University of Southern California Dissertations and Theses

Empirical essays on trade liberalization and export diversification

(USC Thesis Other)

Empirical essays on trade liberalization and export diversification

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content EMPIRICAL ESSAYS ON TRADE LIBERALIZATION AND EXPORT
DIVERSIFICATION
by
Bahar Kartalciklar
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(ECONOMICS)
DECEMBER 2016
Copyright 2016 Bahar Kartalciklar
To my beloved, relentless mother.
In memory of my father.
ii
Acknowledgments
First and foremost, I want to express my deepest gratitude to my advisor, Robert
Dekle, for his endless support and guidance. I have not only benefitted from
his broad knowledge in international economics, but also learned how to be a
researcher from him. His tremendous encouragement, patience and understand-
ing was instrumental in completion of this dissertation.
I am also extremely grateful to Caroline Betts for her stimulating discussion,
which provided the main inspiration behind this work. Her insightful comments
and challenging questions were exceptionally helpful in improving the quality of
my work. I also want to extend my special gratitude to her for her support and
kind words during tough times.
I would like to thank Aiichiro Nakano for his valuable time and accepting
to be a member of my dissertation committee as well as to Jeffrey Nugent, Guil-
laume Vandenbroucke and Aris Protopapadakis who took part in the qualifying
committee. I am also indebted to Joshua Aizenman, who kindly provided his
support and advice during my job search process.
I cannot thank department staff members Morgan Ponder and Young Miller
at the Department of Economics enough for their help in maneuvering various
iii
administrative hurdles during my stay at USC. They made sure bureaucratic con-
cerns do not impede my academic progress. I am also grateful for the funding
provided by USC’s Dornsife College of Letters and Science, Department of Eco-
nomics, and the Graduate School.
I am forever grateful to Haluk Levent, my econometrics professor and men-
tor during my undergraduate years at Galatasaray University. He introduced me
to the exciting world of research. His encouragement and support were crucial
in my pursuit of higher education.
I would also like to thank all my friends, who cherished me during my Ph.D.
journey. In particular, I want to thank Burcin Isguden and Berna Basak from the
bottom of my heart, for always remaining close and being true friends despite
the physical distance.
Last, but definitely not least, I would like to extend my heartfelt gratitude
to my family, for their unconditional love and tremendous support during this
adventure. My sister Seda Kartalciklar, for the endless late night conversations;
my mother, Zuhal Kartalciklar, from whom I learned the meaning of hard work,
dedication and believing in someone; and my dearest father, Nuri Kartalciklar,
who taught me the true value of sacrifice and discipline. They have sacrificed so
much in order for me to be where I am now. None of my achievements would
have been possible if it were not for them.
iv
Table of Contents
Dedication ii
Acknowledgments iii
List of Tables vii
List of Figures ix
Abstract x
Chapter 1: Export Diversification and Output Volatility – Empirical Evi-
dence from Central and Eastern European Countries 1
1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Trade Openness and Export Diversification . . . . . . . . . . . . . . 7
1.3 Related Literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
1.4 Data and Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . 23
1.5 Empirical Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
1.5.1 Aggregate Output Volatility and Extensive Margin of Exports 34
1.5.2 Sectoral Output Volatility and Extensive Margin of Exports 40
1.5.3 European Union Agreement and the Impact of Export Diver-
sification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
1.6 Robustness Checks . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
1.7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
Chapter 2: The Role of Trade Liberalization on the Export Variety of Cen-
tral and Eastern European Countries 58
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
2.2 Related Literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
2.3 Data and Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . 65
2.4 Total Export Growth and the Extensive Margin . . . . . . . . . . . . 70
2.4.1 Unweighted Measures of Export Variety . . . . . . . . . . . 70
2.4.2 Weighted Measures of Export Variety . . . . . . . . . . . . . 85
v
2.5 Regression Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
2.5.1 Specification . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
2.5.2 Estimation Results . . . . . . . . . . . . . . . . . . . . . . . . 96
2.6 Decomposition of the Relative Export Growth . . . . . . . . . . . . 99
2.7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
Bibliography 108
Chapter A: Appendix to Chapter 1 114
Chapter B: Appendix to Chapter 2 119
vi
List of Tables
1.1 Descriptive statistics: Export diversification, variety: product-market
pair . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.2 Change in the average export diversification (in %), between pre-
(1995-2003) and post-EU (2004-2012) periods, variety: product-market
pair . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.3 Change in the average openness and volatility of the economic
indicators (in %), between pre- (1995-2003) and post-EU (2004-
2012) periods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.4 Descriptive statistics: Sectoral export diversification, variety: product-
market pair . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
1.5 Descriptive statistics: Selected variables, by economic activity, before
(1995-2003) and after (2004-2012) EU . . . . . . . . . . . . . . . . . . 13
1.6 Average change in the export diversification by economic activity
(in %), between pre- (1995-2003) and post-EU (2004-2012) periods,
variety: product-market pair . . . . . . . . . . . . . . . . . . . . . . . . 14
1.7 Change in the selected variables (in %), in the sectors with the
largest decline in output volatility, between pre- (1995-2003) and
post-EU (2004-2012) periods . . . . . . . . . . . . . . . . . . . . . . . 15
1.8 Regression results for per-capita GDP volatility on export diversifi-
cation - variety (product-market pair) . . . . . . . . . . . . . . . . . 35
1.9 Regression results for per-capita GDP Volatility on product and mar-
ket diversification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
1.10 Regression results for sectoral output volatility on sectoral export
diversification - variety (product-market pair) . . . . . . . . . . . . 41
1.11 Regression results for per-capita GDP volatility on export diversifi-
cation - variety (product-market pair) and EU accession . . . . . . 44
1.12 Regression results for sectoral output volatility on sectoral export
diversification - variety (product-market pair) and EU accession . . 46
1.13 IV Regression results for per-capita GDP volatility on export concen-
tration - Herfindahl index . . . . . . . . . . . . . . . . . . . . . . . . 50
1.14 Random effects regression results for per-capita GDP volatility on
export concentration - Herfindahl index . . . . . . . . . . . . . . . . 52
vii
1.15 IV Regression results for sectoral output volatility on export concen-
tration - Herfindahl index . . . . . . . . . . . . . . . . . . . . . . . . 53
1.16 IV Regression results for sectoral output volatility on export concen-
tration - Theil index . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
2.1 Summary Statistics - Export Variety: Unweighted Count, Product:
HS-6 level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
2.2 Summary Statistics - Export Variety: Unweighted Count, Product:
HS-4 level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
2.3 Summary Statistics - Export Variety: Unweighted Count, Product:
HS-2 level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
2.4 Summary Statistics - Overall Export Variety: Unweighted Count . . 79
2.5 Change in the Average Export Variety (in %) - Weighted Count . . . 87
2.6 Change in the Overall Export Variety (in %) - Weighted Count . . . 89
2.7 Gravity Regression results for Extensive Margin of Exports - variety:
product-market pair . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
2.8 Decomposition of Relative Export Growth with the Rest of the
World (%) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
2.9 Decomposition of Relative Export Growth with the EU Partners (%) 102
2.10 Decomposition of Relative Export Growth with the Non-EU Part-
ners (%) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
2.11 Summary Statistics: Decomposition of Relative Export Growth (%) 104
A.1 Correlation Matrix for the Export Diversification (Concentration)
Measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
A.2 List of Economic Activity Codes and Names . . . . . . . . . . . . . 115
A.3 Summary Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
A.4 Variable Definitions and Sources - Aggregate Analysis . . . . . . . 117
A.5 Variable Definitions and Sources - Sectoral Analysis . . . . . . . . . 118
B.1 Partner Country List . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
viii
List of Figures
2.1 Fraction of the least traded goods in total exports: Czech Republique 82
2.2 Fraction of the least traded goods in total exports: Estonia . . . . . 82
2.3 Fraction of the least traded goods in total exports: Hungary . . . . 83
2.4 Fraction of the least traded goods in total exports: Latvia . . . . . . 83
2.5 Fraction of the least traded goods in total exports: Lithuania . . . . 84
2.6 Fraction of the least traded goods in total exports: Poland . . . . . 84
2.7 Fraction of the least traded goods in total exports: Slovakia . . . . 85
2.8 Fraction of the least traded goods in total exports: Slovenia . . . . 85
ix
Abstract
This dissertation is composed of two essays, whose objective is to better under-
stand the effects of trade liberalization on the export variety, and the impact of
the change in the export structure on the output volatility during liberalization
episodes.
In the first chapter, I investigate the effects of trade liberalization on output
volatility. Traditional comparative advantage theory such as Heckscher-Ohlin
and the New Trade Theory pioneered by Krugman and Melitz are at odds when
predicting the effects of trade liberalization on export diversification. The tra-
ditional view is that trade openness results in greater trade specialization and
generates output volatility by exposing countries to external shocks. The new
trade theory on the other hand implies that trade liberalization leads to growth
in the extensive margin of exports. This expansion in the extensive export margin
leads to greater export diversification. More diversified markets mean that with
more trade liberalization; countries are less exposed to external shocks.
Matching highly disaggregated export data with aggregate- and industry-
level production data of Central and Eastern European Countries (CEECs) that
joined the EU in 2004, I revisit the openness and volatility link with a particular
x
focus on the effects of the extensive margin of exports. The entry of the CEECs
into trade relationships with the West after the end of the Cold War represents a
unique natural experiment. My statistical analysis shows that although more
openness increases volatility, this effect can be stabilized through sufficiently
diversified export baskets. I find that trade liberalization episodes result in major
growth of the extensive margin of exports. I demonstrate that this export expan-
sion has a consistent and significantly negative effect on both per-capita GDP
and sectoral output volatility. Specifically, for the average country and the sec-
tor, a 10% increase in the extensive margin of exports leads to 4.6% decline in
per-capita GDP volatility and 1.12 % decline in sectoral output volatility. I also
find that the geographical diversification of exports is a more dominant factor in
reducing output volatility, than increase in the product variety.
In the second chapter, I study, empirically, the impact of trade liberalization
on the extensive margin of exports. Using the highly disaggregated trade data
used in the previous chapter, I investigate the evolution of the extensive margin of
exports of the CEECs following their accession to the EU. I employ different mea-
sures of export diversification including simple count of products-market pairs,
the least-traded goods approach developed by Kehoe and Ruhl (2013) and the
weighted count measure developed by Hummels and Klenow (2005). I show that
the startling growth rates in the exports between pre- and post-EU periods are
strongly associated with a sizable expansion in the variety of the product-market
pairs the CEECs trade in. Specifically, I find that following the EU accession; the
CEECs experienced a 42.4% growth in the extensive margin of exports, whereas
the intensive margin growth remained at 18.9%. I demonstrate that this expan-
sion largely depends upon the growth in the number of destination markets. I
further show that the share of the least traded goods in 1995 rises by at least 30%
xi
by the end of the period in study, pointing to a major enlargement of the export
baskets of the CEECs. My empirical analysis shows that the growth in the exten-
sive margin starts five years ahead of the EU agreement. I further decompose the
relative export growth (with respect to the rest of the world) into its extensive
and intensive margin components and find that on average the relative exports
of the CEECs grow by 61.3%, to which the contribution of the extensive margin
is 80.5%, whereas the contribution of the intensive margin is 19.5%.
xii
Chapter 1
Export Diversification and Output Volatility
– Empirical Evidence from Central and
Eastern European Countries
1.1 Introduction
Output volatility is one of the main factors that slow down economic growth
(Aizenman and Marion, 1999, Ramey and Ramey, 1995). Although trade open-
ness is known to foster economic growth (Calderon et al., 2005, Jeffrey A. Frankel,
1999), the relationship between trade openness and output volatility is a topic
with little consensus.
A widely held view is that greater openness to trade leads to higher GDP
volatility by exposing countries to external shocks. According to this view, open-
ness results in higher volatility through two separate channels. First, higher
openness is associated with higher specialization, which leaves countries vul-
nerable to sector-specific shocks that result in greater terms-of-trade volatility.
If the sectors that the country specializes in are more prone to terms-of-trade
fluctuations, greater openness will bring about greater output volatility, hurting
economic growth and welfare (Bejan, 2006, di Giovanni and Levchenko, 2009,
1
Jansen, 2004). The second channel is through increased exposure to country-
specific (foreign demand) shocks. Openness to trade increases output volatility
by transmitting demand shocks to other countries (Bacchetta et al., 2009).
Traditional comparative advantage theory suggests trade liberalization and
greater openness lead to specialization in production and exports of a country.
In other words, given that the openness of a country is measured as the ratio of
total trade to the GDP , increased openness mainly comes from the growth in the
intensive margin of trade (i.e., the growth in the trade of goods that have been
previously traded). When all the growth in total trade is assumed to originate
from the intensive margin, as it does in the traditional models, terms-of-trade
fluctuations appear to be the major source through which openness affects output
volatility.
Recently, there has been vast development in a newer trade theory where
firms face fixed costs of exporting (Anderson and Wincoop, 2004, Helpman et al.,
2008, Melitz, 2003). In this newer trade theory, higher openness does not neces-
sarily increase specialization of exports. Many empirical studies, which are based
on the new trade models with monopolistic competition that incorporate fixed
trade costs and sunk market-penetration costs, show the extensive margin of
trade (i.e., trade in the varieties that have not been traded before) plays as impor-
tant a role as the intensive margin of trade in contributing to trade growth (Broda
and Weinstein, 2006, Hummels and Klenow, 2005, Kehoe and Ruhl, 2013).
Hummels and Klenow (2005) compare the implications of Armington model
based on a more traditional theory, which has no extensive margin, with the
implications of Krugman’s monopolistic competition model in which each coun-
try produces an endogenous number of varieties. They find that 60% of the
2
greater exports of larger economies is due to the extensive margin. Kehoe and
Ruhl (2013) show a significant growth in the extensive margin characterizes
trade-liberalization episodes and periods of structural change. What is more,
they find the extensive margin contributes significantly to trade growth dur-
ing episodes of rapid economic growth and development.
1
It is also shown
when there are market-penetration costs, the response of firms in changing their
exporter status to temporary distortions such as business cycles is weak, which
suggests the extensive margin of exports does not co-move with the temporary
fluctuations in the economy.
2
The main purpose of this paper is to contribute to this line of research by
studying the link between export diversification and output volatility, both at
the aggregate and sectoral levels. To this purpose, I use two separate panel
data sets on the trade and production of eight Central and Eastern European
countries (CEECs), that joined the European Union in 2004. The first data set
includes disaggregated bilateral trade data along with GDP data, covering the
period 1995-2012. Looking at aggregate volatility may mask the real impact of
the extensive margin because the correlation between the traded and non-traded
sectors in the economy would also appear in the aggregate volatility. In the
sectoral data set I match the export data to the production data of the traded
sectors at the 6-digit product-level. This allows me examine the immediate effect
of a change in the extensive margin on the volatility of the production.
1
Kehoe and Ruhl (2013) show that the share of the least traded goods in Korean exports to the
United States went from 10% to 60% during 1975-1980. Currently, a similar pattern is present in
Chinese exports.
