3), among the
socio-economics characteristics, the coefcients of the average
education level of the sample (upper secondary) and of the
employment condition turned out to be positive and highly statistically signicant. This implies that people with those characteristics have a greater probability of attending this type of cultural
event. As expected, the coefcient for travel cost was negative and
statistically signicant, signaling that as the travel cost increases
the presence at the event decreases.
To assess the robustness of the coefcients, we extended the
basic specication and included other types of expenses:

72

J.G. Brida et al. / Tourism Management 59 (2017) 67e75

Table 1
Correlation table for authenticity variables.
Importance to
taste local
products
Importance to taste local
products
Importance to stay in a
unique Christmas
atmosphere
Perception about markets
products authenticity
Perception about Meran
market uniqueness
Perception that through CM
one can learn about local
culture
Perception that CM
represents the local
tradition

Importance to stay in a
unique Christmas
atmosphere

Perception about
markets products
authenticity

Perception about
Meran market
uniqueness

Perception that through CM Perception that CM

represents the local
one can learn about local
tradition
culture

1.0000
0.3998a

1.0000

0.0585

0.1606a

1.0000

0.0156

0.1585

0.2651a

0.0315

0.0264

0.3596

0.2157a

1.0000

0.1055

0.0922

0.3079a

0.0945

0.3192a

1.0000

10.000

Notes: adjusted Bonferroni.

a
Denotes 1% level of statistical signicance.

Table 2
Principal component analysis.
Variables
Importance to taste local products
Importance to stay in a unique Christmas atmosphere
Perception about markets products authenticity
Perception about Meran market uniqueness
Perception that through CM one can learn about local culture
Perception that CM represents the local tradition
% of Variance

accommodation costs, transportation costs, food and beverage

costs, and the amount of money spent at the Christmas market
(Table 3, Model 2). Among these expenses, only the food and
beverage cost coefcient was statistically signicant (1%) and
positive, while the other coefcients remained stable. This may be
an indicator of how the public appreciates local food and beverages,
as said, largely represented in the Christmas market.
The third model included a set of controls for the distance from
the visitor's residence that was a proxy for the travel time value
(Table 3, Model 3). Those who came from central Italy were less
likely to attend the event, while coming from northeastern and
southern Italy increases the probability of attending.
The last specication (Table 3, Model 4) included a control for
satisfaction with the event (satisfaction), a control for the probability of suggesting the visit to friends and relatives (suggest), but
also the expected and the experienced authenticity. Among all, suggesting a visit was statistically signicant at the 1% level indicating
that visitors who have a high probability of suggesting a visit to the
Christmas market are also those who tend to repeat the experience.
In addition, expected authenticity displayed a positive and statistically signicant coefcient (5%) meaning that the Meran market is
associated with authenticity that, in turn, encourages visits. This

ran and Roederer (2013)
nding supports previous results by Caste
that showed visitors who nd a Christmas market more authentic
are more likely to repeat their visit to the Christmas market.
When the specications of the four models were compared, the
Akaike information criterion (AIC) and the Bayesian information
criterion (BIC) were minimized in the last model; therefore, there is
statistical evidence to retain the model as a reliable empirical
specication.
This analysis was implemented without differentiating the type
of trip (i.e., a one-day trip or a multiple-day trip). To account for

Experienced authenticity

Expected authenticity
0.7035
0.6967

0.5537
0.3800
0.5687
0.4709
0.3001

0.2363

different types of tourists, we split the selected sample (N 436)

