Content uploaded by Christos Karelakis
Author content
All content in this area was uploaded by Christos Karelakis on Jun 08, 2016
Content may be subject to copyright.
Available online at www.sciencedirect.com
Procedia Economics and Finance 00 (2014) 000–000
www.elsevier.com/locate/
procedia
“The Economies of Balkan and Eastern Europe Countries in the changed world,
EBEEC 2014, Nis, Serbia”
Assessing exports market dynamics: the case of Greek
wine exports
Eleni Koutroupia, Dimitrios Natosb, and Christos Karelakisc*
aMSc. Democritus University of Thrace, Department of Agricultural Development
Pantazidou 193, 68200, Orestiada, Greece
bPh.D, Aristotle University of Thessaloniki, Department of Agricultural Economics,
University Campus, 54124, Thessaloniki, Greece,
c*Assistant Professor, Democritus University of Thrace, Department of Agricultural Development, Pantazidou
193, 68200, Orestiada, Greece
Abstract
Competitiveness in international wine markets has intensified during the recent years. The dynamic and revitalized
presence of non-European producers (e.g. Australia, Chile, South Africa, USA), who have seen their market shares
rising rapidly and the subsequent reduction in the market shares of traditional European producers (e.g. France,
Italy, Greece) alongside a slight decrease of wine consumption worldwide, have shaped new trends for wine
markets internationally and contemporary challenges for the traditional European wine exporters. The aim of the
present study is to investigate the determinants of the Greek wine exports and to assess its competitiveness in the
European market. Primary data regarding wine trade has been collected for the period of 2004-2011, referring to
four main European producer countries, namely Greece, France, Germany and Bulgaria. A gravity equation model
has been estimated in order to examine the factors that determine wine exports and wine exports competitiveness
among those countries. The results offer a clear picture of the wine trade dynamics in the EU market, indicating as
key determinants for its competitiveness the size of each country’s economy, the geographical distance from other
EU countries, the presence of common borders or common language among trading nations as well as the size of
per capita wine consumption to the European wine markets.
© 2014 The Authors. Published by Elsevier B.V.
Selection and/or Peer-review will be under responsibility of Department of Accountancy and Finance, Eastern
Macedonia and Thrace Institute of Technology, Kavala, Greece.
Keywords: wine trade, gravity equation model, competitiveness
1. Introduction
The wine industry has undergone significant changes recently due to market liberalization and the
existence of new producer countries like the United States, Chile and South Africa along with changes
in demand due to favored consumer preferences towards other alcoholic drinks. Despite its drop in
consumption, wine trade has increased significantly over the years, proving that it is a highly
internationally traded product. The industry has globalized, trying to respond effectively to increased
competition in international markets and the emergence of new ones. Within this context, the European
Union plays an important role since most of the trading quantities to international markets involve
European countries. The EU-27 is the world’s leader in wine production, producing approximately 141
million hectoliters (Mhl) of wine, with major producing countries being France, Italy and Spain,
followed by Germany, Portugal, Romania, Greece, and Austria (Table 1).
* Christos Karelakis. Tel.: +30-25520-41110; fax: +30-25520-41191.
E-mail address: chkarel@agro.duth.gr
2212-5671 © 2014 The Authors. Published by Elsevier B.V.
Selection and/or Peer-review will be under responsibility of Department of Accountancy and Finance, Eastern Macedonia and
Thrace Institute of Technology, Kavala, Greece.
E. Koutroupi, D. Natos, C. Karelakis. /Procedia Economics and Finance 00 (2014) 000-000
The EU-27 is also a leader in international trade, exporting 22.3 Mhl valued at 8.3 billion euro,
while importing 13.6 Mhl valued at 2.36 billion euro. The United States remains the leading export
market for the EU-27 as a whole (European Commission 2014). Main suppliers are Australia, Chile,
South Africa, and the United States.
