Download An Application of the Value-Added Based Economic Integration

How to Improve Performance of the Cost-Benefit Analysis of
Monetary Integration? – An Application of the Value-Added
Based Economic Integration Model*
A study by Lars Wang**
Abstract
The aim of this study is the reassessment of the regional integration within the European Union by taking new economic integration measures into account. The core of these economic
integration measures builds the value-added based economic integration (VEI) model improving established economic integration models. The latter models show a poor linkage between
their theoretical foundation and their empirical economic integration measures since these
models only concentrate on the production output for international trade. In contrast to that,
the VEI model emphasizes the national and international production structures and their interconnections allowing its indicators to show the regional distribution of value added induced
by regional trade. Established economic integration models are not able to do so. This is especially true for one of the most frequently applied economic integration measures which is designed by regional exports in relation to gross domestic product. For example, when this
measure exceeds 100 percent then this implies a negative value of domestic non-tradeables.
The new empirical indicators can serve as important ingredients for the cost-benefit analysis
of the choice of an appropriate exchange rate regime for integration areas. This analysis is
based on the theory of optimum currency areas where costs are represented by the economic
stability loss of an economy from joining an exchange rate area and benefits are characterized
by the monetary efficiency gains from pegging the domestic currency.
The results of this study reveal that the actual degree of economic integration decreases when
the VEI model is applied instead of the established gross economic integration (GEI) model.
In addition, the minimum degree of economic integration increases. This indicates the point
where benefits of a country, joining a fixed exchange rate area, outperform the costs. The new
VEI model might show the necessity of reassessing an economy’s consideration to join a
fixed exchange rate area because advises based on the GEI model overestimate the potential
success of a monetary integration process.
JEL classification: C67, E20, F15, F42
Keywords: economic integration degree, optimum currency area, value added approach
*
I am grateful to Ansgar Belke for support of this study. I thank also Philippe Saucier and Sunni Zhitao Sui for
valuable comments. In addition, I show deep appreciation to the Universitätsbund Hohenheim e.V. for strong
financial support of this study.
**
International Economics, Department of Economics, University of Hohenheim, D-70593 Stuttgart, Germany,
[email protected]
Table of contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1
2. Relevance of economic integration measures in the analysis of monetary integration
2
3. Measurement of economic integration with the VEI model . . . . . . . . . . . . . . . . . . . . . . . .
4
3.1 Representing economic interconnections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5
3.2 Modeling the regional distribution of trade-induced value added . . . . . . . . . . . . .
8
3.3 Calculating economic integration measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
11
4. Impact of the VEI model on the analysis of monetary integration . . . . . . . . . . . . . . . . . .
12
4.1 Economic integration measures of members of EU, NAFTA, and MERCOSUR
12
4.2 Systematic differences between alternative economic integration measures . . .
14
4.3 Influence of economic integration models on the cost-benefit analysis . . . . . . . .
18
5. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
23
References
25
........................................................................
1. Introduction
This study presents the impact of a changed foundation of the cost-benefit analysis of monetary integration on the assessment of regional integration. A fundamental part of the analysis
is a country’s degree of economic integration planning to participate in a monetary integration
process. A high level of economic integration indicates a high economic importance of regional trade. The costs and benefits from an economy’s pegging of the domestic currency
depend on the degree of economic integration. Costs represent economic stability losses of an
economy from joining an exchange rate area. They decrease with the degree of economic integration. On the other hand, benefits do increase and characterize monetary efficiency gains.
When benefits are larger then costs at a specific degree of economic integration then a country
should join the other members of a single currency area. But, the value of the economic integration degree in the cost-benefit analysis of monetary integration depends on the operationalization of the economic significance of a country’s trading partners of an integration area. It is
questionable that established economic integration models appropriately represent the importance of regional trading partners. Established economic integration models show a poor
linkage between their theoretical foundation and their empirical economic integration
measures since these models only concentrate on the production output of regional trade
(Wang 2003, pp. 6 f.). E.g., the regional export ratio (RER) of the gross economic integration
(GEI) model can exceed 100 percent implying a negative value of domestic non-tradeables.1
In contrast to the GEI model, the value-added based economic integration (VEI) model of this
study estimates the regional distribution of value added induced by regional trade.2 The VEI
model takes the production input into account by focusing on national and international production structures and the linkages between them.3 This shows an improved theoretical foundation of the empirical indicators and thus enhances their accuracy of measuring the importance of regional trading partners. Hence, the choice of an economic integration model
influences the outcomes of the cost-benefit analysis.
1
This economic integration measure puts regional exports in relation to gross domestic product within a period
of one year to indicate the importance of regional trade at the export side of a country. Furthermore, the regional
import ratio measures it analogously at the import side.
2
This kind of induced value added measures the economic performance of trading economic sectors and their
supplying sectors of a country. The measures of the VEI model are based on that part of regional trade-induced
value added that represents income of production factors in the producer country.
3
The VEI model enhances the value-added based economic openness model of Wang (2003) by splitting up the
trading partners into two groups – those in an integration area and the others outside of it.
-1-
The study proceeds as follows: Section 2 presents the framework of the cost-benefit analysis
of monetary integration. It points out the significance of the economic integration degree for
the analysis. In section 3, the VEI model gets introduced building the theoretical foundation
of a new empirical method to assess the economic relevance of regional trade linkages for an
economy. It shows the modeling of economic relationships within a country, within its trading
partners inside an integration area and those in the rest of the world, and between all of these
economies. Based on this, the section presents a concept to regionally distribute value added
created by trade. With these instruments the economic integration measures get developed.
Subsequently, section 4 presents an empirical analysis of the VEI model’s influence on the
results of the cost-benefit analysis of monetary integration. It calculates economic integration
measures of the GEI and VEI model. Then, the section characterizes the cross-sectional sample consisting of member countries of EU, NAFTA, and MERCOSUR and finally it compares
the consequences of the cost-benefit analysis based on the two models. Section 5 concludes
and discusses the implications of the outcomes for the assessment of monetary relations.
2. Relevance of economic integration measures in the analysis of monetary integration