and performed the analysis according to the segment to which the
visitors belonged. We repeated the analysis for two groups of visitors: one-day visitors (N 276) and multiple-day visitors (N 160)
to empirically assess the differences in the consumer surplus according to the different experiences.
The comparison of the two samples demonstrated that one-day
visitors drove the results of the whole sample (Tables 1A and 2A in
the Appendix). Overall, the ndings for one-day visitors are in line
with the main model. Travel costs were negatively related to the
number of visits to the Christmas market. Being employed,
expenditure on food and beverages, place of residence, and intention to suggest a visit to the Christmas market were all positively
and signicantly related to the frequency of attending the Christmas market.
Expectations about the authenticity of the Christmas market
were positively related to attendance. One-day visitors displayed a
signicant, but negative, coefcient as far as the experienced
authenticity was concerned. For experienced authenticity, the
result indicated that visitors with fewer visits to the Christmas
market in Meran are more affected by the overall atmosphere at the
Christmas market and tend to perceive it as more authentic. This
suggests that with increased Christmas market experience, visitors
become less sensitive to experienced authenticity and are probably
looking for new experiences. Overall, the ndings suggest that
visitors are attracted to Christmas markets due to their authenticity,
but as experience with the Christmas market increases, the
perceived authenticity decreases.
6.2. Consumer surplus estimation
Consumer surplus (CS) derived for the Christmas market is

J.G. Brida et al. / Tourism Management 59 (2017) 67e75

Table 3
Results of the Zero Truncated Poisson regression: complete sample.
(1)
Socio-Economics Characteristics
Age
0.0309
(0.044)
Age2
0.0005
(0.001)
Male
0.0945
(0.117)
Married
0.0563
(0.140)
Upper Secondary
0.3455***
(0.123)
Income1
0.0067
(0.163)
Income2
0.0077
(0.146)
Income3
0.4223
(0.273)
Employed
1.1370***
(0.320)
Travel Cost
TC
0.0090***
(0.002)
Other costs
Accommodation
Local transport
Food and Beverage
Expenditure at the Market

(2)

(3)

0.0023
(0.046)
0.0001
(0.001)
0.1085
(0.119)
0.0009
(0.142)
0.2938**
(0.127)
0.0136
(0.172)
0.1360
(0.154)
0.1542
(0.279)
0.8965***
(0.322)

0.0123***
(0.003)

0.0082***
(0.003)

0.0090***
(0.003)

0.0002
(0.002)
0.0153
(0.044)
0.0103***
(0.003)
0.0004
(0.001)

0.0002
(0.002)
0.0080
(0.043)
0.0098***
(0.003)
0.0000
(0.001)

0.0003
(0.002)
0.0158
(0.042)
0.0079***
(0.003)
0.0002
(0.001)

0.7942***
(0.283)
1.4855***
(0.368)
0.3676***
(0.133)
0.8136
(0.578)

0.8394***
(0.283)
2.1218***
(0.416)
0.3743***
(0.134)
0.8200
(0.575)

0.7432
(0.866)
436.0000
952.4944
1029.9696

0.0024
(0.145)
0.8256***
(0.197)
0.1288**
(0.060)
0.0463
(0.044)
1.8439**
(0.904)
436.0000
928.5355
1022.3213

Northeast of Italy
Foreign country
Trip factors
Satisfaction
Suggest
Expected authenticity
Experienced authenticity

N
aic
bic

0.4271
(0.851)
436.0000
990.7924
1035.6465

(4)

0.0304
(0.046)
0.0005
(0.001)
0.1192
(0.118)
0.0603
(0.141)
0.2671**
(0.125)
0.1016
(0.170)
0.1373
(0.153)
0.2601
(0.276)
0.9947***
(0.324)

South of Italy

cons

Table 4
Consumer surpluses.

0.0341
(0.044)
0.0005
(0.001)
0.1061
(0.117)
0.0452
(0.140)
0.3282***
(0.123)
0.0442
(0.163)
0.0227
(0.146)
0.3986
(0.272)
1.1102***
(0.321)

Controls for place of residence

Centre of Italy

0.4539
(0.859)
436.0000
986.1486
1047.3133

73

Standard errors in parentheses.

p < 0.10, **p < 0.05, ***p < 0.01.

related to the use value and specically to the recreational value.