Table 1: Wine production* trend in the EU-27 (‘000 Hectoliters)
2007 2008 2009 2010 2011 2012
France 45672 41640 46269 45669 50757 40609
Italy 42514 46245 45800 46737 42705 39300
Spain 36408 35913 36097 35363 33397 31500
Germany 10261 9991 9228 6906 9132 8903
Portugal 6074 5620 5872 7133 5610 5857
Romania 5289 5159 6703 3287 4213 4059
Greece 3511 3869 3366 2950 2750 3150
Austria 2256 2628 2672 2352 1737 2815
Hungary 3222 3460 3198 1762 2750 1874
Other EU-27
Countries
3853 3604 3034 2616 3177 2773
EU-27 159060 158129 162238 154775 155671 140840
As far as per capita wine consumption is concerned, it has fallen during the years mainly because of
the emergence of other alcoholic drinks (i.e. beer) that have seen their market shares rising, changing
consumer preferences and lifestyles towards a more healthy living, along with anti-alcohol campaigns
and prevailing laws in many countries that determine the consumption of alcoholic beverages by age,
but also prohibits the wide use of alcoholic beverages to avoid road accidents . This is mainly evident to
Southern European countries, whereas in the Northern part of the EU, consumption has remained flat
or even slightly increased during the years characterized by the domination of PDO/PGI wines and a
noticeable demand for low-cost bulk wines.
During the last decade, Greek wine exports displayed relative decline that resulted to subsequent
reduced wine production value growth rates. The unfavorable global economic environment as well as
the local economic downturn and a severe recession have resulted to a noticeable decline of wine
exports worldwide. The Greek wine is exported to almost all countries of the world. In all of the
European Union absorbs 85.2% of total wine export, by main suppliers Germany and France. However,
exports to third countries cannot touch on the whole the same proportion as in the EU, but also absorb
large quantities of wine. The U.S. and Canada absorb larger amounts of wine compared with other third
countries, at rates reaching 42.8% and 14.2% respectively of the total (GAIN 2012).
This assignment will be assessed through an econometric model to explain the size of the wine
produced and trade flows from Greece in the major importing countries through the approach of the
“gravity model”. The model used and the results achieved will be useful to predict likely trends in
exports Greek wine. Taken into account macro - variables such as wine production, the AEP,
population, etc. In addition, this model shows the ability of the participants to the disposal of the wine
market and the potential to promote the product.
2. Gravity equation model
The utilization of Gravity Equation Model (GEM) as a method for the empirical analysis of
international trade and foreign direct investment flows is widespread among the relevant literature of
international economics and trade (Cheng and Wall, 2005). Several empirical studies have found, that
the volume of trade between any two countries is given by:
Tij = A * Yi * Yj / Dij (1)
where A is a constant, Tij is trade between country i and country j, Yi is the GDP country i, Yj is the
GDP country j and Dij is the geographical distance between the two countries. The above link shows
that bilateral trade between two countries is proportional to the product of GDP the two countries and
inversely proportional to the geographical distance between the two countries. This is the basic form of
the gravity model based on the theory of Newton, who expressed that the gravitational attraction
between two objects is proportional to the product of the masses and decreases with the distance
between them. Therefore, all GEM applications are based on the notion that the value of exports or
2
E. Koutroupi, D. Natos, C. Karelakis/Procedia Economics and Finance 00 (2014) 000-000
imports among any two trading countries is an increasing function of their economic sizes and a
decreasing function of the physical distance between them. The first major theoretical justification for
the GEM came from Anderson in 1979 (Anderson, 2011) and more recently in 2013 by Bergstrand
Egger and Larch (2013) who built a theoretically consistent GEM based on the standard Krugman
model of monopolistic competition and increasing returns. In empirical analyses, using Yi and Yj as
variables of the economic sizes of two trading nations i and j (national incomes) and Dij as the capitals’
distance among them, the GEM in log-linear form for the natural logarithm of trade value (Tij) is stated
as:
(2)
where lnTij consist the dependent variable of our GEM and lnYi, lnYj and lnDij will function as
independent variables.
3. Model specification and methodology of estimations
Following the scope of this paper, one augmented GEM for exports was estimated for each investigated
country.