This section presents the framework of the cost-benefit analysis of monetary integration derived from the theory of optimum currency areas (see Mundell 1961, Gros and Thygesen
1998, pp. 268 ff.). It highlights the position of the degree of economic integration within the
analysis’ foundation. Moreover, the study catches up again the following framework to contrast the impact of value-added based and established indicators on the results of the analysis.
What influences the decision of an economy to join a monetary integration process? An economy has to assess the potential benefits and costs of pegging its currency to a fixed exchange
rate area (Krugman and Obstfeld 2003, pp. 617 ff.). The outcome strongly depends on the
economic integration of the economy with the members of a monetary integration area. This
is the case since a country’s level of economic integration influences the costs and benefits. A
high level of importance of the member countries of an integration area for an economy is
associated with a high degree of economic integration of the economy.4 The potential benefits
for an economy of joining an exchange rate area might occur through gains in efficiency and
credibility. The monetary efficiency occurs from pegging to a fixed exchange rate area instead
of letting the exchange rate float since this, e.g., lowers transaction costs and inflation differ4
Beside trade, there is also the regional mobility of the production factors labor and capital of relevance for an
economy’s integration within a region because their migration supports markets to adjust to a shock.
-2-
ences. Hence, the higher the degree of economic integration of the economy is the more the
country benefits. The potential costs for an economy of joining an exchange rate area occur
through additional instability. The economic stabilizing of output and thus employment gets
more difficult for an economy by pegging to a fixed exchange rate area instead of letting the
exchange rate float – the country gives up exchange rate and monetary policy to stabilize its
economy. Exchange rate policy cannot influence relative prices of domestic and foreign products and monetary policy is not able to effect domestic output anymore to adjust to a product
demand shock. Thus, it costs fewer for the economy the higher the degree of economic integration is because the economy and the member countries of the integration area are supposed
to respond similarly to shocks.
When should an economy join the monetary integration? Figure 1 brings the described relationships together in a diagram and consequently offers a framework to answer the question.
Figure 1: Cost-benefit analysis of a monetary integration
Costs and benefits
B
0
d0
C
Degree of economic integration
The figure’s horizontal axis measures the economic integration of an economy with other
economies of a region. The vertical axis measures benefits of the monetary efficiency and
costs of the economic stability loss for the economy deciding to join a fixed exchange rate
area. Benefits as well as the costs increase from zero in the diagram’s origin. Schedule B
shows the relation between the degree of economic integration of an economy and the benefits from joining the area. B has a positive slope since an economy’s benefits of monetary
efficiency by joining a fixed exchange rate area raise as its economic integration with that
area increases. Schedule C shows the relation between the degree of economic integration and
the costs – C shows a negative slope. An economy’s costs of economic instability by joining a
fixed exchange rate area decrease with a higher level of its economic integration with the area. Figure 1 shows that the minimum degree of economic integration is d 0 being determined
by the intersection of B and C in point 0. Thus, when an economy shows a degree of econom-3-
ic integration equal to d0 it is inferior with its decision. With a level higher (lower) than d0 the
country should (not) join them. In that case the potential benefits are (not) high enough to
outperform the potential costs for an economy of joining the fixed exchange rate area.
The presentation of the cost-benefit analysis of a monetary integration revealed a high relevance of economic integration measures for the framework’s foundation. This model of the
theory of optimum currency areas shows that the costs and benefits of pegging the currency to
a fixed exchange rate area depend on the importance of an economy’s trading partners within
the region. The next section questions the indication of this importance by established economic integration models because the operationalization influences the development of economic integration measures and thus the value of a country’s degree of economic integration.
3. Measurement of economic integration with the VEI model
The section introduces the VEI model serving as an alternative to the GEI model for measuring economic integration. How can the performance of the analysis of monetary integration be
increased? This question builds the motivation for the VEI model. The answer could be an
improvement in the correctness of the degree of economic integration because of its influence
on the results of the cost-benefit analysis. A closer connection between theoretical foundation
and empirical measurement of the economic integration degree can achieve the enhancement.
Different ways of indicating the importance of trading partners within a region for an economy exist since operationalization depends on the interpretation of economic significance. In
general, a high relevance is associated with a high degree of economic integration.
Established economic integration models, like the GEI model, interpret the importance of
regional trading partners by putting their focus on the production output for trade between an
economy and member countries of an integration area. The GEI model offers two economic
integration measures – the regional export ratio (RER) at the export side of an economy and
the regional import ratio (RIR) at the import side of it. The RER puts the value of goods and
services, sold by the country to the trading partners within a region, into relation to all products the economy produced for final demand within a year. On the other hand, the RIR shows
the value of the economy’s regional imports as share of the gross domestic product (GDP).
This output-orientation of the GEI model leads to economic integration measures lacking of
accuracy. When, e.g., the RER shows a value more than 100 percent then this correctly indicates a high degree of economic integration. But how good is the theoretical foundation of the
empirical measures since this implies a negative value of domestic non-tradables.
-4-
The new VEI model interprets the importance of an economy’s trading partners within a region in another way.5 It concentrates on that part of regional trade-induced value added that
corresponds to the production factors’ income in the producer country. Hence, the VEI model
takes not the total value of regional trade into account since the model’s interpretation shifts
from an output-oriented to an input-oriented one. The indicators of the VEI model are the
regional value-added based export ratio (RVER) and the regional value-added based import
ratio (RVIR). The RVER shows the relation of domestic value added, induced by regional
exports of the home country, to the GDP. Similarly, the RVIR compares the regional value
added, induced by regional imports of the home country, with the GDP. 6 These economic
integration measures incorporate a closer link to their theoretical foundation than those of the
GEI model. E.g., the RVER cannot exceed 100 percent. It is not possible for the home country to use more than all of its production factors to produce goods and services for its regional
trading partners in export sectors and their supplying economic sectors.7
The following subsections show the theoretical foundation of the empirical measures. The
first one presents the modeling of economic interconnections within countries and between
them. Then, the second subsection describes the regional distribution of trade-induced value
added. The third one applies these instruments to develop economic integration measures.
3.1 Representing economic interconnections
This subsection introduces the VEI model’s modeling of economic interdependencies within
an economy, its regional trading partners, and the countries outside the region. In addition, it