2
Alessandria and Choi (2007), Ruhl (2008)
3
I begin my analysis by constructing an extensive margin measure, which
evaluates the variety of the export basket of a country depending on its weight
in the world trade. I construct this measure using highly disaggregated bilateral
trade data specifically, disaggregated at the 6-digit level of Harmonised System
(HS) Classification for the period 1995-2012.
Second, using panel regressions, I investigate the relationship between
export diversification in terms of product-market pairs and per-capita GDP
volatility, while controlling for openness and terms-of-trade volatility, along with
other sources of income volatility that have been established in the literature.
Next, I decompose the impact of diversification into its product and market com-
ponents by employing product and market diversification measures one at a time
in my regressions.
In addition to aggregate volatility, I also examine the link between export
diversification and output volatility at the sectoral level. To this purpose, I com-
bine the trade data with the sectoral output data at the 6-digit product level.
Then, through a set of panel regressions, I explore the link between sectoral
export diversification and sectoral output volatility, while controlling for export
openness, productivity and government funding levels (of each sector).
My findings show that the diversification of exports in terms of product-
market pairs plays a significant role in reducing output volatility, both at the
aggregate and sectoral levels. At the same time, openness to trade and terms-
of-trade fluctuations raise output volatility as predicted in the existing literature.
When per-capita GDP volatility is considered, the export diversification measure
4
consistently enters the regressions with a negative and highly significant coef-
ficient. This result is robust to different estimation techniques, though my pre-
ferred model is the instrumental variables (IV) regression that accounts for the
endogeneity between trade openness and output volatility. The estimates of the
benchmark IV regression suggest that for the average country, a 10% increase in
the extensive margin of exports leads to 4.6% decline in per-capita GDP volatility.
Furthermore, using the aggregate data set, I also show diversification in terms of
destination markets plays a more important role in reducing aggregate volatil-
ity compared to diversification due to different products, whose effect, while
reducing volatility, is insignificant. When I study the sectoral data set, I find
the extensive margin of exports remains a significant factor in reducing sectoral
output volatility. Although the stabilizing effect is weaker at the sectoral level,
the extensive margin measure has a coefficient that is consistently negative and
significant in various estimation techniques.
I present additional evidence showing the stabilizing effect of diversifica-
tion is strongest during trade-liberalization episodes, both at the aggregate and
sectoral levels. When I control for the accession of the CEECs to the EU, I find
that even though the effect of the extensive margin is ambiguous prior to the
EU accession, it becomes stronger and negative afterward. The coefficient of
the extensive margin effect is always greater (in absolute value) for the post-EU
period, which results in a negative overall effect during the period following the
EU accession.
This paper contributes to the existing literature in several different ways.
First, even though an ample amount of empirical literature investigates the effects
of trade openness on output volatility, very little work has focused on the direct
5
effect of export diversification.
3
This paper bridges this gap by directly relating
the extensive margin of exports to both aggregate and sectoral output volatility.
Second, existing research either uses traditional concentration indices (Cav-
allo et al., 2008, Haddad et al., 2013) or a simple count (Buch et al., 2009, di Gio-
vanni and Levchenko, 2009) of varieties to measure export diversification. In this
paper, my main diversification measure is the extensive margin measure devel-
oped by Hummels and Klenow (2005), which weights the export varieties of a
country depending on their importance in world trade. This way of measuring
diversification prevents the extensive margin from being overestimated simply
because a country exports a single good to a specific partner in high volumes.
Third, the data set in this paper consists of countries that underwent a major,
exogenously induced trade-liberalization episode. The entry of Eastern and Cen-
tral European countries into the world-trading regime - mainly dictated by post-
Cold War political events - represents a unique natural experiment on the role of
trade liberalization on the volatility. Previous empirical studies utilize data sets
(cross-sectional or panel data) that include a large set of countries over a long
time horizon. Such large data sets lack the ability to draw specific conclusions
concerning the relationship between the liberalization process and the structural
change in the exports and the output volatility. The selection of countries and
the time frame covered in this study provide a unique opportunity to conduct
a natural experiment on the relationship between openness, diversification and
volatility, because, following their accession to the EU in 2004, the CEECs expe-
rienced a substantial increase in their trade openness along with a significant
3
Among the few are Bejan (2006), Cavallo et al. (2008), Buch et al. (2009), and Haddad et al.
(2013).
6
expansion in their export variety. This characteristic of the data gives me the
opportunity to examine the actual effect of extensive margin growth on output
volatility in the economies that are experiencing an increase in their openness.
The remainder of the paper is organized as follows. Section 2 presents some
preliminary findings relating trade openness to export diversification, and gives
descriptive information for the key variables of interest. The CEECs have under-
gone major trade liberalizations and other structural transformations that affect
the degree of openness of their economies during the period 1995-2012. Here, I
document that a significant spike in trade openness characterizes this period,
accompanied by a major growth in the extensive margin of exports, for the
CEECs. I find that while the terms-of-trade volatility also increases, the volatil-
ity of GDP per capita declines. Section 3 reviews the prior literature. Section 4
describes the data and the empirical strategy that I adopt. Section 5 reports and
discusses the main regression results. Section 6 includes some robustness checks
of the relationship between extensive margin and volatility. Section 7 concludes.
1.2 Trade Openness and Export Diversification
Although in general, trade liberalization and greater trade openness is expected
to lead to higher output volatility, a theoretical ambiguity exists. On the one
hand, greater trade openness leads to greater specialization according to com-
parative advantage, which leaves countries more vulnerable to terms-of-trade
fluctuations and foreign demand shocks. On the other hand, the new trade the-
ory of Krugman and Melitz predicts that trade liberalization will lead to a greater
variety of products being exported, given the lowering of trade costs.
7
Table 1.1
Descriptive statistics: Export diversification, variety: product-market pair
Variable Obs. Mean Std. Dev. Min Max
EM
v
144 0.405 0.149 0.114 0.706
Theil-between
v
144 1.327 0.439 0.574 2.566
HI
v
144 0.057 0.025 0.024 0.138
Num. of Varieties 144 44318 28931 9565 135087
Notes: Table reports the descriptive statistics export diversification measures between the periods pre- (1995-2003) and
post-EU (2004-2012). Each HS-6 category defines a product and each product-market combination is defined as a variety.
Diversification measures are calculated for varieties. EM is the extensive margin, Theil-Between is the between compo-
nent of the Theil index and HI is the Herfindahl-Hirschman index of concentration. Numbers are based on the author’s
calculations from UN COMTRADE data.
Echoing this theoretical ambiguity, the empirical research yields mixed
results. Some of the existing literature relates openness to higher output volatil-
ity, mainly through terms-of-trade shocks (Cavallo et al., 2008, Easterly and
Kraay, 2000, Jansen, 2004). On the other hand, other research find that trade lib-
eralization leads to less volatility, especially for industrialized countries (Bejan,
2006, Krishna and Levchenko, 2013).
The prediction that trade liberalization and openness lead to higher special-
ization is based on comparative-advantage theory. Research based on traditional
comparative advantage theory assume the growth in total trade following a lib-
eralization episode comes fully from the intensive margin growth, which results
in higher specialization of the economy.
Recent theoretical and empirical studies, however, argue the extensive mar-
gin of trade plays an equivalent role in increasing the aggregate volume of trade
(starting from Melitz (2003)). Thus, a trade-liberalization episode leading to
higher openness does not necessarily increase specialization of exports; on the
contrary it can result in a more diversified export basket (Anderson and Win-
coop, 2004, Broda and Weinstein, 2006, Debaere and Mostashari, 2010, Helpman
8
et al., 2008, Kehoe and Ruhl, 2013, Melitz, 2003). Kehoe and Ruhl (2013) in par-
ticular show a significant and robust link exists between total trade growth and
growth of the extensive margin. Moreover, these authors document that signif-
icant growth in the extensive margin characterize trade-liberalization episodes
and periods of structural-change.
Table 1.1 reports descriptive statistics for various export diversification mea-
sures along with the simple count of varieties exported for the aggregate data set.
The percentage change in export diversification in terms of product-market pairs
between the pre- (1995-2003) and post-EU (2004-2012) periods, for all the coun-
tries in the sample are reported in Table 1.2. Contrary to the traditional view,
the CEECs clearly experienced a substantial growth in their extensive margin
following their accession to the EU. The data reveal the growth in the exten-
sive margin ranges between 14% (Hungary) and 114% (Poland). On average, the
extensive margin of exports grew by 47.66% during the post-EU period. Tra-
ditional concentration measures also confirm these findings. Both Theil index
between component and the Herfindahl index display a negative trend pointing
to a declining concentration of exports.
The change in per-capita GDP volatility, average trade openness, and terms-
of-trade volatility over the periods 1995-2003 and 2004-2012 are reported in Table
1.3. Note that accession to the EU is a significant liberalization episode for the
CEECs. The data reveal accession to the EU is associated with a major spike in
the openness of CEECs. The growth rate of openness ranges from 51% (Slovenia)
to 141% (Slovakia) with an average of almost 97%. At the same time, the major-
ity of the countries experienced a decline in per-capita GDP volatility, though
unlike openness, the change in volatility varies greatly among countries. In
9
Table 1.2
Change in the average export diversification (in %), between pre- (1995-2003)
and post-EU (2004-2012) periods, variety: product-market pair
Country EM
v
Theil-Between
v
HI
v
Czech Rep. 16.96 -28.97 0.80
Estonia 34.00 -19.26 -3.98
Hungary 14.00 -20.11 -29.38
Latvia 88.05 -39.88 -50.25
Lithuania 53.07 -35.51 -15.85
Poland 114.74 -69.12 -13.37
Slovakia 32.73 -25.83 -2.52
Slovenia 27.73 -32.19 -14.62
Overall 47.66 -33.86 -16.15
Notes: Table reports the percentage change in the average export diversification measures between the periods pre- (1995-
2003) and post-EU (2004-2012). Each HS-6 category defines a product and each product-market combination is defined as
a variety. Diversification measures are calculated for varieties. EM is the extensive margin, Theil-Between is the between
component of the Theil index and HI is the Herfindahl-Hirschman index of concentration. Numbers are based on the
author’s calculations from UN COMTRADE data.
Czech Republic, output volatility drops by only 3.6%, whereas the decline is
50.3% in Estonia and 70% in Hungary. On the other hand, overall, terms-of-trade
volatility increases by 35%.
In summary, the data presented in Table 1.3 reveal that the trade-
liberalization episode in CEECs resulted in greater trade openness along with
higher terms-of-trade volatility but lower GDP volatility. This result contra-
dicts the traditional view that increased openness leads to increased volatility.
Although I leave the detailed empirical analysis to the following sections, con-
sidering the substantial growth in the extensive margin of exports during this
period, I argue at this point that the main factor behind the decline of the out-
put volatility in the CEECs is the greater trade openness combined with greater
export diversification.
The logic behind this argument is twofold. First, greater openness increases
a country’s exposure to different types of external shocks. On the one hand,
10
Table 1.3
Change in the average openness and volatility of the economic indicators (in %),
between pre- (1995-2003) and post-EU (2004-2012) periods
Country GDP/cap volatility Trade openness terms-of-trade volatility
Czeck Rep. -3.57 111.44 -41.97
Estonia -50.26 62.52 111.96
Hungary -70.22 100.04 133.55
Latvia -45.99 92.46 121.14
Lithuania -32.18 111.10 74.07
Poland 0.41 106.25 -62.40
Slovakia 22.89 140.61 -43.71
Slovenia -48.23 50.70 -15.01
Overall -28.39 96.89 34.71
Notes: Table reports the percentage changes in the per-capita GDP volatility, average openness, and terms-of-trade
volatility, between the periods pre- (1995-2003) and post-EU (2004-2012). Trade openness is measured as (Exports+
Imports)/GDP. Volatility is measured as the relative standard deviation of the relevant variable (standard deviation as a
percentage of the mean). Numbers are based on the author’s calculations from UN COMTRADE and World Bank data.
the country will be vulnerable to terms-of-trade shocks, which are regarded as
negative product-specific shocks. On the other hand, greater openness increases
a country’s exposure to different markets, which will increase output volatility
through spillover effects, especially if the majority of a country’s trading partners
experience high volatility. However, if a country’s exports are sufficiently diversi-
fied in terms of both products and markets, the effect of openness can be reversed
because the high diversity in the variety of exports will provide insurance against
both product- and market-specific shocks.
Second, while penetrating new markets requires large sunk costs, firms do
not stop exporting a product to a country when the economy experiences a tem-
porary change because the change in the expected future profits can be ignorable.
Thus, the extensive margin does not respond to the business-cycle fluctuations
and acts as a stabilizer against temporary shocks in the economy. Ruhl (2008)
formalizes this approach through a quantitative general equilibrium model in
which firms face uncertainty about future profits and sunk costs. He shows that
11
Table 1.4
Descriptive statistics: Sectoral export diversification, variety: product-market pair
Variable Obs. Mean Std. Dev. Min Max
EM
vs
2872 0.484 0.259 0.001 1.000
Theil-Between
vs
2872 1.363 0.487 0.154 4.234
HI
vs
2869 0.195 0.152 0.007 1.000
Num. of Varieties
s
2872 44344 28868 9565 135087
Notes: Table reports the descriptive statistics of the export diversification measures in terms of variety (product-market
pair). The diversification measures are calculated by economic activity (sector). The data are obtained from UN COM-
TRADE and Eurostat Economy and Finance database.
although the extensive margin responds to permanent changes such as trade lib-
eralization, its response to business cycle-fluctuations is weak.
Now I move to the sectoral analysis of the data. Table 1.4 reports the descrip-
tive statistics for the export diversification measures for the sectoral data set.
Although the extensive margin measure and the Theil index have similar aver-
ages to the aggregate data, the Herfindahl index points to a much larger concen-
tration of exports at the sectoral level. Additionally, sectoral data display slightly
higher standard deviations. Among the three measures, the Theil index exhibits
the highest standard deviation, which is not surprising, because unlike the exten-
sive margin measure and the Herfindahl index, the Theil index can take values
greater than 1.
Table 1.5 presents the descriptive statistics of the other variables derived
from the sectoral data, for the two sample periods. When we compare the sec-
toral output volatility between pre- and post-EU periods, we do not see a clear
decline (1.1 percentage point), as in the aggregate data. However, note that con-
siderable heterogeneity exists among sectors in terms of volatility change during
this period.
12
Table 1.5
Descriptive statistics: Selected variables, by economic activity, before (1995-2003)
and after (2004-2012) EU
Variable Obs Mean Std.
Dev.
Min Max
Panel A: Before-EU (1995-2003)
Output volatility 1380 0.233 0.167 0.036 1.202
Export openness 1330 0.532 1.058 0.000 25.790
Output per worker 1093 58.943 76.334 5.591 1094.977
Fixed capital consumption per worker 882 5.146 6.592 0.000 59.040
Taxes paid to government 1305 0.001 0.015 -0.149 0.043
Panel B: After-EU (2004-2012)
Output volatility 1393 0.222 0.095 0.072 0.623
Export openness 1352 1.540 16.694 0.000 462.797
Output per worker 1339 133.786 226.954 0.000 2485.224
Fixed capital consumption per worker 1391 9.187 12.582 0.000 108.497
Taxes paid to government 1352 -0.002 0.023 -0.181 0.040
Notes: Each variable is calculated by economic activity (sector), based on the NACE classification, using the sectoral data
set. The table reports the descriptive statistics of the selected variables, for the pre- (1995-2003) and post-EU (2004-2012)
periods, separately. The data are obtained from the Eurostat Economy and Finance database.