(3)
The variables of the above model are defined as follows:
Xij,t denotes the value of wine exports of Greece ( i=1), France (i=2), Germany (i=3) or Bulgaria (i=4)
respectively, to the EU27 countries (countries j) in each year t,
α0 is the constant term of the equation,
Υi,t and Yj,t represent countries i (Greece, France, Germany and Bulgaria) and j GDP in year t,
Dij denotes the physical distance, originating from the CEPII gravity database (CEPIIa, 2013) between
country i and country j measured as the great circle kilometers distance between capitals,
Euroij represents a binary variable that equals one if country i and j is a member of common euro
currency. That particular binary variable is used in order to investigate firstly the differences between
the value of trade flows to the member states of the EU and to the member states of EMU and
secondly, in order to examine if wine trade is influenced by the adoption of the common currency.
ComLangij represents a dummy variable, originating from the CEPII gravity database (CEPIIb, 2013)
that equals one if Bulgaria, Greece, France or Germany respectively and country j share a common
language. The common language variable is used in order to investigate whether cultural similarities
between wine trading partners exchange more wine in relation to the countries that do not share a
common language.
Landlockedj is a dummy variable that equals one if country j is landlocked. Landlocked dummy is used
in order to assess whether the presence of an inland country as trade partner is favoring or not wine
exports and imports.
Contigij represents a binary variable that equals one if country i and j share a common border and zero
otherwise.
lnWineConi represents countries i natural logarithm of wine per capita consumption. Wine
consumption sizes are used in order to depict that higher per capita consumption in the four
investigated countries leads to higher exports and imports respectively.
lnWineConj represents countries j natural logarithm of wine pre capita consumption. Wine consumption
sizes for the trading partners are incorporated to our estimated model in order to show that higher wine
per capita consumption is a decisive factor to the attractiveness of wine exports and imports.
lnVinei represents countries i natural logarithm of vines cultivation area. Vines area is used to show that
the larger the area of cultivated vines in the four investigated countries, the greater will be the exports
to other EU countries.
lnWineProdj represents countries j natural logarithm of wine production sizes of the EU27 countries
that compose our trading partners.
3
E. Koutroupi, D. Natos, C. Karelakis. /Procedia Economics and Finance 00 (2014) 000-000
αn represents the parameters to be estimated,
εij,t is the error term of the equation.
The model was estimated eight times, one time for each investigated country i=1 (Greece), i=2
(France), i=3 (Germany) and i=4 (Bulgaria). In order to identify and evaluate the factors that affect the
wine trade in EU a series of properly defined dummies (common language, common border, common
currency, the case for landlocked partner) and real variables (wine consumption, wine production and
vines land cultivation) were enriching the estimated model. Methodologically, the inclusion of the
aforementioned dummies and real variables captured the magnitude of various agricultural,
geographical, economical and cultural factors that favor or not EU wine trade. The difference between
the magnitude of each dummy and real variable, among the four investigated countries, offered insights
to the wine exports competitiveness in EU market and the factors influencing wine exports and
imports.
The dependent variable of the model consisted of the values of Greek, Bulgarian, French and
German wine exports with their current EU partners over eight years period (2004-2011). Observations
of wine imports and exports were classified at the 112.1 code, according to the Standard International
Trade Classification Revision 3 (SITC Rev. 3) and originated from the UNcomtrade database (United
Nations Commodity Trade Statistics Database) of United Nations (UN, 2009). Data on wine
consumption and production were extracted from the Wine Institute (2013) while GDP and population
figures of the 27 EU member states were gathered from the statistical web service of the International
Monetary Fund (IMF, 2012). Both trade observations and GDP figures were deflated using the U.S.
consumer price index (Consumer Price Index - CPI). CEPII database constituted the source of
geographical distance figures that comprised the distance variable as well as the origin of data for the
synthesis of all dummy variables utilized (CEPIIa, 2013).