shows the economic interconnections between the economies. The input-output table’s values
build the framework for calculating the economic integration degree of an economy. In the
following, the subsection describes the modeling of these interconnections, it shows the assumptions of the VEI model, and then it illustrates the relationships with equations.
How does the VEI model represent the national, regional, and extra-regional economic linkages in value terms? It shows the output of economic sectors being the delivery of intermediThe foundation of the VEI model’s economic integration measures is a static multiregional input-output table,
describing national and international economic interconnections by their values, and an input-output analysis. In
contrast to this, indicators of the GEI model are based on the national income account and the current account.
5
6
The regional imports consist of direct imports of the home country from the integration area as well as indirect
imports of the home country from the rest of the world importing intermediates from the integration area.
7
But, non-tradeables exist in an economy and thus not all production factors can be moved to produce regional
exports. Hence, the degree of economic integration based on the RVER is in any case lower than 100 percent.
-5-
ate products to domestic sectors as well as to foreign sectors and the supply of goods and services to domestic and foreign final demand. The foreign sectors and the components of foreign final demand are located in economies within a region or outside that region. On the other hand, economic sectors need input to produce their output. Thus, the VEI model presents
these sectors’ obtainment of intermediates from economic sectors at home and abroad. The
imported intermediate inputs are split up by the trading partners’ location within an integration area or as part of the rest of the world. Beside these domestic and imported intermediate
products sectors also require domestic production factors for their production of output.8
Which assumptions are made for modeling the connections between production output and its
input? It is supposed that every sector produces a homogenous product by using a homogenous technology. Hence, the necessity to distinguish between products and economic sectors
is omitted. Furthermore, a proportional relation between total production of a sector and its
necessary intermediate products is assumed. Returns to scale are presumed as constant in the
production. I.e., production coefficients are supposed to be independent from the factor input.
An exogenously given final demand is assumed besides being necessary for the determination
of the economic sector’s total production. Finally, it is presupposed that a given production of
a sector is only achievable by a combination of production factors. Thus, no possibilities of
factor substitution exist. An efficient input of factors is only achievable if all sectors produce
the amount of intermediates being required for the total production of the sector.
Which equations build the input-output table of the VEI model? The illustration of the modeling of linkages within an economy, its aggregated trading partners within an integration area,
those outside the region and between them begins with the output of sectors. The value of the
gross output of sector i of region k (Xik) shows the value of intermediate products of sector i
of region k for all sectors j or region k (Xijkk) and the value of goods and services of sector i of
region k for all components e of final demand of region k, including exports, (Yiekk) as
(1)
4
3
j 1
e 1
X ik   X ijkk   Yiekk , i  1,2,3,4, k  1,2,3.
Region k consists of home country (1), aggregated integration area (2), or aggregated rest of
the world (3). The aggregated region represents all regional trading partners of the home
country and the aggregated rest of the world includes those economies outside the region.
8
The depicted economic linkages show that the VEI model models not only an economy but also its regional and
extra-regional trading partners. In so doing, the representation of economic interdependencies expands the view
of Leontief (1936).
-6-
Sector i and sector j symbolize agriculture (1), other primary production (2), manufacturing
(3), or services (4). Demand e is that one in the home country (1), in the aggregated integration area (2), or in the aggregated rest of the world (3). In addition, economic sectors need
input to produce output. The value of the gross output of sector j of region k (Xjk) consists of
the value of delivered domestic intermediate products (Xijkk), the value of imported intermediate products of all sectors i of region l for sector j of region k (Xijlk), and the value of domestic
production factors of all factors g of sector j of region k (Wgjk) as
4
(2)
4
5
X jk   Xijkk   Xijlk   Wgjk ,
i 1
i 1 lk
j  1,2,3,4, k  1,2,3
g 1
where region l consists of home country (1), aggregated integration area (2), or aggregated
rest of the world (3). Production factor g is unskilled labor (1), skilled labor (2), capital (3),
land (4), or natural resources (5). Thus, the value of gross output in equation (1) equals that
one in equation (2) because production output is of the same value as its input
(3)
X ik  X jk , i, j  1,2,3,4, k  1,2,3.
This relation leads to an additional presentation of the link between gross output and demand
as given in (1). The direct production coefficient of region k (aijk) gets introduced as
(4)
a ijk 
X ijkk
X jk
, i, j  1,2,3,4, k  1,2,3
which shows the value of required intermediate products of sector i of region k for sector j of
region k to produce one unit output of sector j of region k. Hence, (1) can be transformed into
(5)
4
3
j 1
e 1
X ik   a ijk X jk   Yiekk , i  1,2,3,4, k  1,2,3.
Finally, the gross domestic product of region k (Yk) coincides with the value of primary inputs
of region k (Xijlk and Wgjk) as well as the value of domestic final demand of region k (Yiekk) as
4
(6)
4
5
4
4
3
Yk   X ijlk   Wgjk   Yiekk , k  1,2,3.
i 1 j 1 lk
g 1 j 1
i 1 e 1
Equations (1) to (6) represent the economic linkages within an economy, within its aggregated
trading partners inside and outside an integration area, and between them. The next subsection
analysis these interconnections.
-7-
3.2 Modeling the regional distribution of trade-induced value added
The subsection describes how the VEI model distributes parts of the value added created by a
country’s international trade. This input-output analysis and the previous input-output table
build the foundation of economic integration measures. In the following, the subsection explains the export-induced value added and then it presents its regional distribution.
What is the role of the export-induced value added in the VEI model? Assume that the home
country’s export sectors sell goods and services to member countries of an integration area. 9
According to equations (2) and (5) these exports need for their production intermediate inputs
from domestic economic sectors, imported intermediates from sectors inside and outside the
integration area, and production factors of the home country. Thus, exports generate income
for production factors which equals the exports’ value – the export-induced value added. The
distribution of this income is determined by the production structures of export sectors and
their supplying sectors. These production structures reflect the international competitive position of the sectors and thus the degree of the home country’s participation in the international
division of labor. In this concept of the VEI model the export-induced domestic value added
represents the value of required production factors in the home country. Whereas, the exportinduced international value added characterizes the home country’s demand of imported intermediate products from the region aggregated integration area or aggregated rest of the
world. Thus, the imported intermediate inputs do not create income in the home country.10
How much income do exports of the producer country create at home? 11 The answer begins
with a presentation of the gross output of equation (5) in a compact way. Therefore, the vector
of values of gross output of region k (xk) is
(7)
x k  X1k , X 2k , X 3k , X 4k  , k  1,2,3.
T
Then, the vector of final demand values of region k (yk) gets defined as
T
(8)
3
3
3
 3