The data employed in the analysis are linked to the revealed preferences approach; consequently, the CS estimate refers to the event
as a whole. The estimate was computed as the rate of change in
visits to the Christmas market regarding travel cost: CS b1 (Haab
TC
& McConnell, 2002).
Table 4 presents the consumer surplus computed according to
sample segmentation for the four specications. The estimates
ranged from a low of V81.3 to a high of V156.3. Understanding the
sources of this variation is important to policy makers.
The tendency among the entire sample and the two subsamples
was similar. The simple models, that is, the basic TCM, display the
CS1 lowest value for one-day visitors (V98) while the multiple-day

Sample segment

CS1

CS2

CS3

CS4

Average

All
One-day visitors
More days visitors

V 111.1
V 98
V 101

V 81.3
V 84
V 100

V 122
V 144.9
V 103.1

V 111.1
V 156.3
V 106.4

V 106.4
V 120.8
V 102.6

visitors sample has a gain of V10 lower than the one obtained by
not splitting the sample. The growth rate between CS1 and CS2 for
the one-day sample was 59% but 5% for multiple-day visitors.
These differences are attributable to the experiences that for the
one-day visitors could be condensed. There was a signicant increase in the CS for one-day visitors when Model 4 was considered.
The analysis presented here did not measure the recreational
value for visitors on organized tours or non-use values; therefore,
the total economic value of the Meran Christmas market is likely to
be a much higher gure than those reported here.
7. Discussion and conclusions
A Christmas market is a cultural and retail event that creates a
use value, identied as the recreational value, for tourists. As
highlighted in the literature, the main reason for attending a
Christmas market is to relax, enjoy the Christmas atmosphere,
spend time with family and friends, and sample local products
(Brida et al., 2012). The growing importance of Christmas markets
in the tourism industry has created the need to estimate the value
for visitors of the recreational experience at Christmas markets that
goes beyond mere shopping occasions and the estimates of the
economic impact. The estimate of the economic or social value
(consumer surplus) of Christmas markets allows one to understand
whether the event is creating a benet for society as a whole.
However, different types of visitors may encounter different
experiences at Christmas markets, and to account for these differences, a basic distinction between one-day visitors and multipleday visitors could provide the rst insight. The analysis showed
that the value of the experience, that is, the recreational value, is, on
average, higher for one-day visitors. It is likely that the experience
of these visitors is more intense and that the authenticity they
perceive plays an important role. Overall, the ndings suggest that
visitors are attracted to Christmas markets because of their
authenticity, but as experience with Christmas markets increases,
the perceived authenticity decreases.
As far as externalities are concerned, a Christmas market may
have a negative impact on the local population caused by the inux
of people and, in particular, congestion and noise. Additionally, the
event could become so commercial that part of its original foundation of authentic traditions and folklore may be lost. A solution
may be to reduce the total number of visitors who attend the main
event and divert visitors to side events. However, Christmas markets have the potential to inuence positively culture-oriented
economic development and foster tourism during the low season.
Although comparing these results with those of other studies is
difcult, given the peculiarities of the event, the estimate of the
direct and indirect use value of a musical event by Andersson,
Armbrecht, and Lundberg (2012) is a useful point of comparison.
The authors found an indirect use value (similar in nature to the
recreational value here) for music festival visitors of around V137
per entry. This gure gives some indication that the estimates
found in this paper are within a plausible range. This benet value
is important from a policy point of view. Monetary estimates of
Christmas market visitors can assist management decisions, such as
for the application of entry fees or for the introduction of collateral
events. Moreover, the Christmas market in Meran, similar to other

74

J.G. Brida et al. / Tourism Management 59 (2017) 67e75

popular events in Italy, receives public subsidies from the local

government. The calculation of the consumer surplus generated by
the event constitutes one of several measures that governments
can use to allocate scarce public funds to competitive uses, in
particular, in heritage and cultural domains.
A comparison of the present results with other studies concerning Christmas markets was not possible as the present paper is,
to the best of our knowledge, the rst study in which this methodology was applied to Christmas markets.
Although the travel cost model applied in this study estimates
benet measures based on observed behavior, this model provides
only a partial measure of the total benets for Christmas market
visitors. This is because non-use values (existence, option, and
bequest values) are not reected in the benet measures estimated
here. The net values estimated in this paper represent only the use
values of current tourists. Additional studies that measure consumer surplus at different Christmas markets in different countries
are needed in order to compare the use value of this type of event
for tourists and for residents. Christmas markets create, in addition
to economic impacts, a use value for the hosting community that
benets from the cultural exchange brought by tourists and of the
urban revitalization during the low season.