On the other hand, the post-EU period represents a major increase in export
openness. Average export openness of a sector almost triples after EU acces-
sion. A similar trend is observed in labor productivity, which more than doubles
during the post-EU period.
In terms of export diversification growth, the data are similar to (though
slightly lower than in) the aggregate data. Table 1.6 reports the average change
in the export diversification by economic activity, after the accession of CEECs to
EU. The average sectoral extensive margin grew by almost 41% between pre- and
post-EU periods. Both the Theil and Herfindahl indices confirm these findings,
though the Herfindahl index points to a much smaller decline in sectoral export
concentration.
13
Table 1.6
Average change in the export diversification by economic activity (in %),
between pre- (1995-2003) and post-EU (2004-2012) periods, variety:
product-market pair
Country EM
vs
Theil-Between
vs
HI
vs
CzechRep. 17.48 -22.68 -3.11
Estonia 20.18 -14.25 2.68
Hungary 26.35 -12.33 -5.37
Latvia 61.40 -32.38 -19.12
Lithuania 41.83 -26.63 -15.49
Poland 93.97 -62.35 4.10
Slovakia 31.47 -11.57 -4.49
Slovenia 34.26 -27.57 6.07
Overall 40.87 -26.22 -4.34
Notes: The table reports the average percentage change in the export diversification measures in terms of variety (product-
market pair), between the periods pre- (1995-2003) and post-EU (2004-2012). EM
vs
is the extensive margin of the sectoral
exports, Theil-Between is the between component of the Theil index and HI is the Herfindahl-Hirschman index of con-
centration. The diversification measures are calculated by economic activity, based on the NACE classification. The data
are obtained from the UN COMTRADE database.
To examine the link between export diversification and output volatility at
the sectoral level, Table 1.7 documents the percentage changes in selected vari-
ables, for the first three economic activities that experienced the largest decline
in output volatility, between the periods before and after EU. The results show
two stylized facts. First, the sectors with the highest decline in output volatility
almost always increased their export diversification in terms of product-market
categories, the growth rate ranging between 7% and 135%. Second, no clear link
exists, positive or negative, between export openness and output volatility at the
sectoral level. The change in the export openness of the sectors ranges between
-94% to 368%. This observation tells us that a significant degree of heterogeneity
exists in terms of the response of the sectors to higher export openness.
14
Table 1.7
Change in the selected variables (in %), in the sectors with the largest decline in
output volatility, between pre- (1995-2003) and post-EU (2004-2012) periods
Economic Output Export Theil-
Country activity volatility openness TOT risk EM
vs
between
vs
Czech Rep. C26 -63.39 104.94 102.97 17.77 -37.27
C29-C30 -31.71 8.57 4.62 7.11 -30.00
C22-C23 -31.14 3.76 -2.37 16.13 -31.93
Estonia C31-C33 -64.41 31.38 30.24 41.77 -19.41
C16-C18 -59.01 -26.19 -24.65 21.61 -17.81
J -52.61 205.09 198.99 9.98 -34.25
Hungary C26 -78.45 33.74 104.40 17.25 -36.45
J -63.91 26.97 23.58 -0.93 -20.39
M -58.94 -94.89 -94.41 4.90 86.12
Latvia C13-C15 -85.65 62.41 307.05 92.81 -39.15
M -53.14 -93.15 -71.61 -11.65 59.97
J -47.78 -44.09 206.97 115.41 -45.31
Lithuania C16-C18 -66.43 -8.06 25.76 31.49 -27.18
C19 -66.43 -8.06 25.76 71.92 -21.71
J -64.36 83.90 144.74 32.30 -46.09
Poland C13-C15 -58.65 41.01 -25.02 91.66 -70.09
C27 -11.86 41.59 -30.02 135.68 -73.22
C31-C33 -6.09 -0.51 -51.77 101.92 -71.56
Slovakia C29-C30 -44.39 -18.49 -81.05 24.19 -27.36
C13-C15 -42.28 42.48 -63.81 25.46 -23.19
D -28.32 -29.47 -92.57 -0.04 23.68
Slovenia B -44.33 368.94 487.63 22.78 -23.94
A -42.20 244.22 316.49 130.57 -40.84
C21 -34.28 52.89 89.59 27.33 -16.76
Notes: The table reports the percentage changes in the output volatility, average export openness, average TOT risk, aver-
age extensive margin, and average Theil-between index, between the periods pre- (1995-2003) and post-EU (2004-2012),
for the selected sectors. Only data for the first three sectors that experienced the highest output volatility decline are
reported. The names of the economic activities are reported in the Appendix Table A.2.
1.3 Related Literature
Most of the early empirical papers on the subject focus on the relationship
between terms-of-trade variability and output volatility (the sector-specific-
shocks channel). These papers base their analysis on the traditional Ricardian
15
comparative-advantage models that consider only the intensive margin of trade
while ignoring the extensive margin. Even if they include some measure of the
export diversification in their analysis, it is usually considered as a control vari-
able, and no clear conclusion is made stating that higher diversification is asso-
ciated with either higher or lower volatility.
A simple two-country - two-good Ricardian model developed by Razin et al.
(2002, 2003) predicts trade openness may lead to discrete boom-bust cycles of
investment, which causes the terms-of-trade to oscillate and the economy to
destabilize. With the help of an optimal monetary policy model, Gal´ ı and Mona-
celli (2005) find increased openness leads to greater output volatility, while reduc-
ing the volatility of the exchange rate and having an ambiguous effect on con-
sumption. Using a similar monetary model, Guender (2006) shows increased
openness results in higher volatility through the stronger effects of terms-of-
trade changes. He further predicts that whereas domestic inflation targeting has
a U-shaped effect on output volatility, CPI-inflation targeting results in a strict
positive relationship between output variance and openness. On the other hand,
by means of a simple theoretical model, Karras (2006) demonstrates the effects of
openness on output volatility are ambiguous, both theoretically and empirically.
In his paper where he relates the degree of openness of a country to govern-
ment size, Rodrik (1998) finds greater exposure to external risk results in greater
income volatility. However, although he controls for the terms-of-trade volatility
as the external risk, he does not take into account the degree of concentration
of exports. Easterly and Kraay (2000) examine income, growth and volatility in
small states. They find that greater income volatility of small states comes from
16
their greater openness to international trade, which causes large fluctuations in
their terms-of-trade.
Using a general equilibrium Ricardian model of trade based on Eaton and
Kortum (2002), Burgess and Donaldson (2012) investigate the effect of declining
transportation costs on the regional income volatility in India. They find that
as the openness increases, responsiveness of real income to productivity shocks
depends on the price volatility. Kim (2007) distinguishes between openness and
external risk in their effect on volatility and finds that once the effect of term-
of-trade and exchange rate risk is isolated, openness ceases to be a significant
determinant of volatility.
di Giovanni and Levchenko (2009) relate output volatility to trade openness
using industry-level data and find that sectors more open to international trade
are more volatile. They also document that trade increases specialization, which
also plays a role in increasing volatility. Although they do not directly link the
extensive margin to output volatility, they investigate the link between openness
and volatility of number of firms in an industry. They find that openness induces
higher volatility of the entry and exit of the firms and the overall effect of open-
ness on output volatility remains positive.
In the existing literature, even though export diversification is considered
as a control variable in some of the studies, a direct link between diversifica-
tion and volatility has not been established. One of the underlying reasons for
this shortcoming is that these papers are based on the conventional comparative-
advantage models, which fail to account for the fixed and sunk costs the firms
face in making export decisions. When firms do not face any market-penetration
costs, the change in the aggregate trade is mistakenly assumed to be all intensive
17
margin growth. However, as mentioned earlier, recent theoretical and empirical
trade literature implies the extensive margin is also a significant contributor of
trade growth.
4
Research has shown increased openness results in a greater exten-
sive margin, which does not respond to business cycle-fluctuations as strong as
the terms-of-trade. Therefore, any attempt to explain the association between
openness and output volatility without the extensive margin of trade would be,
if not misleading, incomplete.
Ruhl (2008) develops a model that incorporates international real business-
cycle theory, in which firms face uncertainty about future profits and sunk export
costs. Under a general equilibrium setting, they show the response of firms in
changing their exporter status to temporary changes such as business cycles is
weak. These types of changes emerge as price changes in the intensive margin
of trade. However, a permanent change such as a reduction in tariffs induces
more firms to begin exporting, thus increasing the extensive margin. These find-
ings suggest openness to trade brings greater terms-of-trade volatility, but also
leads to growth in the extensive margin, which does not respond to short-term
fluctuations. Therefore, the extensive margin can play a stabilizing role against
aggregate volatility.
5
Tenreyro et al. (2014) develop a multi-sector variation of the Eaton and Kor-
tum (2002) model with stochastic shocks to assess the significance of the two
mechanisms, sectoral specialization and market diversification, in determining
4
Broda and Weinstein (2006), Hummels and Klenow (2005), Kehoe and Ruhl (2013)
5
Alessandria and Choi (2007) construct a similar, standard international real business-cycle model
with export entry costs. They find that the effect of the extensive margin is negligible for aggre-
gate quantities at business-cycle frequencies.
18
the role of trade liberalization in output volatility. When exposure to country-
specific shocks is an important source of volatility, openness to trade can lower
output volatility by reducing exposure to domestic shocks and allowing coun-
tries to diversify the sources of demand and supply across countries. Addition-
ally, the paper shows that higher sectoral specialization does not necessarily lead
to higher volatility. The direction of the GDP volatility depends on the intrinsic
volatility of the sector that the country specializes in, as well as the covariance
of the sectoral shocks. Note that Tenreyro et. al.’s (2014) model differs from
the new trade models in the sense that openness leads to higher specialization
of production while it accounts for the market diversification (a broad range of
intermediate goods are imported from the rest of the world). Therefore, it fails
to capture the full effect of the extensive margin as defined earlier. Although it
predicts a stabilizing impact of market diversification on volatility, it lacks the
ability to make any statements concerning the effect of product diversification.
Jansen (2004) investigates the relationship between income volatility and
terms-of-trade shocks in small and developing economies. He shows that both
higher openness and higher export concentration characterize these countries.
After controlling for several structural determinants, he finds a strong positive
relationship between terms-of-trade volatility and export concentration. In a sec-
ond set of regressions, he confirms that higher terms-of-trade volatility leads to
higher income volatility, which leads to the conclusion that higher export con-
centration leaves small economies vulnerable to terms-of-trade shocks, which in
return increases income volatility. However, this approach does not establish a
direct link between volatility and trade diversification. In a similar study, Bejan
(2006) finds mixed results for developing and developed countries. She argues
19
that in developing countries, openness increases volatility through the terms-of-
trade channel, whereas export concentration has no significant effect. In devel-
oped countries, although both terms-of-trade volatility and export concentra-
tion play an important role in increasing volatility, openness decreases volatility
because it offers more possibilities for hedging against country-specific shocks.
In another study, Cavallo et al. (2008) treat terms-of-trade volatility as a con-
ditioning factor on the relationship between openness and volatility, while intro-
ducing export concentration as another control variable into the regression. They
find that increased volatility due to higher openness mainly comes from terms-
of-trade volatility. Once the effect of terms-of-trade volatility is controlled for,
openness plays a stabilizing role in output volatility. He finds no evidence sug-
gesting export diversification affects output volatility. Buch et al. (2009) relates
export openness to firm-level output volatility and finds export status is a signif-
icant factor in determining the firm-level output volatility. Exporter firms have
lower volatility and most of the reduction in volatility comes from the market
diversification of the firms.
Bacchetta et al. (2009) study the possible role of geographical concentration
of exports and exposure to demand shocks in partner countries as a source of
economic volatility. They decompose the partner countries’ volatility of demand
into two components: a variance and a covariance component. Using panel data
regressions for different country samples, they find geographical diversification
of exports plays an important role in reducing the exposure to external shocks.
Next, they run a regression of output volatility on terms-of-trade volatility, part-
ner countries’ volatility, and a set of control variables, and they find partner coun-
tries’ volatility has a positive and significant impact on exporters’ GDP volatility.
20
In particular, they find that the correlation among the various trading partners’
business cycles (measured by the covariance component) is an important factor
in determining the domestic volatility.
In a recent study that is possibly the closest to this one, Haddad et al.
(2013) investigate the variation of the effect of trade openness on output volatil-
ity depending on the concentration level of exports. Using a comprehensive
set of export diversification measures, in terms of both products and markets,
they run a set of regressions by which they aim to find a cutoff value for the
export-concentration level below which the effects of openness on output volatil-
ity becomes negative. They find trade openness has a decreasing effect on output
volatility in countries where the export concentration (Herfindahl index of prod-
uct concentration) is below 0.24. On the other hand, they do not find significant
evidence suggesting market concentration plays a role in determining the effect
of openness on volatility.
The majority of the prior research that relates trade openness to output
volatility focuses either on aggregate volatility or industry level volatility. In
addition, most of the studies employ data sets that consist of either cross-section
or panel data of a large number of countries that are at different stages of devel-
opment or liberalization episodes. In this study, I attempt to establish a direct
association between output volatility and export diversification, by employing
measures capturing the extensive margin for products, markets and product-
market pairs, both at the aggregate and sectoral levels. By focusing on a set of
countries that underwent a major trade-liberalization episode during the time
frame chosen, I analyze the effects of the extensive margin of exports on the
volatility of per-capita GDP and sectoral output.
21
This study is unique in the sense that the choice of the country set and
time frame constitutes a natural experiment by which I examine the actual link
between export diversification and output volatility in the economies that go
through a major trade-liberalization episode (EU accession) so that their open-
ness to trade increases considerably. My findings support, in part, the conven-
tional view that openness and terms-of-trade fluctuations contribute to output
volatility, both at the aggregate and sectoral levels.
However, I find strong evidence that the effect arising from the diversifica-
tion of exports counteracts this effect. Furthermore, I document that growth in
the extensive margin of exports in terms of destination markets is the driving
force behind the stabilizing effect of export variety growth on output volatil-
ity. An important mechanism behind this result may be that exporting to a
larger number of destinations helps diversify the risks emerging from demand
shocks. Additionally, Malik and Temple (2009) find export concentration in terms
of products is a strong determinant of terms-of-trade volatility. Therefore, the
effect of product diversification on output volatility may be observed through
its effect on terms-of-trade volatility. On the other hand, I find that whereas the
effect of diversification appears to be ambiguous during the pre-EU period, it is
significantly stabilizing during the post-EU period. In this sense, I argue that
as the economies that go through trade liberalizations diversify their exports,
they insure against terms-of-trade and other external risks that are transmitted
through trade.