y k    Y1ekk ,  Y2ekk , Y3ekk , Y4ekk  , k  1,2,3
e 1
e 1
e 1
 e1

which is followed by the matrix of direct production coefficients of region k (Ak)
9
This view can be applied analogously to the aggregated integration area and aggregated rest of the world.
10
The VEI model does not take the redistribution of exported intermediates of the home country into account
because of the low magnitude of exported intermediate inputs of the home country being part of the economy’s
export-induced imported intermediate products.
11
This is the export-induced domestic value added of region k.
-8-
(9)
 a 11k

a
A k  a ijk    21k
a
 31k
a
 41k
a 12k
a 22k
a 32k
a 42k
a 13k
a 23k
a 33k
a 43k
a 14k 

a 24k 
, k  1,2,3.
a 34k 

a 44k 
Now, the gross output of equation (5) can be rewritten as
(10)
x k  A k x k  y k , k  1,2,3.
The next intermediate step links the demanded exports with the required gross output of region k (xk). It begins with the vector of export values of region k (yk) defined as
(11)
y k  Y1lkk , Y2lkk , Y3lkk , Y4lkk  , k  1,2,3, l  k. 12
T
The identity matrix (B) is
(12)
1

0
B  b rs   
0

0

0
1
0
0
0
0
1
0
0

0
1 für r  s
, b rs  

0
0 für r  s


1
which allows to rearrange equation (10) to
(13)
B A k  x k
 y k , k  1,2,3.
Consequently, the gross output of region k xk, required to supply the exports of region k yk, is
(14)
x k  B A k  y k , k  1,2,3.
1
(B–A1)–1 is the Leontief inverse matrix of region k and its coefficients indicate the expenditure of sector i of region k for the production of one unit final demand of sector j of region k.
It follows the last step connecting the gross output of region k (xk) with the income of production factors in region k. The production coefficient of production factors (dgjk) gets introduced
(15)
d gjk 
Wgjk
X jk
, g  1,2,,5,
j  1,2,3,4, k  1,2,3
showing the value of factor g necessary for the production of one unit output of sector j of
region k. Hence, the matrix of production coefficients of production factors of region k (Dk) is
Depending on the analysis’ focus either economies in one of the regions or all foreign countries, demanding
exports, are taken into account.
12
-9-
(16)
 d11k

 d 21k
D k  d gjk    d 31k

 d 41k
d
 51k
d 12k
d 13k
d 22k
d 32k
d 23k
d 33k
d 42k
d 43k
d 52k
d 53k
d14k 

d 24k 
d 34k , k  1,2,3.

d 44k 
d 54k 
This leads to the vector of values of production factors of region k (qk) which is defined as
(17)
q k  Q1k , Q 2k , Q3k , Q 4k , Q5k  , k  1,2,3.
T
It estimates the values of production factors of region k qk for the gross output of region k xk
required to supply the demanded export products of region k (yk)
(18)
q k  D k x k , k  1,2,3.
qk stands for the export-induced domestic value added of region k.
What value of imported intermediates does the producer country create with its exports?13
The linking between the gross output of region k xk and the value of imported intermediates
from region l begins with the production coefficient of imported intermediate products (cijlk)
(19)
c ijlk 
X ijlk
X jk
, i, j  1,2,3,4, k  1,2,3, l  k .
cijlk shows the value of intermediate products of sector i of region k, required to be imported
from region l, for the production of one unit output of sector j of region k. Thus, the matrix of
production coefficients of imported intermediate products of region k from region l (Clk) is
(20)
 c11lk

c
C lk  c ijlk    21lk
c
 31lk
c
 41lk
c12lk
c 22lk
c 32lk
c 42lk
c13lk
c 23lk
c 33lk
c 43lk
c14lk 