Table 1A (continued )
(1)

Local transport
Food and Beverage
Expenditure at the Market
Controls for place of residence
Centre of Italy
South of Italy
Northeast of Italy
Foreign country
Trip factors
Satisfaction

(2)

(3)

(4)

0.0244
(0.054)
0.0005
(0.001)
0.0396
(0.152)
0.0674
(0.172)
0.0300
(0.152)
0.3929*
(0.219)
0.2904
(0.197)
0.2749
(0.399)
1.6378***
(0.505)

0.0333
(0.054)
0.0006
(0.001)
0.0170
(0.152)
0.0207
(0.175)
0.0214
(0.154)
0.3985*
(0.219)
0.3392*
(0.198)
0.3312
(0.397)
1.5748***
(0.508)

0.0042
(0.055)
0.0001
(0.001)
0.0526
(0.156)
0.0110
(0.176)
0.0486
(0.158)
0.2101
(0.224)
0.3249
(0.204)
0.1506
(0.406)
1.4194***
(0.508)

0.0142
(0.004)

***

0.0026
(0.009)
0.0315
(0.048)
0.0114***
(0.003)
0.0009
(0.001)

0.0097
(0.004)

**

0.0094
(0.004)
0.0029
(0.010)
0.0339
(0.047)
0.0073**
(0.004)
0.0007
(0.001)

0.7871**
(0.398)
13.0031
(1356.523)
0.5307***
(0.199)
14.2508
(672.499)

0.7546*
(0.399)
12.6876
(1613.330)
0.4872**
(0.199)
14.5182
(740.249)
0.0484

1.4326
(1.089)
276.0000
587.1767
655.9643

(0.178)
1.2274***
(0.284)
0.2012**
(0.094)
0.1489***
(0.053)
2.8390**
(1.169)
276.0000
554.8200
638.0892

Experienced authenticity
cons
N
aic
bic

1.0654
(1.082)
276.0000
617.9241
657.7485

1.2091
(1.100)
276.0000
613.9704
668.2764

Standard errors in parentheses.

p < 0.10, **p < 0.05, ***p < 0.01.

Table 2A
Results of the Zero Truncated Poisson regression: more-days visitors
(1)
Socio-Economics Characteristics
Age
0.0370
(0.087)
Age2
0.0004
(0.001)
Male
0.1803
(0.191)
Married
0.2564
(0.258)
Upper Secondary
0.8479***
(0.230)
Income1
0.4568*
(0.257)
Income2
0.4030*
(0.229)
Income3
0.7689**
(0.371)
Employed
0.2248
(0.488)
Travel Cost
TC
0.0102***
(0.004)
Other Costs
Local transport
Food and Beverage
Expenditure at the Market

(2)

(3)

(4)

0.0533
(0.089)
0.0006
(0.001)
0.1907
(0.192)
0.2063
(0.260)
0.7736***
(0.233)
0.3963
(0.260)
0.2963
(0.235)
0.6016
(0.379)
0.2148
(0.488)

0.0054
(0.096)
0.0002
(0.001)
0.2674
(0.206)
0.0021
(0.262)
0.7302***
(0.239)
0.3470
(0.278)
0.0924
(0.253)
0.3511
(0.394)
0.2424
(0.516)

0.0096
(0.100)
0.0000
(0.001)
0.2148
(0.213)
0.0385
(0.265)
0.7124***
(0.244)
0.3371
(0.282)
0.1355
(0.253)
0.3101
(0.400)
0.0895
(0.525)

0.0119***
(0.004)

0.0069
(0.004)

0.0064
(0.004)

0.1065
(0.107)
0.0053
(0.006)
0.0024
(0.002)

0.0862
(0.103)
0.0033
(0.007)
0.0014
(0.002)

0.0907
(0.106)
0.0027
(0.007)
0.0008
(0.002)