22
1.4 Data and Methodology
In this study, I use two different cross-sectional time-series data sets. The first
comprises per-capita GDP and trade data of a balanced panel of eight Central
and Eastern European Countries that joined the EU in 2004, covering the period
1995-2012. To be specific, my sample includes Czech Republic, Estonia, Hungary,
Latvia, Lithuania, Poland, Slovakia, and Slovenia. The second data set includes
output and trade data at the sectoral level, for the same set of countries and
time frame. I use these two data sets to test and compare the effect of export
diversification on both aggregate and sectoral output volatility.
Studying the case of CEECs for the period 1995-2012 is interesting because
during this period, the CEECs underwent an extensive trade-liberalization
episode, which resulted in a staggering 162% increase in the average trade open-
ness from 1995 to 2012. This startling growth of trade is accompanied by a large
increase in the export variety. From 1995 to 2012, the average extensive margin of
exports in terms of product-market pairs grew by 94%. The data show the aver-
age relative standard deviation of per-capita GDP declined 38%. Therefore, the
data sample utilized in this study represents a natural experiment whereby we
can observe the actual impact of growth in trade openness on output volatility,
and determine the role of extensive margin growth in the volatility change.
The trade data I use are from the United Nations Comtrade database,
reported in thousand of US dollars, in 6-digit HS 1988/92 classification, cov-
ering the exports of the CEECs to 244 partner countries, in 5,030 unique product
codes, resulting in 388,352 unique product-market pairs, which I treat as a vari-
ety in this study. GDP per-capita data are obtained from the World Banks’s
23
World Development Indicators database, measured in 2011 purchasing power par-
ity (PPP) dollars for comparability. I convert nominal variables into real variables
by deflating each variable by the US real GDP deflator.
I measure the aggregate output volatility based on the rolling relative stan-
dard deviation of per-capita GDP over a three-year window. The main advantage
of this approach is that the rolling-volatility measure yields a longer time-series
dimension, which produces more reliable results. Additionally, by using the
relative standard deviation measure (standard deviation as a percentage of the
mean), I remove the scale effect and render the volatility measure comparable
across cross sections. On the other hand, the main disadvantage of measuring
output volatility this way is that, by definition, the measure is autocorrelated.
However, prior literature has used similar overlapping period measures, making
my results comparable to the previous work in the field.
My central variable of interest is the interaction between openness and
export diversification. I use the extensive margin measure developed by Hum-
mels and Klenow (2005) as my main diversification measure. The extensive
margin is based on the product-varieties-growth measure developed by Feen-
stra (1994), which states the growth in product variety of a single country over
time. However, instead of measuring variety growth over time, Hummels and
Klenow (2005) adopt a cross-sectional comparison approach. They construct the
extensive margin of exports relative to a reference country, so that one can com-
pare the two countries at a point in time. When the rest of the world is chosen
as the reference country, the extensive margin becomes a weighted count of the
varieties of country j relative to the varieties of the rest of the world. In the spirit
24
of Helpman et al. (2008), in my main specification, I define a variety as a product-
market pair. To be clearer, I consider a computer as a product, and a computer
exported to the United States as a variety. Specifically, the extensive margin of
country j’s exports to country m in a given year is defined as
EM
jm,t
=
å
i2I
jm
x
kmi,t
å
i2I
x
kmi,t
(1.1)
where x
kmi,t
is the value of country k’s (rest of the world) exports of product i to
country m at time t, I
jm
is the set of observable products in which country j has
positive exports to country m (i.e. x
jmi
> 0), and I is the set of all possible prod-
ucts in which the rest of the world has positive exports to country m. Therefore,
the numerator of this expression is the aggregate value of the rest of the world
exports to country m in the products in which country j exports to country m.
In the current sample, country j’s product set is always a subset of the products
exported by the rest of the world. The denominator, on the other hand, is the
aggregate value that the rest of the world exports to country m in the set of all
possible products. Hummels and Klenow (2005) summarize the extensive mar-
gin for varieties of country j’s exports in a given year by taking the geometric
average of EM
jm
across all of its export partners:
EM
vj,t
=
Õ
m2M
j
(EM
jm,t
)
w
jm,t
(1.2)
where v denote each variety, M
j
is the set of partner countries of country j and
w
jm
is the logarithmic mean of the shares of m in the total exports of j and the
25
rest of the world, respectively. w
jm
s are normalized to 1 so that å
M
j
w
jm
= 1. In
addition to the extensive margin for varieties, I further decompose EM
vj
into its
product and market components, and include them in the regressions one by
one. This approach gives me the opportunity to further analyze distinct effects
of each dimension (product and market margins) independently.
6
Note that the majority of the existing literature uses one or several of the con-
centration indices that are used in the income-distribution literature (Herfindahl,
Theil, and Gini indices). Although these indices give a relatively good idea about
the diversification level of exports, they all depend on the value of the country’s
own exports, which may cause an overestimation problem. Unlike the traditional
concentration measures, the extensive margin defined in equation (1.2) does not
depend on country j’s export volume; thus, it prevents the extensive margin from
being overestimated simply because country j exports a single good to a specific
partner in high volumes. Nevertheless, following the existing empirical literature
on trade diversification, I also use standard concentration indices for robustness
checks, including the Herfindahl-Hirschman index and the between component
of the Theil index. The correlation matrix for the diversification measures used
in this study can be found in the Appendix Table A.1.
My second variable of interest in the first part of my analysis is the trade
openness. Consistent with the existing literature, I compute openness as the ratio
of total trade (exports plus imports) to the GDP . The GDP data come from the
6
The extensive margin of country j’s exports for products in a given year is defined as the ratio
of rest-of-the world (ROW) exports in the product basket of country j to the ROW exports in all
possible products. The extensive margin for markets is defined in a similar way where it is the
ratio of ROW exports to the destination market basket of country j, to the ROW exports to all
possible markets. Specifically, EM
pj
=
å
i2I
j
x
ki
å
i2I
x
ki
and EM
mj
=
å
m2M
j
x
km
å
m2M
x
km
.
26
World Banks’s World Development Indicators database. In addition to the export
diversification and openness measure, I include several control variables, because
a number of factors determine aggregate volatility. Financial openness is mea-
sured by an index of restrictions on capital account transactions (Chinn and Ito,
2006). Both trade-openness and financial-openness variables reflect a country’s
exposure to the world markets. External volatility is captured by terms-of-trade
risk and exchange rate risk. The terms-of-trade is computed as the ratio of the
export-value index to the import-value index, for which the data are obtained
from the World Banks’s World Development Indicators database. Following Rodrik
(1998), I measure the terms-of-trade risk as the volatility of terms-of-trade (rel-
ative standard deviation of terms-of-trade) multiplied by openness. Exchange-
rate volatility is measured as the relative standard deviation of the real effective
exchange rate with 18 trading partners. The data for exchange rate are obtained
from Eurostat’s Economy and Finance database. Other control variables for the
aggregate volatility analysis include foreign direct investment volatility, inflation
volatility, government spending, population and GDP per capita (PPP). The sum-
mary statistics for the variables used in the aggregate analysis can be found in
the Appendix Table A.3 - Panel A.
In determining the role of export diversification on aggregate output volatil-
ity, I estimate a set of regressions of the following general form:
GDP Volatility
it
= b
0
+b
1
Openness
it
+
b
2
Openness
it
EMv
it
+
b
3
Openness
it
TOT Volatility
it
+
q X
it
+e
it
(1.3)
27
where i and t denote country and year, respectively. The dependent vari-
able is the ratio of the standard deviation of GDP per capita (PPP) to the
mean. Openness
it
is the ratio of total trade to GDP . Openness
it
EMv
it
is the
interaction between openness and export diversification, whereas Openness
it

TOT Volatility
it
is the interaction between openness and terms-of-trade volatil-
ity. X is the vector of control variables, and e
it
is the disturbance term. Control
variables include financial openness, foreign direct-investment volatility, infla-
tion volatility, real effective exchange-rate volatility, government spending, pop-
ulation and GDP per capita (PPP). All volatility measures are calculated as the
rolling relative standard deviations over a three year window (for the period
[t, t+ 2]), whereas the other explanatory variables are measured in t. Following
Rodrik (1998), I measure the terms-of-trade risk and the diversification effect by
using their interaction with openness.
7
A detailed description of the variables
and their sources can be found in the Appendix Table A.4 and Table A.5.
Endogeneity is a major concern in the analysis of volatility and openness. On
the one hand, openness can affect volatility through increased exposure to exter-
nal shocks. On the other hand, fluctuations in the GDP may drive policy makers
to choose policies affecting openness of countries, especially when openness is
considered as a source of volatility. A third mechanism may be such that in the
presence of large fluctuations in GDP , countries may attempt to increase their
exports in order to diversify their external risk through different products and
markets. Therefore, I employ a two-stage GMM instrumental variables regression
(IV-GMM) as the benchmark model. I use weighted average of yearly effective
7
Rodrik (1998) provides a formal justification for the use of interaction terms as a proxy.
28
tariff rates as instruments, along with lagged explanatory variables with up to
two lags.
8
Nonetheless, I use different estimation methods as robustness checks and for
comparison purposes. These methods include: OLS, the Prais-Winsten regres-
sion with panel-corrected standard errors (PW-PCSE), fixed (FE) and random
effects (GEE) regressions, and cross-sectional time-series FGLS regression (CS-
TS FGLS). Period dummies are included in all the regressions to account for
time fixed effects. Because of the construction of the dependent variable, resid-
uals follow a moving-average autoregressive process of order 2. In addition, the
data display cross-sectional dependence. Therefore, for the OLS and fixed-effects
models, I use clustered robust standard errors, which allow for panel-level het-
eroskedasticity and autocorrelation of an arbitrary form. On the other hand,
PW-PCSE, FGLS, random effects (population averaged), and IV regressions all
compute robust standard errors that allow for heteroskedasticity, autocorrelation,
and cross-sectional dependence.
As mentioned earlier, the recent literature argues the extensive margin of
trade plays an important role in increasing the aggregate volume of trade. An
increase in the extensive margin of exports will lead to higher export openness,
which in turn will directly affect the output volatility of the export sectors in
the economy. If Y denotes the aggregate GDP , S
T
and S
N
represent the outputs
of the traded and non-traded sectors in the economy, and given the aggregate
output of the economy equals the sum of the outputs of traded and non-traded
sectors, the variance of the aggregate output would be equal tos
2
Y
= s
2
S
T
+s
2
S
N
+
8
Tariff data are obtained from TRAINS database, through World Integrated Trade Solutions
(WITS).
29
2s
S
T
,S
N
, wheres
2
ands denote the variance and covariance, respectively. Hence,
if diversification through exports changes the effect of openness, it would mainly
do so through the export sectors in the economy. By studying only the aggregate
output volatility, one may fail to observe the full extent of the role of the export
diversification, because the covariance of the traded and non-traded sectors has
the potential to increase or decrease the observed effect of the extensive margin
on the volatility of production.
To observe the direct effect of export diversification on volatility of output,
I repeat my analysis with a second data set, which includes both export and
production data at the sectoral level. Using this data set, I examine the link
between sectoral output volatility and the diversification level of exports, along
with the export openness of each sector. The use of industry-level data set allows
me to isolate the covariance component of the aggregate volatility and determine
the immediate impact of any change in the extensive margin on volatility of the
production.
Eurostat’s Economy and Finance Database provides output data for 38 eco-
nomic activities, reported in NACE (classification of economic activities in the
European Community), revision 2. Along with the output data, Eurostat pro-
vides correspondence tables that allow conversion between a harmonized coding
system and NACE classification at the 6-digit product level.
9
Using these tables,
I first matched each product category of the disaggregated trade data to its cor-
responding NACE code at the 6-digit level. Then, I classified the exports of each
country depending on the 38 sectors defined in the output data. Note that 25 of
9
The correspondence tables between various HS coding systems and CPA and NACE classifica-
tions can be found at: http://ec.europa.eu/eurostat/ramon.
30
the 38 sectors turn out to be export sectors that appear in the trade data. How-
ever, the number of products exported within some sectors is substantially low.
Therefore, I combined these sectors at a higher aggregation level, and I ended
up with 20 export sectors for each country in my sample. I used this data set to
calculate output volatility, export openness, and export diversification measures
at the sectoral level. A complete list of the names and the NACE codes of the
sectors included in the data is presented in Appendix Table A.2.
Similar to the aggregate analysis, I measure sectoral output volatility based
on the rolling relative standard deviation of real sectoral output, over a three-year
window. Different from the aggregate study, instead of using overall openness,
I use export openness of each sector. My central variable of interest is the inter-
action between the export openness and export diversification. Along with the
openness and diversification measures, I include several sectoral control vari-
ables. To account for the external risk, I include terms-of-trade risk, which I
capture by multiplying aggregate terms-of-trade volatility by export openness of
each sector.
Because larger sectors are less likely to be volatile compared to the smaller
ones, I include sector size as a control variable. The sector size is measured as
the ratio of sectoral output to the GDP . Sectors that receive greater funding from
the government have greater insurance against shocks, so they are expected to
be less volatile (Rodrik, 1998). To account for the government funding, I include
total taxes paid by each sector as a control variable. The sectors that receive
government subvention pay negative taxes. To account for the sectoral produc-
tivity, I include two variables: output per worker and fixed capital consumption
per worker. All sectoral data are obtained from Eurostat’s Economy and Finance
31
Database, in current million euros. Euro values are first converted to dollars and
then converted into real values by the US GDP deflator. The summary statistics
for the variables used in the sectoral analysis can be found in the Appendix Table
A.3 - Panel B.
Using the sectoral data, I run a series of regressions to establish the effect of
export diversification on sectoral output volatility. The exact specification of my
empirical model for the sectoral analysis is as follows:
Output Volatility
ist
= b
0
+b
1
Export openness
ist
+
b
2
Export openness
ist
EMv
ist
+
b
3
Export openness
ist
TOT Volatility
it
+
q Z
ist
+e
ist
(1.4)
where i, s, and t denote country, sector, and year, respectively. The dependent
variable is the ratio of the standard deviation of real sectoral output to the mean.
Similar to before, it is measured as the rolling relative standard deviation over a
three year window. Export openness
ist
is the ratio of exports to the sectoral out-
put. The interaction of export openness with terms-of-trade volatility and export
diversification are included to capture the terms-of-trade risk and diversification
effect, whereas the variable Export openness
ist
captures the effect of openness
on output volatility through other channels. Z is the vector of control variables,
and e
ist
is the disturbance term. Control variables include sector share in GDP ,
labor productivity, fixed capital consumption per worker and taxes paid to the
government.
32
Because of the endogeneity concerns, I employ as the benchmark model an
IV regression that accounts for autocorrelation and arbitrary heteroskedasticity.
As instruments, I use tariff rates together with lagged independent variables, up
to two lags. The tariff data are obtained from the TRAINS database, at the 6-
digit level, HS-combined classification. I averaged the disaggregated data first
over product categories and second, over activity codes to attain the tariff rates
at the sectoral level. Along with IV , I estimate OLS, fixed- and random- effects
regressions that compute clustered standard errors that are robust to panel-level
autocorrelation and heteroskedasticity. Moreover, I run pooled OLS with Driscoll
and Kraay (1998) standard errors, which are robust to cross-sectional (spatial) and
temporal dependence as well as heteroskedasticity and autocorrelation.