4. Results
The first model referred to wine exports from Bulgaria to the 27 EU member states and in order to
examine the statistical significance of the coefficients and the variables in the model we used the t-
statistic coefficient. In particular, for the first valued model where i = 1, the following coefficients were
statistically significant (Table 2):
•lngdpp (= 2,192106) with (t = 3.24, P = 0.002). Specifically, for countries with GDP larger than
Bulgaria is statistically significant. Bulgaria has presented increased exports by 2.19%.
•lndist (= -3,938663) with (t = 3,20, p = 0.002). For countries located away from Bulgaria, the
results show a reduction in its exports by 3.93%.
•lnw_pc_con-r (= -2,086325) with (t = 3,20, p = 0.002). The increase in exports to other members
of the EU causes a decrease in the domestic consumption of Bulgaria by 2.08%.
In addition, to examine the statistical significance of the overall model, the relation Prob>F is
compared to 0.05. Based on the results, shown in Table 2, and given that Prob corresponding to F is
equal to 0 < 0.05 implies that the overall model is statistically significant. Another element is the R-
squared, which expresses the explanatory power of the model and according to the value of the 47.45%
of the variability of the dependent variable explained by the variability of the explanatory variables of
the model and the remaining 52.55% of random error terms.
4
E. Koutroupi, D. Natos, C. Karelakis/Procedia Economics and Finance 00 (2014) 000-000
Table 2: Wine export dynamics for Bulgaria- Gravity Model estimation results
The second model refers to the export of wines from France to the 27 member-countries of the
European Union (Table 3). The statistical significance of the coefficients and the variables in the
respected model were assessed though the t-statistic coefficient.
Particularly, for the first valued model i = 2 statistically important are the following factors:
•lngdpp (= 0,4952835) with (t = 4,57, p = 0,000 ). Specifically, for countries with GDP similar to
France , the results shows an increase in exports by 0.49% .
•lndist (= -1,296138) with (t = 5,43, p = 0.000), are statistically significant. For countries that are
located far from France, the results show a decrease in its exports by 1.29%.
•landlock (= 1,058499) with (t = 3,89, P = 0.000) are statistically significant . Neighboring
countries of France, not surrounded by the sea and located in the inland of Europe, exports are
increased by 1.05 % compared with the countries that are surrounded by sea.
•synora (= 0,7981715) with (t = 2,26, P = 0.026 ) are statistically significant , because the neighbor
countries have common borders with France, thus there is an increase in exports by 0.79%.
•lnw_con_par (= 0,2139528) with (t = 2,53, p = 0,013 ). The consumption of French wine by
member-countries of the European Union, the French wine export has increased by 0.21%.
•lnw_pc_con-r (= -0,3471835) with (t = 2,11, p = 0,038 ). The increase in exports to other countries
of the EU, causes a decrease in the domestic consumption of France by 0.34%.
Based on the results of the table and given that Prob corresponding to F is equal to 0 < 0.05 ultimately
means that the whole model is statistically significant. Another element is the R-squared, which
expresses the explanatory power of the model and according to the value of the 78.86 % of the
variability of the dependent variable explained by the variability of the explanatory variables of the
model and the remaining 21.14 % of random error terms.