0.7284*
(0.439)
1.7676***
(0.471)
0.2358
(0.226)
0.0898
(0.654)

0.6862
(0.440)
2.4570***
(0.821)
0.2136
(0.232)
0.1543
(0.655)

0.3479
(1.687)
160.0000
362.8834
418.2365

0.2942
(0.259)
0.2111
(0.266)
0.1705
(0.123)
0.1102
(0.086)
0.4953
(1.775)
160.0000
365.6042
433.2580

Controls for place of residence

Centre of Italy

**

0.0055
(0.009)
0.0373
(0.048)
0.0120***
(0.003)
0.0007
(0.001)

(4)

Expected authenticity

Table 1A
Results of the Zero Truncated Poisson regression: one-day visitors
(1)

(3)

Suggest

Appendix

Socio-Economics Characteristics
Age
0.0253
(0.053)
Age2
0.0004
(0.001)
Male
0.0540
(0.151)
Married
0.0516
(0.169)
Upper Secondary
0.0429
(0.151)
Income1
0.3448
(0.216)
Income2
0.3006
(0.196)
Income3
0.2701
(0.400)
Employed
1.6380***
(0.503)
Travel Cost
TC
0.0099***
(0.003)
Other costs
Accommodation

(2)

South of Italy
Northeast of Italy
Foreign country
Trip factors
Satisfaction
Suggest
Expected authenticity
Experienced authenticity
cons
N
aic
bic

0.7265
(1.601)
160.0000
371.4437
405.2706

Standard errors in parentheses.

p < 0.10, **p < 0.05, ***p < 0.01.

0.9999
(1.631)
160.0000
372.2336
415.2860

J.G. Brida et al. / Tourism Management 59 (2017) 67e75

References
Andersson, T. D., Armbrecht, J., & Lundberg, E. (2008). Impact of mega-events on the
economy. Asian Business & Management, 7(2), 163e179.
Andersson, T. D., Armbrecht, J., & Lundberg, E. (2012). Estimating use and non-use
values of a music festival. Scandinavian Journal of Hospitality and Tourism, 12(3),
215e231.
Armbrecht, J. (2014). Use value of cultural experiences: A comparison of contingent
valuation and travel cost. Tourism Management, 42, 141e148.
Bedate, A., Herrero, L. C., & Sanz, J. A. (2004). Economic valuation of the cultural
heritage: Application to four case studies in Spain. Journal of Cultural Heritage,
5(1), 101e111.
Brida, J. G., Disegna, M., & Osti, L. (2012). Segmenting visitors of cultural events by
motivation: A sequential non-linear clustering analysis of Italian Christmas
market visitors. Expert Systems with Applications, 39(13), 11349e11356.
Brida, J. G., Disegna, M., & Osti, L. (2013a). The effect of authenticity on visitors'
expenditure at cultural events. Current Issues in Tourism, 16(3), 266e285.
Brida, J. G., Disegna, M., & Osti, L. (2013b). Visitors' expenditure behaviour at cultural events: The case of Christmas markets. Tourism Economics, 19(5),
1173e1196.
Brida, J. G., Meleddu, M., & Pulina, M. (2013c). The economic impacts of cultural
tourism (Chapter 2.14). In M. Smith, & G. Richards (Eds.), The book handbook of
cultural tourism (Series: Routledge handbooks) (1st ed., pp. 110e115). , New York:
Routledge, ISBN 9780415523516.
Brida, J. G., & Tokarchuk, O. (2015). Keeping mental budgets: Visitors' spending at a
Christmas market. Tourism Economics, 21(1), 67e82.
Cameron, A. C., & Trivedi, P. K. (2013). Regression analysis of count data. Cambridge
University Press.