In the sectoral model, I choose not to include measures of aggregate volatil-
ity, because country-level volatility measures do not change over sectors and
inclusion of them would only boost their significance artificially. To account for
the effect of macroeconomic volatility measures at the sectoral level, I include
sector-specific fixed effects as well as country and time fixed effects in all regres-
sions.
1.5 Empirical Results
To analyze the impact of export diversification on output volatility, I use 6-digit-
level bilateral trade data along with GDP per capita and sectoral output data.
Using these two data sets, I estimate the effect of the extensive margin of exports
on per-capita GDP volatility and sectoral output volatility based on equations
33
(1.3) and (1.4) and methodology discussed in section ??. Furthermore, to evalu-
ate the particular impact of trade liberalization on the effect of diversification, I
run regressions in which I interact the diversification effect with the EU dummy
variable. The results are presented in Table 1.8 through Table 1.12.
1.5.1 Aggregate Output Volatility and Extensive Margin of
Exports
The results of the benchmark, IV-GMM estimation are presented in the first col-
umn of Table 1.8. Columns (2) to (6) report the results of different estimation
techniques that account for the autocorrelation and cross-sectional dependence
problems. Overall, the results show a significant negative effect of export diver-
sification on per-capita GDP volatility. In the majority of the regressions, the
coefficient of the extensive margin and openness interaction is consistently nega-
tive and highly statistically significant.
The fit of the benchmark IV regression that accounts for the endogeneity
problem between openness and volatility is very good, with an adjusted R-
squared of 0.69, and the Hansen J test of overidentifying restrictions suggests
the instruments are valid. The estimates clearly support my claim that although
higher trade openness is associated with higher output volatility, higher diversifi-
cation of exports decreases this effect. The coefficient of openness is positive and
statistically significant at 5% level, whereas the extensive margin effect enters the
regression with a negative coefficient that is highly statistically significant, at the
1% level.
34
Table 1.8
Regression results for per-capita GDP volatility on export diversification - variety
(product-market pair)
(1) (2) (3) (4) (5) (6)
IV-GMM OLS PW-
PCSEs
FE RE-GEE FGLS
Openness 0.0301
**
0.0194 0.0428
**
0.0199 0.0213
*
0.0350
***
[0.0144] [0.0136] [0.0203] [0.0232] [0.0124] [0.0137]
Openness x -0.0594
***
-0.0418
**
-0.0574
*
-0.0240 -0.0531
**
-0.0688
***
EM-variety [0.0228] [0.0209] [0.0302] [0.0273] [0.0251] [0.0231]
Openness x 0.1579
***
0.1205
*
0.0936
*
0.0348 0.1745
**
0.0884
*
TOT volatility [0.0494] [0.0613] [0.0535] [0.0474] [0.0694] [0.0507]
Financial 0.0185
***
0.0108 -0.0078 -0.0141 0.0084 0.0109
openness [0.0071] [0.0072] [0.0115] [0.0145] [0.0078] [0.0085]
FDI volatility 0.0001
***
0.0001
***
0.0000 0.0000 0.0001
***
0.0001
[0.0000] [0.0000] [0.0001] [0.0001] [0.0000] [0.0001]
Inflation 0.0017
***
0.0019
***
0.0012 0.0017
*
0.0016
***
0.0012
volatility [0.0005] [0.0007] [0.0010] [0.0007] [0.0005] [0.0011]
Exchange rate -0.0307 -0.0255 -0.0566 -0.0605 -0.0537 -0.0272
volatility [0.0392] [0.0472] [0.0427] [0.0584] [0.0411] [0.0411]
Government -0.1118 -0.2216
*
0.0857 -0.0292 -0.0756 -0.0606
spending [0.1095] [0.1181] [0.1239] [0.2931] [0.1854] [0.1086]
Population (Log) -0.0012 -0.0045 -0.0872 -0.0490 -0.0040 -0.0011
[0.0035] [0.0032] [0.0755] [0.0810] [0.0028] [0.0036]
GDP/cap, PPP (Log) -0.0568
***
-0.0570
***
-0.0458 -0.0207 -0.0557
***
-0.0556
***
[0.0096] [0.0107] [0.0302] [0.0466] [0.0095] [0.0124]
Constant 0.6256
***
0.7113
***
1.8626 1.0182 0.6469
***
0.5858
***
[0.1441] [0.1451] [1.3528] [1.6351] [0.1525] [0.1565]
R
2
0.7529 0.7545 0.8910 0.5403
Adj-R
2
0.6891 0.6932 0.4254
Hansen J stat. 29.0200
K-P LM stat. 29.8947
K-P F stat. 22.6467
N 118 136 136 136 136 136
Notes: Models (2) and (4) report clustered robust standard errors, which account for panel-level heteroskedasticity and
autocorrelation of arbitrary form, in parentheses. Models (1), (3), (5), and (6) report robust standard errors that allow
for heteroskedasticity, autocorrelation, and cross-sectional dependence in parenthesis.
*
indicates significance at the 10%
level,
**
indicates significance at the 5% level, and
***
indicates significance at the 1% level. Time fixed effects are included
in all the regressions, but are not reported.
Furthermore, consistent with traditional theory, the terms-of-trade-risk has
a strong positive effect on output volatility. The terms-of-trade risk variable has
a positive coefficient that is highly significant (at the 1% level). Additionally,
35
as the previous empirical literature suggests, financial openness, foreign direct-
investment volatility and inflation volatility all appear to be significant factors
that increase output volatility, whereas GDP per capita has a negative and highly
significant coefficient, suggesting larger economies tend to be less volatile.
The result that export diversification is a significant factor in reducing volatil-
ity is robust to different estimation techniques. Throughout all regressions, the
extensive margin effect remains negative and statistically significant (except for
the fixed-effects model), confirming the hypothesis that greater diversification
of exports plays a significant role in decreasing per-capita GDP volatility. Sim-
ilarly, both the effect of openness and terms-of-trade risk remain positive and
significant in the majority of the regressions, attesting to the prior findings of the
trade literature that higher openness and terms-of-trade shocks are important
sources of aggregate volatility. Moreover, although they fall in and out of signifi-
cance, the coefficients of the control variables remain consistent in their signs all
through different estimation techniques. On the other hand, neither exchange-
rate risk nor population appears to be significant factors affecting aggregate out-
put volatility.
As mentioned earlier, in measuring the effect of terms-of-trade volatility and
export diversification, I prefer to include the interaction of these variables with
the openness measure without including the levels (terms-of-trade volatility and
EM
v
). Generally, exclusion of a component of the interaction term from the
regression may result in omitted-variable bias. However, Rodrik (1998) presents a
theoretical justification that, given that the terms-of-trade volatility affects output
volatility only through trade, the interaction term is the correct variable to con-
trol for the terms-of-trade risk. Similar logic is applicable to the extensive margin
36
measure. Because trade openness is the only channel that extensive margin can
affect the output volatility, I measure the effect of diversification only through the
interaction term. Additionally, both terms-of-trade volatility and extensive mar-
gin variables are highly correlated with their interaction terms with openness.
Specifically, the correlation coefficient is 0.82 for the terms-of-trade volatility and
its interaction, and it is 0.73 for the interaction term and the level of the export
diversification variable. Thus, for these two variables, including the levels along
with their interactions with openness causes a multicollinearity problem, and
when they are excluded, the model does not suffer from omitted-variable bias.
The export diversification measure used in the regressions presented in Table
1.8 is defined in terms of product-market pairs. To investigate the effect of each
dimension separately, I run a second set of regressions in which I include product
and market diversification measures one at a time. The results of these regres-
sions are presented in Table 1.9. Columns (1) to (3) report the results for the IV
estimator, whereas columns (4) to (6) report the results for the random-effects
model.
The fit of the IV regression is still good such that the model explains 76% of
the variation in aggregate volatility. The overidentification test suggests instru-
ments are valid for all three models. The estimates show that market diversifi-
cation clearly matters more in providing insurance against external shocks. The
coefficient of the market diversification effect has a negative sign and is highly
statistically significant in both (2) and (3), and the magnitude of its coefficient is
significantly high. On the other hand, the product diversification effect remains
insignificant in both (1) and (3).
37
Table 1.9
Regression results for per-capita GDP Volatility on product and market
diversification
IV-GMM RE-GEE
(1) (2) (3) (4) (5) (6)
Openness -0.0105 0.5923*** 0.6201*** 0.0106 0.0987 0.1220
[0.0205] [0.2223] [0.2297] [0.0102] [0.1342] [0.1347]
Openness x 0.0069 -0.0154 -0.0156 -0.0165
EM-product [0.0209] [0.0188] [0.0174] [0.0172]
Openness x -0.6075*** -0.6212*** -0.1046 -0.1135
EM-market [0.2265] [0.2282] [0.1331] [0.1369]
Openness x 0.2069*** 0.1588*** 0.1501*** 0.2109*** 0.2115*** 0.2103***
TOT volatility [0.0477] [0.0558] [0.0557] [0.0785] [0.0809] [0.0798]
Financial 0.0144** 0.0156** 0.0162** 0.0046 0.0043 0.0053
openness [0.0072] [0.0070] [0.0068] [0.0093] [0.0088] [0.0084]
FDI volatility 0.0001*** 0.0001*** 0.0001*** 0.0001*** 0.0001*** 0.0001***
[0.0000] [0.0000] [0.0000] [0.0000] [0.0000] [0.0000]
Inflation 0.0019*** 0.0020*** 0.0020*** 0.0016*** 0.0016*** 0.0017***
volatility [0.0005] [0.0005] [0.0005] [0.0005] [0.0005] [0.0005]
Exchange rate -0.0137 -0.0137 -0.0193 -0.0472 -0.0447 -0.0440
volatility [0.0404] [0.0385] [0.0394] [0.0407] [0.0402] [0.0415]
Government -0.2093** -0.3022*** -0.3140*** -0.1543 -0.1560 -0.1685
spending [0.1016] [0.0931] [0.0973] [0.1506] [0.1569] [0.1541]
Population (Log) -0.0090*** -0.0096*** -0.0091*** -0.0090*** -0.0094*** -0.0091***
[0.0020] [0.0017] [0.0018] [0.0019] [0.0021] [0.0019]
GDP/cap, PPP (Log) -0.0724*** -0.0653*** -0.0644*** -0.0703*** -0.0695*** -0.0680***
[0.0073] [0.0076] [0.0078] [0.0088] [0.0073] [0.0084]
Constant 0.9283*** 0.8903*** 0.8766*** 0.8886*** 0.8882*** 0.8703***
[0.0801] [0.0762] [0.0786] [0.1017] [0.0789] [0.0924]
R
2
0.7595 0.7619 0.7618
Adj-R
2
0.6975 0.7004 0.6970
Hansen J stat. 31.9128 29.9420 29.7637
K-P LM stat. 23.9210 16.1470 17.1551
K-P F stat. 9.8417 0.7296 0.7600
N 118 118 118 136 136 136
Notes: Robust standard errors that allow for heteroskedasticity, autocorrelation, and cross-sectional dependence are
reported in parentheses.
*
indicates significance at the 10% level,
**
indicates significance at the 5% level, and
***
indicates
significance at the 1% level. Time fixed effects are included in all the regressions, but are not reported.
Furthermore, estimates in column (3) confirm my previous findings that
trade openness is an important factor that increases output volatility. Openness
enters the regression with a positive and highly significant coefficient. Similarly,
38
terms-of-trade risk, financial openness, FDI volatility, and inflation volatility have
positive and highly significant coefficients, suggesting they are among the factors
that contribute to the aggregate volatility. Last, the coefficients of government
spending, population, and GDP per capita all have a negative sign and are sta-
tistically significant at the 1% level, which is in line with the theory that larger
economies are less vulnerable to shocks, and greater government spending has a
stabilizing effect on the economy.
The last three columns of Table 1.9 presents the estimates of the random-
effects model. Although all the variables have the same signs as in the IV regres-
sions, most of them fall in and out of significance. Mainly, neither any of the
diversification measures nor the openness variable have significant coefficients,
though they have the expected signs. Coefficients of terms-of-trade risk, FDI and
inflation volatility remain positive and highly significant. Moreover, the random-
effects model again confirms the size of the economy is an important stabilizing
factor in aggregate volatility. Both population and GDP per capita have negative
and highly significant coefficients.
In summary, I find strong evidence that export diversification plays an
important role in providing insurance against global shocks. Although greater
trade openness leaves a country more vulnerable, by diversifying its exports in
terms of product-market pairs, a country can considerably reduce (if not elim-
inate) the volatility of its aggregate output. On the other hand, diversifying
exports in terms of destination markets reduces the vulnerability of countries to
the terms-of-trade shocks and other foreign risk better than diversification of the
product basket of exports.
39
1.5.2 Sectoral Output Volatility and Extensive Margin of Exports
The regression results for equation (1.4) are reported in Table 1.10. Equation (1.4)
relates sectoral output volatility to export openness and the extensive margin
of exports, controlling for terms-of-trade volatility, sectoral share, sectoral pro-
ductivity and government funding. The estimates of the benchmark IV-GMM
regression are presented in column (1), whereas columns (2) to (5) report the
results for OLS, fixed-effects and random-effects models. In all the regressions, I
include country- and sector-specific fixed effects along with time fixed effects to
account for the aggregate volatility.
The results indicate there is a clear negative relationship between export
diversification and output volatility at the sectoral level, as well as the aggregate
level. In all the specifications, the interaction of the extensive margin with export
openness enters the regression with a negative and statistically significant coef-
ficient. Additionally, the effect is the strongest in the IV specification, when the
endogeneity of volatility and openness is taken into account. On the other hand,
the coefficient of the variable measuring export openness of a sector is consis-
tently positive and significant in most of the regressions, pointing to a positive
correlation between export openness and sectoral output volatility. Similarly, the
effect of export openness is the strongest when the IV regression is employed.
These results support my hypothesis that even though openness leaves coun-
tries exposed to external risk, through terms-of-trade shocks and other sources
of volatility, the effect of openness can be dampened if the exports are sufficiently
diversified in terms of product-market categories. The results presented in Table
1.10 provide evidence that this claim is valid at the sectoral level, as well as the
aggregate level, such that sectors that succeed in diversifying their exports in
40
terms of product-market pairs will be less affected by the destabilizing effect of
trade-liberalization episodes on sectoral output.