Source SS df MS Number of obs = 104
F(9, 94) = 9.43
Prob > F= 0.0000
R-squared = 0.4745
Adj R-squared = 0.4242
Root MSE= 3.1098
Model
Residual
820.755568
909.075006
9
94
91.1950631
9.6710107
Total 1729.83057 103 16.7944716
lnwx Coefficient Std. Err. t P>|t| [95% Confidence
Interval]
lngdpr
lngdpp
lndist
euro
com_lang
landlock
synora
lnw_con_par
lnw_pc_con~r
lnvineyard~r
lnwine_pro~r
_cons
-5.692022
2.192106
-3.938663
(omitted)
(omitted)
-.1840906
-2.183171
-1.007722
-2.086325
.3017095
-2.300665
174.4149
7.95708
.6758371
1.230919
1.018163
1.862407
.727342
.651725
10.80571
8.908997
208.9416
-0.72
3.24
-3.20
-0.18
-1.17
-1.39
-3.20
0.03
-0.26
0.83
0.476
0.002
0.002
0.857
0.244
0.169
0.002
0.978
0.797
0.406
-21.49099
.8502152
-6.3826
-2.205677
-5.881024
-2.451877
-3.38034
-21.15328
-19.98969
-240.4436
10.10695
3.533996
-1.494645
1.837496
1.514683
.4364322
-.7923093
21.7567
15.38836
589.2734
5
E. Koutroupi, D. Natos, C. Karelakis. /Procedia Economics and Finance 00 (2014) 000-000
Table 3: Wine export dynamics for France- Gravity Model estimation results
Source SS df MS Number of obs =104
F(11, 92) = 31.19
Prob > F = 0.0000
R-squared = 0.7886
Adj R-squared = 0.7633
Root MSE = .91991
Model
Residual
290.37859
77.8534547
11
92
26.3980536
.846233204
Total 368.232044 103 3.57506839
lnwx Coefficient Std. Err t P>|t| [95% Confidence
Interval]
lngdpr
lngdpp
lndist
euro
com_lang
landlock
synora
lnw_con_par
lnw_pc_con~r
lnvineyard~r
lnwine_pro~r
_cons
5.002332
.4952835
-1.296138
-.1804601
.3727013
1.058499
.7981715
.2139528
-.3471835
-.1980385
1.818317
-155.6294
6.73686
.1084336
.2389079
.2387684
.46357
.2718132
.3530379
.0845581
.1646621
6.993049
3.677952
282.1821
0.74
4. 57
-5.43
-0.76
0.80
3.89
2.26
2.53
-2.11
-0.03
0.49
-0.55
0.460
0.000
0.000
0.452
0.423
0.000
0.026
0.013
0.038
0.977
0.622
0.583
-8.377652
.279925
-1.77063
-.6546748
-.5479888
.5186542
.0970076
.0460131
-.6742166
-14.08684
-5.486413
-716.0673
18.38232
.710642
-.8216467
.2937546
1.293391
1.598343
1.499335
.3818924
-.0201503
13.69076
9.123047
404.8086
The third model (Table 4) represents Germany and its wine exports to the 27 countries of the
European Union. Therefore, to examine the statistical significance of the coefficients and the variables
in the appreciated model via software, the statistical coefficient t-statistic is used. More specifically, if
t> 1.69 leads to a coefficient that is statistically significant. Otherwise, when t <1.69, this means that
the coefficient concerned is not statistically significant. In particular, for the first valued model i = 3
statistically significant are the following factors:
•lnngdpp (= 0,6425027) with (t = 9,36, P = 0.000) . Specifically, for countries with GDP higher
than Germany’s, Germany has increased its exports to these countries by 0.64%.
•lndist (= -1,261428) with (t = 4,54, P = 0.000) are statistically significant . For countries that are
far from Germany, Germany shows a decrease in its exports by 1.26%.
•com_lang (= 0,8732134) with (t = 2,37, p = 0,020) is statistically significant. Among the countries
that share a common language with Germany, Germany 's exports to these countries have
increased by 0.87%.
•lnw_pc_con-r (= -0,5453863) with (t = 3,92, p = 0,000 ) . The increase in exports to other
members of the EU causes a reduction in the domestic consumption in Germany by 0.54%.
Therefore, to examine the statistical significance of the whole model the relation Prob>F is compared
to 0.05. Based on the results of the table and given that Prob corresponding to F is equal to 0 < 0.05
one can say that the whole model is statistically significant. Furthermore, another factor is the R-
squared, which reflects the explanatory power of the model and according to the value of the 78.44 %
of the variability of the dependent variable explained by the variability of the explanatory variables of
the model and the remaining 21.56 % of random error terms.