c 24lk 
, k  1,2,3, l  k
c 34lk 

c 44lk 
leading to the vector of values of imported intermediate products of region k from region l
(plk) where
(21)
p lk  P1lk , P2lk , P3lk , P4lk  , k  1,2,3, l  k
T
shows the values of the required imported intermediates of region k from region l plk for the
gross output of region k xk being essential to produce the export products of region k yk
13
This is the value added which exports of the producer country create abroad in region l represented by the
export-induced international value added of region k in region l.
-10-
(22)
p lk  Clk x k , k  1,2,3, l  k.
plk symbolizes the export-induced international value added of region k in region l.
The presented concepts of export-induced domestic value added and export-induced international value added complete the framework for the following economic integration measures.
3.3 Calculating economic integration measures
The subsection presents the economic integration measures of the VEI model. These indicators show the importance of a country’s trading partners within an integration area with the
degree of economic integration. In the following, the subsection applies the instruments of the
previous subsections to develop the economic integration measures RVER and RVIR.
How does the VEI model calculate the degree of economic integration? The RVER measure
shows the export-induced domestic value added of exports to the integration area (q1) as share
of the GDP (Y1) in percent as
y1  Y1211, Y2211, Y3211, Y4211  , x 1  B A1  y1 , q1  D1 x 1 ,
1
T
(23)
RVER 
q1
100.
Y1
In addition, the indicator RVIR puts the export-induced regional value added (q2 and p23) in
relation to the GDP (Y1) in percent as
y 2  Y1122 , Y2122 , Y3122 , Y4122  , x 2  B A 2  y 2 , q 2  D 2 x 2 ,
1
T
(24)
y 3  Y1133 , Y2133 , Y3133 , Y4133  , x 3  B A 3  y 3 , p 23  C 23 x 3 ,
1
T
RVIR 
p
q2
100  23 100.
Y1
Y1
The export-induced regional value added consists of the income created in the integration area
by international trade with the home country. q2 represents the export-induced domestic value
added of the region aggregated integration area of exports to the region home country and p23
symbolizes the export-induced international value added of the region aggregated rest of the
world in the region aggregated integration area.
After the introduction of the VEI model, showing the different foundation of its economic
integration measures in contrast to those of the GEI model, the next section empirically assesses the indicators within the framework of monetary integration analysis.
-11-
4. Impact of the VEI model on the analysis of monetary integration
The section presents an empirical comparative analysis of the VEI model. Regular variations
of the calculated degrees of economic integration are very likely to occur because of the different theoretical backgrounds of the GEI and VEI model’s economic integration measures. It
is reasonable to suppose that their importance increase in line with the intensity of an economy’s participation within the regional division of labor. A country needs not only domestic
production factors but also intermediate products from abroad. Thus, the section tries to find
out if economic integration measures of the VEI model can empirically show a smaller impact
of regional trade on economic variables of the producer country. This leads to the question:
Do differences of economic integration degrees between the presented models show a relevant magnitude to influence the results of the cost-benefit analysis of monetary integration?
The following subsections try to answer the question. The first one calculates economic integration measures of the GEI and VEI model. Then, the second subsection characterizes the
cross-sectional sample consisting of member countries of the EU, NAFTA, and MERCOSUR.
The third one compares the results of the cost-benefit analysis based on the two models.
4.1 Economic integration measures of members of EU, NAFTA, and MERCOSUR
The subsection presents calculated degrees of economic integration based on the GEI and VEI
model. It is the starting point for the empirical analysis in the next subsections. The crosssectional sample includes 21 economies of the integration areas EU, NAFTA, and MERCOSUR.14 Paraguay is not contained in the latter one since data were not available. The sample in this comparative analysis is based on the GTAP Data Base Version 5 (Center for Global
Analysis 2001) described in Dimaranan and McDougall (2002).15 In the following, the subsection presents the economic integration degrees of interest and then it shows the corresponding country’s ranking of economic integration.
How well integrated are the economies of the three integration areas? Table 1 shows the sample. It presents the calculation outcomes of the RVER and the RVIR of the VEI model together with the GEI model’s economic integration measures RER and RIR.
14
In addition to the EU, NAFTA and MERCOSUR are taken into account to compare results with other regions.
15
Before this data basis was applied to calculate economic integration measures it was transformed to fit the VEI
model introduced in the previous section 3.
-12-
Table 1: Degrees of economic integration based on the VEI and GEI model
Percent of GDP,
1997
MERCOSUR
Argentina
Brazil
Paraguay
Uruguay
NAFTA
Canada
Mexico
United States
EU
Austria
Belgium
Denmark
Finland
France
Germany
Greece
Ireland
Italy
Luxembourg
Netherlands
Portugal
Spain
Sweden
United Kingdom
Export side
Import side
RVER
RER
RVIR
RIR
2.4
0.8
....
5.7
2.7
0.9
....
7.1
2.0
1.1
....
8.3
2.2
1.2
....
9.0
19.2
17.7
2.2
27.1
23.2
2.6
20.0
16.3
2.4
22.5
18.2
3.3
14.8
24.8
16.1
15.0
11.8
11.3
6.7
29.3
9.9
25.9
25.7
16.1
12.4
15.7
10.5
21.1
48.4
21.7
20.7
14.5
14.1
7.8
49.8
12.9
50.6
42.1
21.7
16.4
22.1
13.2
23.4
42.3
18.0
16.7
12.1
11.4
14.4
37.2
11.3
47.3
27.4
26.1
15.1
19.5
12.3
26.8
48.6
20.7
18.9
14.3
13.6
16.3
41.9
13.0
54.1
31.0
30.3
17.5
22.3
14.2
Source: Center for Global Trade Analysis (2003) and own calculations.
Which relative positions do the economies show according to their degrees of economic integration? The following table 2 gives this additional view on the sample. The table records the
rank order of the RVER and the RVIR as well as the RER and the RIR. These rank orders
begin with one for the country with the lowest degree of economic integration, continue with
two, three, …, and end with the total number of countries for the most integrated economy.
-13-
Table 2: Rank orders of economies by economic integration based on the VEI and GEI model
Rank order, 1997
RVER
RER
Rank order, 1997
RVIR
RIR
Export side
Brazil
United States
Argentina
Uruguay
Greece
Italy
United Kingdom
Germany
France
Spain
Finland
Austria
Denmark
Portugal
Sweden
Mexico
Canada
Netherlands
Belgium
Ireland
Luxembourg
Paraguay
1
2
3
4
5
6
7
8
9
10
12
11
15
14
13
16
17
19
18
21
20
....
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
....
Import side
Brazil
Argentina
United States
Uruguay
Italy
Germany
United Kingdom
France
Greece
Spain
Mexico
Finland
Denmark
Sweden
Canada
Austria
Portugal
Netherlands
Ireland
Belgium
Luxembourg
Paraguay
1
2
3
4
5
6
8
7
9
10
11
12
13
14
15
16
17
18
19
20
21
....
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
....
Source: Center for Global Trade Analysis (2003) and own calculations.
Based on the presented sample the next subsection attempts to reveal regular disparities between the economic integration degrees calculated by the GEI model and the VEI model.
4.2 Systematic differences between alternative economic integration measures
The subsection characterizes the economic integration models of the study to assess if systematic differences of their economic integration measures might be of relevance for the costbenefit analysis of monetary integration. In the following, the subsection gives a visual impression of the sample with an interpretation of the outcomes and then it completes the characterization with a frequency distribution, correlation, and regression analysis.