Table 1.10
Regression results for sectoral output volatility on sectoral export diversification -
variety (product-market pair)
(1) (2) (3) (4) (5)
IV-GMM OLS OLS-DK-
SEs
FE RE
Export openness 0.0272*** 0.0110** 0.0110 0.0070 0.0102**
[0.0048] [0.0055] [0.0073] [0.0052] [0.0047]
Export openness x -0.0303*** -0.0138** -0.0138* -0.0107* -0.0133***
EM-variety [0.0049] [0.0060] [0.0076] [0.0055] [0.0051]
Export openness x 0.0774 0.1373* 0.1373*** 0.1720** 0.1449**
TOT volatility [0.0869] [0.0751] [0.0503] [0.0692] [0.0724]
Sector share in GDP (Log) -0.0059 -0.0079 -0.0079 -0.0196 -0.0081
[0.0054] [0.0054] [0.0053] [0.0236] [0.0055]
Output per worker (Log) 0.0026 0.0028 0.0028 -0.0431** -0.0075
[0.0086] [0.0092] [0.0097] [0.0201] [0.0093]
Capital per worker (Log) -0.0143* -0.0213** -0.0213*** -0.0016 -0.0166*
[0.0085] [0.0100] [0.0081] [0.0148] [0.0103]
Taxes paid to -0.2196* -0.2094* -0.2094** -0.2788* -0.2305*
government [0.1200] [0.1267] [0.0903] [0.1483] [0.1270]
Constant 0.0476 0.0640 0.0336** 0.1792 0.0581*
[0.0365] [0.0395] [0.0147] [0.1317] [0.0339]
R
2
0.2826 0.3009 0.3009 0.1744 0.2992
Adj-R
2
0.2622 0.2843 0.1651
R
2
-within 0.1744 0.1630
R
2
-between 0.0322 0.5877
R
2
-overall 0.0894 0.2992
Hansen J stat. 14.5297
K-P LM stat. 30.1800
K-P F stat. 32.4526
N 1702 2162 2162 2162 2162
Notes: Models (2) and (4) report clustered robust standard errors, which account for panel-level heteroskedasticity and
autocorrelation of arbitrary form, in parentheses. Models (1), (3), and (5) report robust standard errors that allow for het-
eroskedasticity, autocorrelation, and cross-sectional dependence in parentheses.
*
indicates significance at the 10% level,
**
indicates significance at the 5% level, and
***
indicates significance at the 1% level. Time-, country- and sector- specific
fixed effects are included in all the regressions, but are not reported.
Furthermore, I include the interaction of export openness and the aggre-
gate terms-of-trade volatility in each regression to capture the effect of aggregate
price shocks on the sectoral output. The coefficient of the terms-of-trade risk is
41
positive and significant in most of the regressions, suggesting aggregate terms-
of-trade volatility is positively related to sectoral output volatility. Government
funding is another factor that has a significant impact on sectoral volatility. The
coefficient of the variable that measures the taxes that each sector pays is always
negative and statistically significant, indicating sectors that receive higher gov-
ernment funding are less volatile. Recent theoretical work also argues that the
level of productivity is a major factor in a firm’s export decision (Krugman, 1979,
Melitz, 2003). Firms with higher productivity prefer to enter the international
market, whereas firms with lower productivity remain domestic. To account for
the productivity, I include output per worker as a control variable. My results
point that sectors that have higher productivity tend to be more volatile. The
coefficient of labor productivity is always positive, except in one specification;
however, in most cases, it is not significant. Moreover, the size of the sector does
not have a significant impact on sectoral volatility. The coefficient of sector share
in aggregate output is never significant, though it is always negative. Finally, I
find evidence that sectors that are more capital intensive tend to be less volatile.
The capital per worker enters the regressions, almost always, with a negative and
significant coefficient.
1.5.3 European Union Agreement and the Impact of Export
Diversification
As discussed earlier, previous empirical studies on openness, diversification, and
output volatility make use of data sets - cross-sectional or panel data - that
include a large set of countries. The studies that use cross-sectional data sets
42
provide a basic understanding of the relationship between volatility and diversi-
fication, but they yield only a snapshot and lack the ability to explain this rela-
tionship depending on the evolution of the openness and export structure of the
economy over time. The studies using panel data sets that include a long time
dimension as well as a large set of countries provide a better explanation of this
relationship, but they remain too general of a sample to draw specific conclusions
on the relationship between the liberalization process and the structural change
in the exports and the output volatility. The selection of countries and the time
frame covered in this study provide a unique opportunity to conduct a natural
experiment on the relationship between openness, diversification, and volatility.
The CEECs joined the EU in 2004. Hence, the data used in this study cover nine
years before and nine years during a major trade-liberalization episode for the
CEECs. A tremendous increase in the trade of CEECs with the rest of the world,
as well as a significant expansion in their export variety, characterize this lib-
eralization episode. Furthermore, along with liberalization in trade, the CEECs
underwent significant structural change during and after their accession to the
EU, which further improved their trade performance in various ways. There-
fore, the data set employed in this study gives the opportunity to examine the
actual effect of the extensive margin on output volatility, in the economies that
experience an increase in their openness during trade liberalization.
To investigate the impact of trade liberalization on the effect of export diver-
sification on output volatility, I estimate a third set of regressions, using both
aggregate and sectoral data sets, in which I include the interaction of a dummy
variable that takes the value of 1 during the EU period and zero otherwise, and
43
Table 1.11
Regression results for per-capita GDP volatility on export diversification - variety
(product-market pair) and EU accession
(1) (2) (3)
IV-GMM RE PW-PCSEs
Openness 0.0050 0.0164 0.0343**
[0.0132] [0.0144] [0.0135]
Openness x 0.0867*** 0.0335 -0.0179
EM-variety [0.0260] [0.0451] [0.0318]
Openness x -0.4801*** -0.3696*** -0.2810***
TOT volatility [0.1068] [0.1012] [0.1077]
EU -0.0009 0.0133 0.0253*
[0.0105] [0.0105] [0.0133]
Openness x -0.0808*** -0.0565** -0.0468**
EM-variety x EU [0.0164] [0.0248] [0.0234]
Openness x 0.7072*** 0.5777*** 0.4884***
TOT volatility x EU [0.1127] [0.1245] [0.1202]
Government -0.3508*** -0.2086 -0.0475
spending [0.1058] [0.1924] [0.0979]
GDP/cap, PPP (Log) -0.0996*** -0.0804*** -0.0700***
[0.0094] [0.0154] [0.0110]
Population (Log) -0.0111*** -0.0075*** -0.0043
[0.0024] [0.0027] [0.0028]
FDI volatility 0.0001*** 0.0001** 0.0001
[0.0000] [0.0000] [0.0001]
Inflation 0.0018*** 0.0018*** 0.0015
volatility [0.0004] [0.0006] [0.0010]
Constant 1.2583*** 0.9740*** 0.7863***
[0.1247] [0.2177] [0.1369]
R
2
0.8046 0.7964 0.8408
Adj-R
2
0.7526
Hansen J stat. 19.6504
K-P LM stat. 32.4494
K-P F stat. 9.1441
N 120 142 142
Notes: Robust standard errors that allow for heteroskedasticity, autocorrelation, and cross-sectional dependence are
reported in parentheses.
*
indicates significance at the 10% level,
**
indicates significance at the 5% level, and
***
indicates
significance at the 1% level. Time fixed effects are included in all the regressions, but are not reported.
the diversification effect.
10
In addition, I also include an interaction of the terms-
of-trade risk and the EU dummy to measure the change in the terms-of-trade
10
Kiel and McClain (1995) use a similar technique to study the impact of a new garbage inciner-
ator on the housing values in North Andover, Massachusetts. In addition to different controls,
they include the interactions of period dummies and the distance to the incinerator, which
44
effect. I also include an EU dummy variable on itself, to control for the other
structural changes that affect the output volatility following the EU accession.
Table 1.11 presents the results of these regressions using the aggregate data
set. The estimates of the IV regression are reported in column (1), whereas
columns (2) and (3) report the results for the random-effects model and OLS
estimation. The results indicate export diversification has a significantly stabiliz-
ing effect on output volatility during liberalization episodes. The coefficient of
the interaction between the diversification effect and the EU dummy variable is
consistently negative and highly significant in all the regression models. It can
be seen that although the effect of diversification on aggregate volatility appears
to be ambiguous prior to the accession to the EU (the openness-extensive mar-
gin interaction term is only significant in the IV specification, and changes sign
and turns insignificant in the random-effects and OLS regressions), its impact
declines significantly and even becomes negative following the EU accession.
The results of different estimation techniques using the sectoral data are pre-
sented in Table 1.12. The export diversification effect on sectoral output volatil-
ity appears to be significantly positive prior to the EU accession. The export-
openness and diversification interaction term enters the regression with a pos-
itive and significant coefficient in all the models. However, the interaction of
the diversification effect and the EU dummy has a negative coefficient that is
highly statistically significant at the 1% level in all the different specifications.
This result suggest that even though higher diversification of exports leads to
higher volatility of sectoral output prior to the EU, this effect loses its strength
measure the change in the effect of distance on the housing prices between different periods of
implementation.
45
Table 1.12
Regression results for sectoral output volatility on sectoral export diversification -
variety (product-market pair) and EU accession
IV-GMM RE OLS
(1) (5) (3)
Export openness 0.0200*** 0.0014 0.0026
[0.0057] [0.0087] [0.0127]
Export openness x 0.0591* 0.0973*** 0.1001***
EM-variety [0.0302] [0.0263] [0.0274]
Export openness x 0.0226 -0.2577 -0.0702
TOT volatility [0.4534] [0.2968] [0.2400]
EU 0.0225** 0.0332*** 0.0333***
[0.0091] [0.0086] [0.0072]
Export openness x -0.0824*** -0.1013*** -0.1050***
EM-variety x EU [0.0288] [0.0284] [0.0228]
Export openness x 0.0898 0.4131 0.2150
TOT volatility x EU [0.4547] [0.3009] [0.2416]
Sector share in GDP (Log) -0.0091 -0.0183*** -0.0167***
[0.0067] [0.0053] [0.0051]
Output per worker (Log) 0.0095 0.0078 0.0145**
[0.0097] [0.0093] [0.0072]
Capital per worker (Log) -0.0269*** -0.0216*** -0.0244***
[0.0088] [0.0082] [0.0064]
Taxes paid to -0.2477** -0.1708 -0.1650*
government [0.1250] [0.1324] [0.0960]
Constant 0.0498 0.0207 0.0053
[0.0369] [0.0333] [0.0263]
R
2
0.2074 0.2134 0.2152
Adj-R
2
0.1902 0.2020
Hansen J stat. 56.4992
K-P LM stat. 39.9328
K-P F stat. 22.5803
N 1702 2162 2162
Notes: Models (1) and (2) report robust standard errors that allow for heteroskedasticity, autocorrelation, and cross-
sectional dependence in parentheses. Model (3) reports clustered robust standard errors, which account for panel-level
heteroskedasticity and autocorrelation of arbitrary form, in parentheses.
*
indicates significance at the 10% level,
**
indi-
cates significance at the 5% level, and
***
indicates significance at the 1% level. Country- and sector-specific fixed effects
are included in all the regressions, but are not reported.
and even becomes negative during the trade-liberalization period. The coeffi-
cient of the extensive margin effect is always greater (in absolute value) for the
46
post-EU period, which results in a negative overall effect of the extensive margin
during the period following the EU accession.
Overall, the results show the diversification of exports in terms of product-
market pairs, plays a particularly important role in decreasing the output volatil-
ity (both at the aggregate and sectoral levels) during the episodes of trade lib-
eralization and structural transformation. Both aggregate and sectoral analyses
show the effect of export diversification on output volatility becomes negative
during the period following the CEECs’ accession to the EU.
1.6 Robustness Checks
The existing empirical literature relies heavily on the traditional concentration
indices that are used in the income-distribution literature to measure the export
diversification. Even though these indices give a relatively good idea about the
diversification level of exports, they depend on the value of a country’s own
exports, which is argued to be somewhat problematic because it may cause over-
estimation of the extensive margin of exports (Hummels and Klenow, 2005).
In this study, I use the extensive margin measure developed by Hummels and
Klenow (2005) that measures the export diversification of a country by its weight
in the total world trade. Nonetheless, in testing the robustness of the above
results, I run additional regressions employing standard concentration indices.
These include the Herfindahl-Hirschman index (HI) and the Theil index. The
Herfindahl index is defined as
47
HI
jv
=
s
n
jv
å
v=1
(
x
jv
X
j
)
2

q
1
n
jv
1
q
1
n
jv
(1.5)
where v represents each variety (product-market pair), n
jv
is the total number
of varieties in which country j trades, x
jv
is the value of country j’s exports in
variety v, and X
j
=
n
jv
å
v=1
x
jv
is the total exports of j.
The Theil index is given by
T
jv
=
1
n
jv
n
jv
å
v=1
x
jv
m
j
ln(
x
jv
m
j
), where m
j
=
1
n
jv
n
jv
å
v=1
x
jv
(1.6)
The Theil index can be decomposed into within- and between-groups com-
ponents such that T
W
+ T
B
= T, where changes in the between-groups com-
ponent measure the proportional changes in the number of active export lines,
which is equivalent to the changes in the extensive margin. It can be easily shown
that
T
B
=
1
å
k=0
n
k
n
jv
m
k
m
j
ln(
m
k
m
j
) (1.7)
where k2f0, 1g, and n
0
and n
1
represent the number of inactive (exports = 0)
and active (exports> 0) export lines, respectively. Given the fact that m
0
= 0 and
n
0
is unobservable, we can show that
lim
m!0
T
B
= ln(
m
1
m
j
)= ln(
n
jv
n
1
)= ln(n
jv
) ln(n
1
) (1.8)
48
which is the log difference of the possible number of export lines and the actual
number of active lines in that specific year.
11
First, I employ HI-variety, HI-product, and HI-market measures in my
benchmark IV regression model of aggregate output volatility. Table 1.13 reports
the results of the benchmark IV regression of per-capita GDP volatility on concen-
tration indices of exports, along with trade openness and other control variables.
Note that both HI and the Theil index measure the concentration of exports.
Therefore, a positive coefficient on these indices points to a negative relationship
between export diversification and output volatility.
The results in column (1) show the extensive margin of exports in terms of
varieties continues to carry the correct sign and is highly statistically significant.
The results indicate a 10% increase in the variety concentration of exports results
in a 4.3% increase in per-capita GDP volatility, which is very close to the results
obtained from the preferred benchmark model from Table 1.8, where I found that
a 10% increase in the extensive margin of exports in terms of varieties results in
a 4.6% decline in output volatility.
The results of the regressions that include the effects of HI-product and HI-
market separately are presented in columns 2 to 4 of Table 1.13. The estimates in
column 4 indicate that when included together, both product concentration and
market concentration are significant factors in increasing output volatility. Both
variables have positive coefficients that are statistically significant at the 1% level.
On the other hand, different from my previous findings, the openness variable
11
See Cadot et al. (2011) for detailed derivations of equations 1.7 and 1.8. I take the distinct
number of product-market pairs over the entire sample period as the possible number of export
lines.