ran, H., & Roederer, C. (2013). Does authenticity really affect behavior? the case
Caste
of the Strasbourg Christmas Market. Tourism Management, 36, 153e163.
CRR Research. (2014). Christmas markets: 2014 review. Retrieved September 24, from
Centre for retail research (CRR) website. http://www.retailresearch.org/
christmasmarkets.php.
Diamond, P. A., & Hausman, J. A. (1994). Contingent Valuation: Is some number
better than No number? Journal of Economic Perspectives, 8(4), 45e64.
Haab, T. C., & McConnell, K. E. (2002). Valuing environmental and natural resources:
The econometrics of non-market valuation. Northampton, MA: Edward Elgar
Publishing Inc.
Hair, J. F., Anderson, R. E., & Tatham, R. L. (1989). Multivariate data analysis. New
York: MacMillan Publishing Company.
Hicks, J. R. (1946). Value and capital: An inquiry into some fundamental principles of
economic theory. Oxford: Oxford University Press.
Hotelling, H. (1947). Letter cited in the economics of public recreation: An economic
study of the monetary evaluation of recreation in the national parks. Washington,
D.C., U.S.: National Park Service.
Hurwicz, L., & Uzawa, H. (1971). On the integrability of demand functions. In
J. S. Chipman, L. Hurwicz, M. K. Richter, & H. F. Sonnenschein (Eds.), 1971.
Preferences, utility, and demand: A Minnesota symposium (pp. 114e148). New
York: Harcourt, Brace, Jovanovich (Chapter 6).
Kaiser, H. F. (1974). An index of factorial simplicity. Psychometrika, 39(1), 31e36.
Nunnally, J. C. (1978). Psychometric theory (2nd ed.). New York: McGraw-Hill.
Poor, J. P., & Smith, J. M. (2004). Travel Cost Analysis of a Cultural heritage site: The
case of Historic St. Mary's City of Maryland. Journal of Cultural Economics, 28,
217e229.
Prayaga, P., Rolfe, J., & Siden, J. (2006). A travel cost analysis of the value of special
events: Gemfest in General Queensland. Tourism Economics, 12(3), 403e420.
Smith, H. D., Su

arez de Vivero, J. L., & Agardy, T. S. (2015). Routledge handbook of
ocean resources and management (p. 612). London and New York: Routledge,

75

ISBN 978-0-415-53175-7.
Sohngen, B., Lichtkoppler, F., & Bielen, M. (1999). The Value of Day Trips to Lake Erie
Beaches. Technical Bulletin OHSU e TB e 039. Dept. of Agricultural, Environmental,
and Development Economics. Columbus, Ohio: Ohio State University. Ohio Sea
Grant College Program.
Tourkolias, C., Skiada, T., Mirasgedis, S., & Diakoulaki, D. (2015). Application of the
travel cost method for the valuation of the Poseidon temple in Sounio, Greece.
Journal of Cultural Heritage, 16(4), 567e574.
Vicente, E., & De Frutos, P. (2011). Application of the travel cost method to estimate
the economic value of cultural goods: Blockbuster art exhibitions. Revista de
Economia Pblica, 196(1), 37e63.
Willis, K. G., Snowball, J. D., Wymer, J. G., & Grisola, J. (2012). A count data travel
cost model of theatre demand using aggregate theatre booking data. Journal of
Cultural Economics, 36(2), 91e122.

Juan Gabriel Brida is Full Professor of Economic Dynamics

at the Faculty of Economics and Management e Universidad de la Repblica - Uruguay. His research interests
and expertise are in the areas of tourism economics and
economic growth. He has a degree in Mathematics from
the Universidad de la Republica (Uruguay) and a PhD in
Economics from the University of Siena.

Marta Meleddu, is a Research Fellow at the Department of

Political Science, Communication, Engineering and Information Technologies (POLCOMING) and CRENoS, University of Sassari, Piazza Universita 11, 07100 Sassari, Italy. Her
main research interests cover applied economics (cultural,
tourism, environment, consumer choice and behavior) and
applied econometrics.

Oksana Tokarchuk, is assistant professor of Management

and Marketing at the Competence Centre in Tourism
Management and Tourism Economics (TOMTE) of the
Faculty of Economics and Management at Free University
of Bolzano. She holds PhD in Economics and Management.
Her main research interests include behavioral economics,
consumer behavior, quality of life and applied statistics/
econometrics.