Does a visual impression of the sample show us systematic differences between the GEI and
VEI model’s indication of the importance of regional trading partners for an economy? The
RVER gets presented by contrasting it with the corresponding widespread RER. Figure 2
shows estimations of these two indicators. The horizontal axis arranges the economies of the
sample in an increasing order by their position within the rank order of the RER measure. The
vertical axis shows the empirical outcomes of the RVER and the RER, respectively.
-14-
Figure 2: Degrees of economic integration at the export side
60
Share of GDP (%)
50
40
30
20
10
0
1
3
5
7
9
11
13
15
17
19
21
Rank of RER
RVER
RER
Source: Center for Global Trade Analysis (2003) and own calculations.
Figure 2 illustrates that (1) the RVER is in all cases lower than the RER. Thus, the VEI model
leads to empirical lower degrees of economic integration in comparison to the often applied
GEI model. The RVER cannot exceed 100 percent. It is not possible to use all production
factors of an economy to produce exclusively export products since, e.g., non-tradeables exist.
In case of the GEI model, its RER measure can be larger than 100 percent because this model
does not take the structure of production input for international trade into account. It clearly
reveals (2) the tendency of the RVER to increase with the RER. This means that the more
products the economic sectors of an economy sell to their regional trading partners the more
domestic production factors they and their previous supplying economic sectors need to produce. The value of the production factors exactly corresponds to the export-induced domestic
value added. In addition, the figure shows that (3) the spread between the RVER and the RER
increases with the rank order. This spread reflects export-induced imported intermediate
products as share of the GDP. A country requires for its exports intermediates from abroad for
production. The growing spread between the RVER and the RER discloses that a more regionally open economy demands for domestic production factors at a relatively lower magnitude. Since the RVER is less steep than the RER (4) the economies show smaller differences
in their economic integration when the VEI model is applied. This implies that the importance
of regional trade becomes more similar for the countries within an integration area. Finally,
the jitter of the RVER reflects that (5) some positions of countries within the rank order
change when the degree of economic integration is estimated by the RVER instead of the
RER. Export-induced imported intermediate products do disturb the rank order with a low
magnitude since they show a tendency to increase with the degree of economic integration.
-15-
How can value-added based economic integration measures be empirically described at the
import side? Figure 3 illustrates the values of the RVIR as well as the RIR for the investigated
countries. The figure’s horizontal axis puts the economies in an increasing order of their estimated RIR values. Its vertical axis shows the realizations of the RVIR and the RIR.
Figure 3: Degrees of economic integration at the import side
60
Share of GDP (%)
50
40
30
20
10
0
1
3
5
7
9
11
13
15
17
19
21
Rank of RIR
RVIR
RIR
Source: Center for Global Trade Analysis (2003) and own calculations.
Figure 3 indicates corresponding results for the import side than for the export side but at a
distinct lower level.
Does a frequency distribution analysis support our results? The following table 3 gives a
summary of the sample in table 1.
Table 3: Results of the frequency distribution analysis
Sample 1 21
Observations 21
Mean
Median
Maximum
Minimum
Range
Standard deviation
Variation coefficient
Skewness
Kurtosis
Jarque-Bera
Probability
Export side
RVER
RER
14.01
20.98
14.78
20.67
29.28
50.59
0.84
0.94
28.44
49.65
8.05
15.21
0.57
0.72
0.17
0.77
2.29
2.64
0.54
2.19
0.7648
0.3340
Import side
RVIR
RIR
18.31
20.95
16.25
18.24
47.33
54.12
1.08
1.19
46.25
52.93
12.38
14.11
0.68
0.67
0.83
0.83
3.17
3.20
2.42
2.47
0.2980
0.2907
Source: Center for Global Trade Analysis (2003) and own calculations.
The common statistical measures include also the Jarque-Bera test testing the distribution for
normality (Jarque and Bera 1987). A small probability value leads to a rejection of the null
-16-
hypothesis that the distribution is a normal distribution. Table 3 confirms the empirical outcomes of figure 2 and figure 3.
What do results of a correlation analysis show? Table 4 summarizes the sample in table 2.
The correlation analysis is applied to characterize the different rank orders of economies,
which are sorted by their economic integration degrees, according to the RVER and the RER
as well as the RVIR and the RIR. The analysis incorporates the rank order correlation
measures of Spearman (ρR) and Kendall (τ), respectively (Kendall and Dickinson Gibbons
1990).
Table 4: Results of the correlation analysis
Sample 1 21
Observations 21
RVER
RVIR
RER
RIR
0.990909 (ρR)
0.952381 (τ)
/
/
/
/
0.998701 (ρR)
0.990476 (τ)
Source: Center for Global Trade Analysis (2003) and own calculations.
The measures ρR and τ also show that the positions of the economies within the rank order do
scarcely change when the VEI model is applied instead of the GEI model.
What additional insights between the relationship of regional trade and induced income can a
regression analysis offer (Greene 2002)? The specifications of the regression equations are
(25)
log RVER t  cˆ1  cˆ2 log RER t  uˆt , t  1,2,,21 and
(26)
log RVIR t  cˆ1  cˆ2 log RIR t  uˆt , t  1,2,,21
where the index t represents the economy with the number t in the sample. The estimator ĉ2 in
equation (25) measures the induced percentage change of RVERt when RERt increases for
one percent. Equation (26) is to be analogously interpreted. We apply the ordinary least
squares (OLS) method since the specifications do not disturb assumptions of functionality, the
residual, and variables. The following table 5 sums up the estimation results.
-17-
Table 5: Results of the regression analysis
Sample 1 21
Observations 21
RVER
RVIR
RER
RIR
0. 87***
/
/
1.00***
Source: Center for Global Trade Analysis (2003) and own calculations.
Note: *** 1 percent significance level
Table 5 indicates that a raise of RER of 1 percent goes in line with a 0.87 percent increase of
the RVER. This supposes that the importance of domestic production factors in relation to
imported intermediate products to produce goods and services for exports declines with the
level of an economy’s participation within the international division of labor. The estimates
for the import side show that the RVIR increases 1 percent when the RIR raises 1 percent because the RVIR deviates from the RIR at a similar rate for each country. Table 5a and table 5b
reproduce table 5 in detail to present all relevant estimation results.
-18-
Table 5a: Regression of value-added based economic integration at the export side
Dependent Variable
Method
Variable
C
LOG(RER)
R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
Durbin-Watson stat
LOG(RVER)
Least Squares
Coefficient
0.033038
0.872078
0.989690
0.989148
0.094272
0.168858
20.84609
0.965361
Sample
Included observations
Std. Error
t-Statistic
0.058531
0.564452
0.020420
42.70726
Mean dependent var
S.D. dependent var
Akaike info criterion
Schwarz criterion
F-statistic
Prob(F-statistic)
1 21
21
Prob.
0.5791
0.0000
2.373258
0.904943
-1.794866
-1.695387
1823.910
0.000000
Source: Center for Global Trade Analysis (2003) and own calculations.
Table 5b: Regression of value-added based economic integration at the import side
Dependent Variable
Method
Variable
C
LOG(RIR)
R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
Durbin-Watson stat
LOG(RVIR)
Least Squares
Coefficient
-0.150631
1.004423
0.998254
0.998162
0.041622
0.032916
38.01473
2.530844
Sample
Included observations
Std. Error
t-Statistic