6
E. Koutroupi, D. Natos, C. Karelakis/Procedia Economics and Finance 00 (2014) 000-000
Table 4: Wine export dynamics for Germany - Gravity Model estimation results
Source SS df MS Number of obs =104
F(11, 92) = 30.44
Prob > F = 0.0000
R-squared = 0.7844
Adj R-squared = 0.7587
Root MSE = .80532
Model
Residual
217.124409
59.6661766
11
92
19.7385826
.648545397
Total 276.790585 103 2.68728724
lnwx Coefficient Std. Err t P>|t| [95% Confidence
Interval]
lngdpr
lngdpp
lndist
euro
com_lang
landlock
synora
lnw_con_par
lnw_pc_con~r
lnvineyard~r
lnwine_pro~r
_cons
9.932341
.6425027
-1.261428
.2508775
.8732134
.4730747
-.0538
.1236742
-.5453863
-281.6306
-1.925507
3246.147
14.45745
.0686718
.2775535
.2254683
.3685271
.3003561
.3222247
.0676498
.1390176
351.7552
3.217183
3993.201
0.69
9.36
-4.54
1.11
2.37
1.58
-0.17
1.83
-3.92
-0.80
-0.60
0.81
0.494
0.000
0.000
0.269
0.020
0.119
0.868
0.071
0.000
0.425
0.551
0.418
-18.7814
-.5061147
-1.812674
-.1969221
.1412867
-.1234584
-.6937661
-.0107311
-.8214872
-980.2468
-8.31511
-4684.694
38.64608
.7788907
-.7101833
.6986771
1.60514
1.069608
.5861661
.2579855
-.2692855
416.9856
4.464095
11176.99
The fourth model simulates Greece and its exports of wine to the 27 (twenty seven) members of the
European Union (Table 5). In order to test the statistical significance of the coefficients and the
variables in the model via software, the statistical coefficient t-statistic is used. More specifically, if t>
1.69 leads to a coefficient that is statistically significant. Otherwise, when t <1.69, this means that the
coefficient concerned is not statistically significant.
Table 5: Wine export dynamics for Greece- Gravity Model estimation results
Source SS df MS Number of obs = 104
F(11, 92) = 11.20
Prob > F = 0.0000
R-squared = 0.5725
Adj R-squared = 0.5214
Root MSE = 3.0135
Model
Residual
1118.94307
835.468948
11
92
101.7220198
9.08118421
Total 1954.41202 103 18.974874
lnwx Coefficient Std. Err t P>|t| [95% Confidence
Interval]
lngdpr
lngdpp
lndist
euro
com_lang
landlock
synora
lnw_con_par
lnw_pc_con~r
lnvineyard~r
lnwine_pro~r
_cons
1.101641
1.838511
.8280754
-1.870969
9.33692
-2.458451
5.032894
.457515
-.1449168
58.85727
-5.986906
-736.6743
7.967419
.2627432
.9539041
.7971099
1.792385
.844662
1.997603
.2622422
.5288912
97.63017
7.831569
1294.221
0.14
7.00
0.87
-2.35
5.21
-2.91
2.52
1.74
-0.27
0.60
-0.76
-0.57
0.890
0.000
0.388
0.021
0.000
0.005
0.013
0.084
0.785
0.548
0.447
0.571
-14.72234
1.31668
-1.06646
-3.459098
5.777089
-4.136023
1.065483
-.0633207
-1.19534
-135.0447
-21.54108
-3307.109
16.92562
2.360341
2.722611
-.2878403
12.89675
-.7808797
9.000305
.9783506
.9055068
252.7592
9.567266
1833.76
Particularly, for the first valued model i = 4 statistically significant are the following factors:
7
E. Koutroupi, D. Natos, C. Karelakis. /Procedia Economics and Finance 00 (2014) 000-000
•lnngdpp (= 1,838511) with (t = 7,00, P = 0.000) . Specifically, for countries with GDP in
comparison to Greece, Greece has increased its exports by 1.83%.
•euro (= -1,876909) with (t = 2,35, P = 0.021 ) are statistically significant. For countries that are
outside the euro zone, Greece has increased its exports by 1.87%.
•com_lang (= 9,33692) with (t = 5,21, p = 0.000),are statistically significant . Among the countries
that have a common language with Greece, Greece's exports to these countries have increased by
9.33 %.