49
Table 1.13
IV Regression results for per-capita GDP volatility on export concentration -
Herfindahl index
(1) (2) (3) (4)
Openness -0.0292*** -0.0129 -0.0003 -0.1477***
[0.0097] [0.0084] [0.0204] [0.0352]
Openness x 0.3917***
HI-variety [0.0915]
Openness x 0.0585** 0.2314***
HI-product [0.0230] [0.0462]
Openness x -0.0077 0.3096***
HI-market [0.0520] [0.0795]
Openness x 0.1521*** 0.1667*** 0.2206*** 0.1316***
TOT volatility [0.0497] [0.0544] [0.0486] [0.0509]
Financial 0.0179** 0.0146** 0.0108 0.0215***
openness [0.0070] [0.0072] [0.0076] [0.0071]
FDI volatility 0.0001*** 0.0001*** 0.0001*** 0.0000
[0.0000] [0.0000] [0.0000] [0.0000]
Inflation 0.0014*** 0.0018*** 0.0018*** 0.0022***
volatility [0.0004] [0.0004] [0.0005] [0.0004]
Exchange rate -0.0033 -0.0140 -0.0082 -0.0365
volatility [0.0389] [0.0388] [0.0403] [0.0394]
Government -0.3166*** -0.2653*** -0.1607 -0.4405***
spending [0.1013] [0.1009] [0.1011] [0.1210]
Population (Log) -0.0059*** -0.0087*** -0.0092*** -0.0159***
[0.0018] [0.0018] [0.0025] [0.0025]
GDP/cap, PPP (Log) -0.0483*** -0.0653*** -0.0722*** -0.0712***
[0.0091] [0.0078] [0.0084] [0.0074]
Constant 0.6677*** 0.8694*** 0.9199*** 1.1226***
[0.0943] [0.0784] [0.1151] [0.1118]
R
2
0.7688 0.7708 0.7599 0.7811
Adj-R
2
0.7073 0.7098 0.6960 0.7199
Hansen J stat. 28.8180 30.0659 29.6056 32.3614
K-P LM stat. 35.3656 32.8338 29.1408 33.2825
K-P F stat. 25.2025 35.1186 17.7688 7.9952
N 120 120 120 120
Notes: Robust standard errors that allow for heteroskedasticity, autocorrelation, and cross-sectional dependence are
reported in parentheses.
*
indicates significance at the 10% level,
**
indicates significance at the 5% level, and
***
indicates
significance at the 1% level. Time fixed effects are included in all the regressions, but are not reported.
enters all the specifications with a negative coefficient, which is only significant
in models (1) and (4). This result is in line with Haddad et al. (2013), who uses HI
50
as the preferred concentration measure and they find trade openness coefficient
to be negative. Furthermore, the coefficients on the control variables tend to carry
the expected signs and they appear to be statistically significant. Whereas greater
terms-of-trade volatility, financial openness, and FDI and inflation volatility are
associated with greater output volatility, higher per-capita GDP , population, and
government spending are factors that decrease output volatility.
Table 1.14 presents the estimates for the random-effects regression. The
results are similar, though not that strong, for both concentration measures and
the control variables. Column (1) shows that HI in terms of product-market
pairs enters the regression with a positive and statistically significant coeffi-
cient, whereas control variables mostly have significant coefficients with expected
signs.
Second, I repeat a similar analysis with different concentration measures for
the sectoral data. Table 1.15 reports the results of the benchmark IV specifica-
tion of sectoral output volatility on HI-variety, HI-product, and HI-market, along
with sectoral export openness and other control variables. The results confirm a
clear negative relationship exists between export diversification and sectoral out-
put volatility. Column (1) shows that HI-variety has a positive and statistically
significant coefficient. Furthermore, the results in columns 2 to 4 indicate diver-
sification of exports in terms of new destination markets is a more important
factor in reducing sectoral output volatility than diversification in terms of prod-
ucts. Whereas HI-market has a positive and significant coefficient, HI-product
enters models (2) and (4) with a positive but insignificant coefficient.
51
Table 1.14
Random effects regression results for per-capita GDP volatility on export
concentration - Herfindahl index
(1) (2) (3) (4)
Openness -0.0243 -0.0131 -0.0022 -0.0718
[0.0189] [0.0129] [0.0228] [0.0441]
Openness x 0.2838*
HI-variety [0.1671]
Openness x 0.0660* 0.1326*
HI-product [0.0351] [0.0718]
Openness x 0.0090 0.1431
HI-market [0.0584] [0.0930]
Openness x 0.1427** 0.1399** 0.1545** 0.1397**
TOT volatility [0.0662] [0.0690] [0.0614] [0.0706]
Financial 0.0125 0.0118 0.0063 0.0136
openness [0.0098] [0.0096] [0.0118] [0.0115]
FDI volatility 0.0001** 0.0001*** 0.0001*** 0.0001
[0.0000] [0.0000] [0.0000] [0.0000]
Inflation 0.0016*** 0.0019*** 0.0019*** 0.0020***
volatility [0.0005] [0.0005] [0.0006] [0.0005]
Exchange rate -0.0231 -0.0115 -0.0215 -0.0342
volatility [0.0617] [0.0651] [0.0644] [0.0656]
Government -0.2960* -0.3008** -0.2751* -0.3598**
spending [0.1533] [0.1385] [0.1409] [0.1794]
Population (Log) -0.0064** -0.0078*** -0.0090*** -0.0111***
[0.0026] [0.0022] [0.0016] [0.0015]
GDP/cap, PPP (Log) -0.0517*** -0.0601*** -0.0713*** -0.0665***
[0.0152] [0.0109] [0.0117] [0.0123]
Constant 0.7017*** 0.8031*** 0.9223*** 0.9358***
[0.1443] [0.0943] [0.1204] [0.1307]
R
2
0.7625 0.7540 0.7468 0.7623
N 136 136 136 136
Notes: Robust standard errors that allow for heteroskedasticity, autocorrelation, and cross-sectional dependence are
reported in parentheses.
*
indicates significance at the 10% level,
**
indicates significance at the 5% level, and
***
indicates
significance at the 1% level. Time fixed effects are included in all the regressions, but are not reported.
Last, I repeat the same experiment using Theil index of concentration. Table
1.16 presents the results of the IV estimation of sectoral output volatility on vari-
ous Theil indices. Column (1) of Table 1.16 shows that concentration of exports in
52
Table 1.15
IV Regression results for sectoral output volatility on export concentration -
Herfindahl index
(1) (2) (3) (4)
Export openness -0.0720** -0.0262 -0.0592*** -0.0507*
[0.0316] [0.0204] [0.0089] [0.0262]
Export openness x 0.1571**
HI-variety [0.0702]
Export openness x 0.0226 0.0205
HI-product [0.0189] [0.0192]
Export openness x 0.1460*** 0.0689***
HI-market [0.0216] [0.0237]
Export openness x 0.5000*** 0.2719** 0.2583*** 0.2962**
TOT volatility [0.1858] [0.1149] [0.0805] [0.1247]
Sector share in GDP (Log) -0.0160*** -0.0126*** -0.0128*** -0.0122***
[0.0042] [0.0037] [0.0035] [0.0037]
Output per worker (Log) 0.0444*** 0.0324*** 0.0300*** 0.0340***
[0.0090] [0.0072] [0.0048] [0.0075]
Capital per worker (Log) -0.0286*** -0.0146** -0.0129*** -0.0150**
[0.0089] [0.0069] [0.0040] [0.0071]
Taxes paid to -0.1970** -0.1854** -0.2111** -0.1829**
government [0.0872] [0.0891] [0.0825] [0.0882]
Constant -0.0897*** -0.0689*** -0.0706*** -0.0728***
[0.0280] [0.0253] [0.0239] [0.0254]
R
2
0.1550 0.2386 0.1953 0.2298
Adj-R
2
0.1415 0.2259 0.1825 0.2165
Hansen J stat. 13.3462 22.7612 18.1263 22.3656
K-P LM stat. 62.2814 99.8265 19.3676 48.8499
K-P F stat. 4.2780 10.4810 0.8576 4.2924
N 1854 1770 1854 1770
Notes: Robust standard errors that allow for heteroskedasticity, autocorrelation, and cross-
sectional dependence are reported in parentheses.
*
indicates significance at the 10% level,
**
indi-
cates significance at the 5% level, and
***
indicates significance at the 1% level. Time-, country-
and sector-specific fixed effects are included in all the regressions, but are not reported.
terms of varieties remains a significant factor in increasing volatility. The Theil-
variety has a positive coefficient that is highly statistically significant. Moreover,
the remaining control variables are statistically significant and have the expected
signs. Whereas higher terms-of-trade volatility and labor productivity contribute
53
Table 1.16
IV Regression results for sectoral output volatility on export concentration - Theil
index
(1) (2) (3) (4)
Export openness -0.0410*** -0.0119 -0.0063*** -0.0200***
[0.0052] [0.0108] [0.0018] [0.0036]
Export openness x 0.0322***
Theil-variety [0.0041]
Export openness x 0.0125 0.0128***
Theil-product [0.0163] [0.0036]
Export openness x 0.0083*** 0.0177***
Theil-market [0.0021] [0.0019]
Export openness x 0.2126*** 0.3778** 0.1552** 0.2550***
TOT volatility [0.0755] [0.1787] [0.0768] [0.0765]
Sector share in GDP (Log) -0.0107*** -0.0137*** -0.0124*** -0.0125***
[0.0036] [0.0035] [0.0035] [0.0035]
Output per worker (Log) 0.0290*** 0.0307*** 0.0274*** 0.0314***
[0.0048] [0.0050] [0.0048] [0.0047]
Capital per worker (Log) -0.0126*** -0.0132*** -0.0112*** -0.0146***
[0.0040] [0.0043] [0.0040] [0.0039]
Taxes paid to -0.2258*** -0.2624*** -0.2184*** -0.2418***
government [0.0834] [0.0832] [0.0843] [0.0834]
Constant -0.0565** -0.0764*** -0.0637*** -0.0704***
[0.0247] [0.0245] [0.0242] [0.0240]
R
2
0.2489 0.2257 0.2420 0.2481
Adj-R
2
0.2369 0.2134 0.2300 0.2358
Hansen J stat. 27.5838 37.3413 23.8737 34.2046
K-P LM stat. 19.7149 24.7230 6.4953 63.2025
K-P F stat. 21.3765 3.8949 15.6378 12.5030
N 1856 1856 1856 1856
Notes: Robust standard errors that allow for heteroskedasticity, autocorrelation, and cross-sectional dependence are
reported in parentheses.
*
indicates significance at the 10% level,
**
indicates significance at the 5% level, and
***
indicates
significance at the 1% level. Time-, country- and sector-specific fixed effects are included in all the regressions, but are
not reported.
to greater sectoral volatility, sector share in GDP , capital per worker and gov-
ernment funding are the factors that reduce sectoral volatility. These results
prove that my prior findings, suggesting that export diversification reduces out-
put volatility both at the aggregate and sectoral levels, do not depend on the
choice of diversification measure.
54
1.7 Conclusion
A commonly held belief is that greater openness to trade, despite its benefits
in terms of output growth, increases GDP volatility by exposing countries to
external shocks. This belief stems from the assumption that higher openness
leads to higher specialization, which leaves countries vulnerable to terms-of-
trade shocks. However, this line of thought ignores the effects of trade liber-
alization and increased openness on the export structure of a country. That is,
while traditional comparative advantage theory predicts that a liberalizing coun-
try may specialize in a narrower range of products, the lowering of trade costs
accompanying trade liberalization often leads to a greater variety in products
being exported to a larger set of countries. This increase in trade variety may
help decrease the risks of greater trade openness.
In this paper, using two separate data sets, I test the impact of export diver-
sification on both per-capita GDP volatility and sectoral output volatility. For
my analysis, I match highly disaggregated trade data with per-capita GDP data
and sectoral production data of eight Central and Eastern European countries
for the period 1995 to 2012. I construct the extensive margin of exports, both at
the aggregate and sectoral levels, that measures the diversification of exports by
their weight in total world trade. I present new empirical evidence suggesting
that although higher openness has a positive effect on output volatility, this effect
can be stabilized through sufficiently diversified export baskets in terms of both
products and new destination markets.
My results show that product-market diversification of exports has a signif-
icantly negative effect on per-capita GDP volatility. I find that, after accounting
55
for the likely endogeneity of the trade openness, diversification of exports in
terms of product-market pairs plays an important role in providing insurance
against both product- and market- specific shocks, and thus reducing per-capita
GDP volatility. This relationship holds for the industry-level data as well. The
findings suggest that greater diversification of exports results in lower sectoral
output volatility. Moreover, the results suggest market diversification plays a
more important role in providing insurance against the negative effects of exter-
nal shocks, whereas diversification of the product basket of exports does not
have a significant impact. Furthermore, I find export diversification has a par-
ticularly stabilizing effect on both aggregate and sectoral volatility during trade-
liberalization episodes. Although the effect of the extensive margin of exports
on output volatility is always negative following a major trade-liberalization
episode, the results prior to the liberalization remain mixed. These findings are
robust to different estimation techniques and choice of diversification measures.
Along with the benchmark IV estimation, models such as OLS with Driscoll and
Kraay standard errors, random and fixed effects, and FGLS and Prais-Winsten
regression yield similar results pointing to a negative relationship between the
extensive margin of exports and volatility of output. Last, I experiment with
traditional measures of concentration including the Herfindahl-Hirschman index
and the Theil index, which support my prior findings.
The results presented here provide significant evidence that greater open-
ness does not necessarily bring about greater volatility, as the traditional view
suggests. My empirical findings suggest trade-liberalization episodes that result
in a growth in the extensive margin of exports can lead to lower output volatility.
Any policy advice that ignores the structural change in the export basket of an
56
economy would be ill suited to evaluating the effects of liberalization on output
volatility.
57
Chapter 2
The Role of Trade Liberalization on the
Export Variety of Central and Eastern
European Countries
2.1 Introduction
The effect of trade-cost reductions on the aggregate bilateral trade flows is a
widely researched topic in international trade. Numerous models with variable
trade costs are developed and used to estimate the changes in the trade pat-
terns through gravity equations (Anderson, 1979, Baier and Bergstrand, 2001,
Bergstrand, 1985, 1989, Lim˜ ao and Venables, 2001). Although the importance of
product variety in trade is well recognized, since the seminal work of Krugman
(1979), the international trade literature only recently focused on the changes
in the trade patterns stemming from the increased variety of exports following
trade liberalization.
Recently developed models with monopolistic competition, where firms face
market penetration costs (Arkolakis et al., 2008) or fixed trade costs (Anderson
and Wincoop, 2004, Helpman et al., 2008, Melitz, 2003) helped researchers to
explain the impact of the variations in extensive and intensive margins of trade
on the increased aggregate trade volumes. The extensive margin growth of trade
58
refers to the growth in the range of goods that a country trades, while the inten-
sive margin growth refers to the growth in the trade of the goods that have been
previously traded.
The standard monopolistic competition model assumes that all of the pro-
duced goods are exported. Therefore, the growth of total trade solely depends on
the intensive margin. Models, in which firms face some kind of fixed trade costs,
predict that only a subset of the firms are able to export. Relatively lower pro-
ductivity firms, that cannot cover the sunk costs, produce only for the domestic
market. This finding implies that a decline in fixed trade costs will result in more
firms entering foreign markets, which will result in an increase in the extensive
margin of trade. Hence, follows my central question. How big is this effect of
reductions in trade costs on the extensive and intensive margins of trade? How
important is the extensive margin growth in terms of aggregate trade growth?