0.027835
-5.411636
0.009637
104.2224
Mean dependent var
S.D. dependent var
Akaike info criterion
Schwarz criterion
F-statistic
Prob(F-statistic)
1 21
21
Prob.
0.0000
0.0000
2.591568
0.970846
-3.429974
-3.330496
10862.31
0.000000
Source: Center for Global Trade Analysis (2003) and own calculations.
After this subsection empirically characterized the economic integration measures of the GEI
and VEI model the following subsection investigates the relevance of the findings for the
cost-benefit analysis of monetary integration of the theory of optimum currency areas.
4.3 Influence of economic integration models on the cost-benefit analysis
The subsection assesses effects of a change in the theoretical foundation of the economic integration degree on the results of the monetary integration analysis. A shift from the established GEI model towards the new VEI model might lead to a deviated valuation of pegging
an economy’s currency to a fixed exchange rate area. This might be the case because the VEI
model shows a closer linkage between theoretical foundation and empirical measurement of
the degree of economic integration than the GEI model. In the following, the subsection illustrates the significance of the discovered systematic differences of the degrees of economic
integration for the judgment of monetary integration. Then it presents the potential influence
-19-
of the changed foundation on the analysis results within the cost-benefit framework. Finally,
the subsection compares the outcomes of estimated degrees of economic integration for the
EU and the two other investigated integration areas NAFTA and MERCOSUR.
How relevant are the divergences of economic integration degrees for the analysis of monetary integration? The degree of economic integration shows independent from its theoretical
foundation a high relevance within the cost-benefit framework for answering the question
whether an economy should join a monetary integration process (see section 2). A change of
the economic integration model from the GEI model towards the VEI model leads to a systematic difference of the economic integration degree’s value. The potential candidate and the
member countries of an integration area are less integrated but at a more similar level (see
subsection 4.2). Table 6 reviews the deviations of economic integration degrees. It characterizes the investigated integration areas by the measures mean, maximum, and minimum.
Table 6: Effects of the VEI model on the degree of economic integration
Sample 1 21
Observations 21
Mean
Maximum
Minimum
RVER as share of
RER in percent
74.19
90.12
51.22
RVIR as share of
RIR in percent
87.13
91.32
74.63
Source: Center for Global Trade Analysis (2003) and own calculations.
The VEI model estimates at the export (import) side of an economy a degree of economic
integration that is on average 26 percent (13 percent) lower than the value calculated by the
GEI model. Thus, the magnitude of different values of the economic integration degree due to
its changed economic integration model might be relevant for the cost-benefit analysis because of its potential impact on the analysis’ results.
What does the closer theoretical linkage of empirical economic integration measures of the
VEI model mean for the cost-benefit analysis? In the following we pick up again the framework of section 2 to reflect the effects on the analysis’ outcomes.16 According to the VEI
model, the present members of a fixed exchange rate area and the possible participant are less
economic integrated with each other than the GEI model suggests. Thus, the candidate’s value
of the realized economic integration degree is lower. The economic integration measures of
the VEI model indicate the significance of an economy’s trading partners within an integration area by focusing on the regional trade-induced value added in the probable participant as
16
The following diagrams use the same cost-benefit framework as in figure 1. For a description of their construction refer to section 2.
-20-
well as aggregated member countries. This includes in addition that the VEI model does not
take into account the trade with the rest of the world as the GEI model partly does. The established GEI model is not able to distinguish between regional trade-induced imported intermediate products delivered from trading partners within the integration area and from those suppliers outside the region. Hence, the GEI model overestimates the regional economic integration because it includes these extra-regional intermediates to indicate the importance of member countries of an integration area.17 Furthermore, this implies that the GEI model indicates a
too high impact of the regional integration on economic variables of the economies within a
region. Figure 4 shows the impact of a shift in the theoretical foundation of the concrete degree of economic integration for an economy deciding to join the monetary integration.
Costs and benefits
Figure 4: Influence of the VEI model on the realized degree of economic integration
d'1
d1
Degree of economic integration
It illustrates the move of the actual economic integration degree from d1 to d'1 when the VEI
model is applied instead of the GEI model for measuring regional economies’ significance.
In addition to the lower concrete degree of economic integration, an application of the VEI
model as a replacement for the GEI model might lead to a higher minimum degree of economic integration – the level at which benefits outperform costs. An economy’s benefits of
monetary efficiency gains by linking its currency to the fixed exchange rate area are lower as
the GEI model suggests because regional trade-induced intermediates from outside the integration area do not make a contribution to the monetary gains. In contrast to this, the costs of
economic instability losses by joining the area are higher when the VEI model is used. The
adjustment of distortions of asymmetric shocks is for the country more painful because the
regional trade incorporates intermediate inputs from trading partners outside the region. The
following diagram in figure 5 gives a summary of these results.
17
This implies that an extended version of the GEI model, modeling in addition the extra-regional trade, would
undervalue the significance of economies outside the region.
-21-
Figure 5: Impact of the VEI model on the minimum economic integration degree
Costs and benefits
B
B'
0
0'
C'
d0
d'0
C
Degree of economic integration
Figure 5 picks up the country’s critical degree of economic integration d 0 of figure 1. It is
assumed that d0 is derived by the established GEI model. An application of the VEI model
shifts d0 to d'0 because the intersection 0 of schedule B and C moves to 0' of B' and C'.
A reassessment of an economy’s decision of pegging its currency to a fixed exchange rate
area might be necessary when the VEI model is applied instead of the GEI model to measure
economic integration. The following figure 6 illustrates this result.
Figure 6: Influence of the VEI model on the judgment of joining an exchange rate area
Costs and benefits
Costs and benefits
B
1
0
B'
2'
1'
0'
C'
2
d0
d1
C
d'1
Degree of economic integration
d'0
Degree of economic integration
(a) GEI model based foundation
(b) VEI model based foundation
Figure 6a shows that according to the GEI model an economy’s actual degree of economic
integration d1 is higher then its minimum economic integration degree represented by d0.
Since the benefits of joining the fixed exchange rate area in point 1 outweigh the costs in
point 2 the result of this cost-benefit analysis of monetary integration is a recommendation for
the economy to link its currency to the area’s currency. Figure 6b draws another conclusion
for the same economy facing an unchanged economic environment. A change in the economic
integration model and thus in the foundation of the degree of economic integration by applying the VEI model leads to an opposite advice than before with the well established GEI mod-22-