•landlock (= -2,458451) with (t = 2,91, p = 0.005), are statistically significant. For neighboring
countries and not surrounded by the sea and located in the inland of Europe, Greece’s exports have
increased by 2.45 % compared with the countries that are surrounded by sea.
•synora (= 5,032894) with (t = 2,52, P = 0.013 ) are statistically significant for Greece , for the
bordering countries of Greece an increased export by 5.03 %.
Then, to examine the statistical significance of the whole model we focus on the Prob> F again
comparing with 0.05. Based on the results of the table and given that Prob corresponding to F is equal
to 0 < 0.05 we can say that the whole model is statistically significant. Finally, another factor is the R-
squared, which reflects the explanatory power of the model and according to the value of the 57.25 %
of the variability of the dependent variable explained by the variability of the explanatory variables of
the model and the remaining 42.75 % of random error terms.
5. Conclusions
Having completed the study of winemaking, we conclude that one of the major problems of this factor
is the fragmentation of the wine production in Greece, resulting in a dispersion of the supply. There is,
namely, a reduction of production costs, which has an impact on the selling price of the product and
consequently on the consumer. Concerning “the gravity model”, is a great opportunity for the exporters
of Greek wine to export in Europe.
In addition, the wine industry displays problems in the field of consumption. The cost of wine
production configures the selling price of the product on the market. Therefore, the cost of the product
has created ripples in the market, resulting in a decline of the number of the consumers. Furthermore,
the industry of wine faces great competitiveness of other alcoholic beverages both in the domestic
market and in other countries of the European Union. However, in recent years the availability of wines
from the U.S., Argentina, Chile and other countries, has affected the wine market and has as an result
reduced the market share of European wines. The low price of wines from countries like U.S, Chile
etc. has attracted customers, resulting ultimately in the improvement of their wine production. These
problems have led in recent years to reduce exports and sales of Greek wines resulting in an increasing
product stocks to a great extent, which causes deterioration of the quality of production. The production
of Greek wines should increase to seize the beneficial opportunities on the international market.
References
Anderson, J. E. (2011). The Gravity Model. Annual Review of Economics, 3(1), 133-160. doi:
10.1146/annurev-economics-111809-125114.
Bergstrand, J. H., Egger, P., & Larch, M. (2013). Gravity Redux: Estimation of gravity-equation
coefficients, elasticities of substitution, and general equilibrium comparative statics under
asymmetric bilateral trade costs. Journal of International Economics, 89(1), 110-121. doi:
http://dx.doi.org/10.1016/j.jinteco.2012.05.005.
Cheng IH, Wall HJ (2005) Controlling for heterogeneity in gravity models of trade and integration.
Federal Reserve Bank of St. Louis Review 87:49-63.
CEPIIa 2013. Geodesic distances. [Accessed at 09/07/2013 at: http://www.cepii.fr/distance/dist_cepii.zip].
CEPIIa 2013. Geodesic distances. [Accessed at 09/07/2013 at: http://www.cepii.fr/distance/dist_cepii.zip].
CEPIIb, 2013. Data on countries. [Accessed at 09/07/2013 at: http://www.cepii.fr/distance/geo_cepii.xls].
UN, 2009. United Nations Commodity Trade Statistics Database. [Accessed at 15/03/2011 at:
http://comtrade.un.org/db/].
European Commission (2014), Agriculture and Rural Development, Agricultural Markets, “Wine”
[Accessed at 13/10/2014 at:
http://ec.europa.eu/agriculture/markets/wine/facts/index_en.htm
8
E. Koutroupi, D. Natos, C. Karelakis/Procedia Economics and Finance 00 (2014) 000-000
Global Agricultural Information Network, (2012), “Wine Annual Report and Statistics”.
IMF, 2012. International Financial Statistics Online Service, International Monetary Fund. [Accessed
at 15/03/2012 at: http://ifs.apdi.net/imf/].
Wine Institute, 2013. [Accessed at 15/04/2013 at: https://www.wineinstitute.org/resources/statistics]
9