el. In this case, the realized degree of economic integration d'1 is lower than the critical degree
of economic integration d'0. Hence, the benefits in point 1' are less than the costs in point 2'
and the economy should not join the monetary integration process of the region. Thereupon,
outcomes of the cost-benefit analysis of monetary integration based on the VEI model might
deviate from those analysis results backed up by the GEI model.
Are the differences of the calculated economic integration degrees between the economic integration models for the EU, NAFTA, and MERCOSUR significant to influence the results of
the integration areas’ cost-benefit analysis of monetary integration? In the following, this part
investigates the actual degrees of economic integration based on the VEI model and compares
their values with those realizations calculated by the GEI model. An answer to the given question is very questionable when only concrete economic integration degrees are taking into
account but not the corresponding minimum degrees of economic integration. The critical
levels are in addition necessary to assess the influence of the VEI model on the results of the
cost-benefit analysis for a possible member country of an integration area.18 Only the minimum economic integration degree can show whether in the concrete economic situation of a
potential candidate benefits of joining the fixed exchange rate area surpass costs. Nevertheless, a closer look at the deviations of actual economic integration degrees should give some
preliminary insights. When a country shows a high concrete economic integration degree,
according to the GEI model, but after the application of the VEI model it faces a small level
of integration then this might indicate a reassessment of the advice for the economy to peg its
currency to a fixed exchange rate area. The following table 7 presents the results of a comparative analysis of the study’s economic integration models.19
18
The estimation of critical degrees of economic integration is not part of this study.
19
Table 7 represents the sample of the integration areas in table 1. It disaggregates the results of table 3 and table
6 by the investigated regions.
-23-
Table 7: Effects of the VEI model on the concrete degree of economic integration
Export side
Sample 1 21
Observations 21
EU
Mean
Maximum
Minimum
Range
Standard deviation
NAFTA
Mean
Maximum
Minimum
Range
Standard deviation
MERCOSUR
Mean
Maximum
Minimum
Range
Standard deviation
Import side
RVER
RER
RVER as
share of RER
RVIR
RIR
RVIR as
share of RIR
16.40
29.28
6.66
22.62
6.82
25.13
50.59
7.75
42.83
14.78
70.97
85.95
51.22
34.72
10.72
22.29
47.33
11.31
36.02
11.67
25.56
54.12
12.99
41.13
13.25
86.98
88.65
84.21
4.44
1.30
13.06
19.25
2.23
17.02
9.42
17.64
27.10
2.58
24.52
13.18
77.91
86.46
71.03
15.42
7.84
12.91
20.03
2.43
17.60
9.27
14.68
22.55
3.26
19.29
10.12
84.19
89.10
74.63
14.46
8.28
2.97
5.66
0.84
4.82
2.46
3.56
7.07
0.94
6.12
3.16
86.55
90.12
80.09
10.03
5.61
3.79
8.26
1.08
7.18
3.90
4.16
9.04
1.19
7.85
4.27
90.82
91.32
90.13
1.19
0.62
Source: Center for Global Trade Analysis (2003) and own calculations.
The table compares the concrete degrees of economic integration based on the VEI model
with the values of the GEI model for the different regions by statistical measures. The deviations of the concrete economic integration degrees seem to be not significant to influence the
results of the cost-benefit analysis of monetary integration. Either the net benefits of participants are lower as before or costs already outperform benefits for countries decided not to
migrate towards a monetary integration process. Thus, the sample indicates no revision of the
ongoing monetary integrations with the limitation that no critical degrees of economic integration are taken into the account of this analysis.
After the investigation of the VEI model’s influence on the results of the cost-benefit analysis
of monetary integration is presented the next section draws conclusions and gives some implications for the analysis of monetary integration.
5. Conclusions
The previous sections illustrated how the new VEI model improves the performance of the
cost-benefit analysis of monetary integration compared to the established GEI model. A candidate for joining a fixed exchange rate area is less economic integrated with the member
-24-
countries of the region and net benefits of a participation are lower.20 Furthermore, the countries within the integration area show a more similar level of integration between each other.
Thus, the theoretical foundation of the degree of economic integration matters for the analysis’ outcomes. For a closer link of these results to economic policy advices further research is
necessary. Anyhow, already this very early stage of research indicates that it might be reasonable to think about changing the perspective on indicating the importance of regional trading
partners by the degree of economic integration from an output-oriented towards an inputoriented theoretical core.
How could further work advance this study? An estimation of a candidate’s costs and benefits
of joining to a fixed exchange rate area would be of interest. This would allow us to calculate
the critical degree of economic integration. Its comparison with the actual level of integration
would give an advice whether the country should participate. Also, a comparison between the
significance of trading partners inside and those outside a region could reveal additional insights about the regional integration. This should also lead to an additional knowledge about
the regional distribution of the trade-induced value added and thus the interpretation of the
intra- and extra-regional economy’s importance. Furthermore, the presented model of the theory of optimum currency areas does not use all possibilities which the VEI model offers. A
slightly enhanced version of the economic integration model shows beside the already introduced economic integration measures and the mentioned extra-regional indicators also structures of the international trade. These are the traded products, the demanding sources, and the
incorporated production factors.21 Finally, a sample, including more integration areas as well
as additional years, should enrich the work.
20
This includes the fact that net benefits can show a more negative value.
21
Exports get delivered to final demand as well as economic sectors, using intermediates to produce products for
the own country or economies abroad.
-25-
References
Center for Global Trade Analysis (2003), GTAP Data Base Version 5.4, Purdue University,
West Lafayette, IN.
Dimaranan, B.V./McDougall, R.A. (2002), Global Trade, Assistance, and Production: The
GTAP 5 Data Base, West Lafayette, IN.
Greene, W.H. (2002), Econometric Analysis, 5th Ed., Upper Saddle River, NJ.
Gros, D./Thygesen, N. (1998), European Monetary Integration, 2nd Ed., Harlow.
Jarque, C.M./Bera, A.K. (1987), A Test for Normality of Observations and Regression
Residuals, in: International Statistical Review, Vol. 55, pp. 163-172.
Kendall, M./Dickinson Gibbons, J. (1990), Rank Correlation Methods, 5th Ed., Oxford.
Krugman, P.R./Obstfeld, M. (2003), International Economics: Theory and Policy, 6th Ed.,
Reading, MA.
Leontief, W.W. (1936), Quantitative Input and Output Relations in the Economic System of
the United States, in: Review of Economics and Statistics, Vol. 18, No. 3, pp. 105-125.
Mundell, R.A. (1961), A Theory of Optimum Currency Areas, in: American Economic
Review, Vol. 51, No. 4, pp. 657-665.
Wang, L. (2003), How Important is International Trade for a Country Really?: A ValueAdded Based Approach to Measure Economic Openness, paper for the conference “VIII
Conference on International Economics”, Spanish Chapter of the International
Economics and Finance Society (AEEFI) and University of Castilla-La Mancha
(UCLM), Ciudad Real, June 25-27.
-26-