Even though there is an ample amount of empirical literature investigating
the effects of North American Free Trade Agreement (NAFTA), Canada-U.S. Free
Trade Agreement (FTA) and other free trade agreements on the product variety,
relatively little work has been done concerning the effect of European Union
(EU) membership. Based on the assumption that EU accession reduces fixed
trade costs, I attempt to answer these questions by studying exports of eight
Central and Eastern European Countries (CEECs) that joined the EU in 2004.
1
Using disaggregated data at the 6-digit level of the Harmonized System (HS)
Classification (revision 1988-92) for the period 1995-2012, I investigate the role of
the EU agreement on the export variety of the CEECs. To determine the changes
1
My sample includes Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Slovak
Republic and Slovenia
59
in the trade patterns of these countries following their accession to the EU, I
begin by decomposing exports into their extensive and intensive margins, using
different measures developed in the recent literature. I compare the average
values of the two margins of exports for the pre- and post-EU periods and find
that there is a significant correlation between the growth of total exports and the
growth in the export variety of CEECs, following their accession to the EU.
This paper makes three main contributions to the existing literature. First, I
show that the significant growth in total exports following EU accession mainly
comes from the growth in the extensive margin, rather than the intensive margin.
I find that while extensive margin accounts for 80% of the relative export growth,
intensive margin accounts for only 20%, suggesting that EU accession reduces the
fixed rather than variable trade costs. Broda and Weinstein (2006) find significant
increases in productivity resulting from growth in product variety. Therefore,
EU membership has potentially large welfare effects through trade facilitation.
Second, my decomposition allows me to understand the mechanism behind the
extensive margin growth. While in most studies a variety is defined either a
good-category, a market-category or a market-good category, I study the exten-
sive margin in terms of both definitions. My results show that the growth in
the extensive margin mainly comes from penetration of new markets rather than
enlargement of the product range. Third, unlike many other studies, I provide
cleaner analysis because the sample period I consider do not suffer from the data
discontinuity problems associated with the adoption of the HS in the late 1980s.
I find that after joining the EU, CEECs experienced on average 212.2%
growth in their total exports. A simple count of goods and destination mar-
kets of these exports reveals that the average number of goods increased by 6.3%,
60
while the average number of markets they export to increased by 70.6%. Com-
bined, I find a 77.9% increase in the number of good-market pairs. In order to
see the evolution of export sectors, I do the same analysis at higher aggregation
levels. I find that the growth in the average number of goods exported drops by
half to 3% at the 4-digit level and to zero at 2-digit level, suggesting that at the
more aggregated level, export patterns are not affected significantly by the EU
accession.
I also look at the evolution of the share of the least traded goods based
on their average export value between 1995-1997. I see that between 1995 and
2012, this share increases from 10% to 41% in 2004, and to 46.2% in 2012, which
provides evidence that the new goods margin grows significantly over the period
1995-2012. An interesting point is that the fraction of the 10% least traded goods
of 1995, reaches to at least 20% by 1999, showing that the effect of EU starts five
years ahead of the actual beginning of the EU agreement.
Next, I construct weighted measures of the extensive and intensive margins
and find that following the EU accession, CEECs experienced a 42.4% growth
of the extensive margin, while the intensive margin growth remained at 18.9%.
This results points out the importance of the extensive margin in total export
growth following a trade liberalization. Furthermore, I find that the growth in the
extensive margin depends largely upon the growth in the number of destination
markets. When the extensive margin is calculated by using a 6-digit HS category
as a variety instead of a good-market pair, the growth in the extensive margin
drops down to 3.25%, which provides evidence that models that account for both
new goods margin and new markets margin are needed to evaluate the effects of
FTAs on international trade flows.
61
To better document the effect of EU accession on the extensive margin of
exports, I estimate an augmented version of the standard gravity equation, with
the dependent variable being log of the extensive margin. Controlling for as
many natural factors affecting trade as possible through the standard gravity
variables, I include a dummy variable which is unity for 2004 and afterwards.
The estimated coefficients on the EU accession dummy further solidifies my
hypothesis that joining the EU has a significant and large effect on the exten-
sive margin. My estimates show that the increase in the extensive margin is
around 55% following the EU accession.
After establishing the relationship between the EU Agreement and extensive
margin growth, I try to determine the importance of each margin in the growth
of the relative exports with respect to the rest of the world. I decompose relative
export growth into its extensive and intensive margin components. On average,
the relative exports of CEECs grow by 61.3%, where the extensive margin con-
tributes 80.5% and the intensive margin contributes 19.5% to this growth. From a
policy point of view, this result suggests that to improve the export performance,
looking for the ways to generate growth in the extensive margin yields better
results than simply boosting the exports of the previously traded goods.
Furthermore, I find that when the exports to EU and non-EU partners are
taken separately, the contribution of extensive margin to export growth is larger
for non-EU partners. On average, following EU membership the relative exports
to EU member states increase by 84%, where extensive margin accounts for
59.3%. On the other hand, 66.6% of the 70% increase in relative exports to non-EU
countries come from the extensive margin growth. Therefore, I conclude that the
growth in extensive margin following trade liberalization not only boosts trade
62
flows between member countries, but also it is a major factor in improving the
export performance to the rest of the world.
Overall, I find strong evidence linking trade liberalization through EU acces-
sion with growth in the extensive margin. The spike in the total exports of the
CEECs following their accession to EU is accompanied by a large increase in
the extensive margin. Moreover, I show that the extensive margin growth is not
only a contributing factor, but also the driving force behind the export growth
experienced by these countries.
2.2 Related Literature
The findings of this study fall in line with the previous empirical work of others.
Hummels and Klenow (2005) decompose exports and imports of a large set of
countries into extensive and intensive margins. Based on a cross-section com-
parison, they conclude that the extensive margin accounts for two-thirds of the
greater exports of the larger economies and one-third of the greater imports of
the larger economies.
Broda and Weinstein (2006) study the number of varieties in U.S. imports
for the period 1972-2001. Though their focus is on the effects of extensive mar-
gin growth on aggregate price indices, they find that during the implementa-
tion period of NAFTA, the average number of goods imported by U.S. increased
12.5%, while the average number of exporters increased by 32%.
Evenett and Venables (2002) study export growth of a group of developing
countries and decompose their trading patterns. They measure the extensive
63
margin in terms of exporting existing products to new destination markets and
find that about a third of the export growth is in exports of long-standing exporta-
bles to new markets.
In a similar study, Debaere and Mostashari (2010) focus on the changes in
the extensive margin of U.S. imports between 1989 and 1999, and find substan-
tial extensive margin growth across most countries. However, the role of tariff
changes in the extensive margin growth is found to be statistically significant,
but quantitatively moderate.
Kehoe and Ruhl (2013) study NAFTA along with the Canada-U.S. FTA
and find that the extensive margin grows significantly following a decrease in
trade barriers. Their results show that the share of the least traded goods that
account for 10% of the exports, increase up to 30% during the trade liberalization
episodes.
In a recent study, Bergin and Lin (2012) investigate the dynamic effects of the
European Monetary Union (EMU) on trade and find similar results that the effect
become significant starting four years prior to EMU. Magee (2008) estimates the
effects of regional trade agreements on trade flows and finds significant antici-
patory effects of regional agreements, going back to four years leading up the
beginning of the agreement.
Using Colombian customs data, Eaton et al. (2007) track firms’ entry and exit
into and out of destination markets. They find that although new exporters ship
very low volumes, the surviving firms increase their foreign sales very quickly,
and gradually expand into additional destinations. Their results imply that the
64
growth in the extensive margin (even if it is moderate) have amplifying effects
on the intensive margin in the long-run.
In a more recent study, Buono and Lalanne (2012) investigate the effect of
the Uruguay round on the two margins of trade. Using data on French manu-
facturing firms, they estimate gravity equations, and find that the effect of tariff
reductions on the extensive margin disappears, when the panel dimension of
the data is taken into account. Of the 3% of the increase in exports, only 0.5%
comes from the extensive margin growth. Similar results are found by Hanson
and Xiang (2011). They use U.S. movie exports data to measure the effect of the
global and bilateral fixed trade costs on the two margins of exports. They find
little variation in the number of countries U.S. export, but large variation in the
box-office revenues per movie.
Dalton (2013) uses the methodology developed in Kehoe and Ruhl (2013)
to measure the change in the new goods margin of trade between Austria and
the ten expansion countries of the EU 2004 enlargement. He finds that on aver-
age, the new goods account for 42% of the bilateral trade flow after enlargement.
Using the same methodology, Mukerji (2009) studies India’s unilateral trade lib-
eralization and finds that the bottom 10% of the total exports and imports account
for 26.5% and 33.8%, respectively, following the liberalization.
2.3 Data and Methodology
I study annual export values by good for the CEECs that joined the European
Union in 2004. To be specific, my sample includes Czech Republic, Estonia,
65
Hungary, Latvia, Lithuania, Poland, Slovak Republic and Slovenia. Addition-
ally, I include Turkey, who is not a member of European Union, but joined the
European Customs Union in 1996, in my analysis as a placebo.
The export data I use are from the United Nation’s COMTRADE database.
The data are reported in the Harmonized System (revision 1988-92) 6-digit level,
covering 248 partner countries and the period 1995-2012, which includes 9 years
before and 8 years after the year of accession. The length of my sample period
allows me better identify the effects of trade liberalization on the extensive mar-
gin.
Population, land area and GDP data have been obtained from the World
Banks’s World Development Indicators database. Data on the traditional gravity
variables are obtained from the CEPII GEODIST database (www.cepii.fr).
The notion of extensive margin was first introduced to international trade
theory by Krugman (Krugman, 1979, 1980). The prominent monopolistic com-
petition model of Krugman endogenizes the number of varieties a country pro-
duces. Combined with the taste for variety in consumer preferences, the model
predicts that the number of varieties produced in a country is proportional to
the country’s size. However, the Krugman model generates an extensive margin
through fixed costs of production. Therefore, it assumes that only the firms with
higher productivity levels will be able to produce, and all the firms that produce
will also export to all markets. Thus, although it takes into account the number
of varieties exported, it does not explain the variation in the destination markets,
that I observe in the data.
66
The empirical evidence shows us that not all the firms that have positive
production export. Most of the time, only a subset of the existing firms engage
in exporting activities. Besides, given that a firm exports, the destination mar-
kets are only a subset of all the markets with a demand for the variety produced
by the specific firm. This fact points out that additional to the fixed production
costs, firms may also face fixed export costs. Melitz (2003) develops a model that
incorporates fixed export costs faced by the firms into the Krugman’s monopolis-
tic competition model. Helpman et al. (2008) extends Melitz (2003) by allowing
the profitability of exports to vary across different destination markets, which
results in an estimation that matches the data closer.
Following the theoretical framework set up by Helpman et al. (2008), I ask:
when opened up to free trade, do countries export the same set of goods in high
volumes, or do they export a larger set of goods? What is the effect of trade
liberalization on the number of destination markets? To answer these questions,
I decompose the exports of each country in my sample into the extensive and
intensive margins.
In order to characterize the extensive margin, one first needs to define a vari-
ety. Empirical studies define a variety in two ways, either a good category or a
market-good category. A good category is defined as a 6-digit commodity, while
a market-good category is defined as a 6-digit commodity exported to a particu-
lar country. In spirit of Helpman et al. (2008), in my main specification, I define
a variety as a market-good category. To be more clear, I consider an automobile
as a good, while an automobile exported to U.S. as a variety. However, I also
compare the extensive margin growth rates with respect to different definitions
of variety. This way, I expect to test the hypothesis that the countries experience
67
a growth in extensive margin, not only export a larger set of goods to the same
markets (variety as a 6-digit commodity), but also they export same goods to
different markets (variety as a good-market category).
My decomposition methodology is based on the product varieties growth
measure developed by Feenstra (1994), which states growth in product variety
of a country over time. However, instead of measuring the variety growth over
time, I adopt a cross sectional comparison approach. Following Hummels and
Klenow (2005), I construct the extensive and intensive margins of exports relative
to a reference country (I choose rest of the world as the reference country) so that
I can compare the two countries at a point in time.
An alternative way of calculating the extensive margin would be the simple
count of the varieties, where the change in the extensive margin is measured by
counting the distinct number of good-market categories for every year. Given the
fact that every variety exported does not have the same importance in a nation’s
exports, the unweighted count tends to overestimate growth in the extensive
margin. On the other hand, the weighted count measure allows varieties to be
traded in different prices and quantities, thus it gives more reliable results.
The extensive margin of a country’s exports in a given year is defined as:
EM
j
=
å
k6=j
å
v2I
j
x
kv
å
k6=j
å
v2I
x
kv
(2.1)
where x
kv
is the value of country k’s exports of variety v and I
j
is the set of
observable varieties in which country j has positive exports, i.e. x
jv
> 0. There-
fore, the numerator of this expression is the aggregate value of the rest of the
68
world exports in the varieties in which country j exports. In my sample, country
j’s varieties is always a subset of the varieties exported by the rest of the world.
The denominator is the aggregate value of the rest of the world exports in the set
of all possible varieties (I). I have 5,038 unique 6-digit HS product codes and 248
partner countries in the sample. On average I have 710,318 different market-good
categories per year.
Equation (2.1) can be rewritten as:
EM
j
=
å
v2I
j
X
row
v
X
row
(2.2)
where X
row
is the value of aggregate exports of the rest of the world.
The extensive margin can be regarded as a weighted count of the varieties of
country j relative to the varieties of the rest of the world (I). It should be noted
that the extensive margin of country j is measured without taking into account
the value of its exports, but only its variety set. This approach prevents the
extensive margin being overestimated simply because country j exports a single
good to a specific partner in high volumes (Hummels and Klenow (2005)).
The corresponding intensive margin of exports is defined as:
IM
j
=
å
v2I
j
x
jv
å
k6=j
å
v2I
j
x
kv
=
X
j
å
v2I
j
X
row
v
(2.3)
69
where X
j
is the value of aggregate exports of country j. The intensive margin
measures relative exports of country j with respect to the rest of the world in
those varieties j exports. Note that the product of the two margins is:
EM
j
IM
j
=
å
v2I
j
x
jv
å
k6=j
å
v2I
x
kv
=
X
j
X
row
(2.4)
which equals aggregate exports of country j as a fraction of aggregate exports of
the rest of the world, which I call relative exports of country j. I will decompose
the growth of the relative exports into its components using equation (2.4).
2.4 Total Export Growth and the Extensive Margin
2.4.1 Unweighted Measures of Export Variety
Many studies show that reduction in trade costs and elimination of trade barri-
ers through FTAs have a significant positive impact on a country’s trade volume
(Anderson and Wincoop, 2004, Baier and Bergstrand, 2001, Bergstrand, 1985,
Lim˜ ao and Venables, 2001). However, one may ask: do countries that expe-
rience large increases in their total trade after joining an FTA also experience
large increases in their export variety? In order to answer this question, I start
with some descriptive evidence by presenting the unweighted decomposition
of exports into extensive and intensive margins. Table 2.1 reports the sample
70
statistics for each country in my dataset, based on the HS of 6-digit commod-
ity classification codes. In order to make comparisons across time, I divide the
sample into pre- (1995-2003) and post-EU (2004-2012) periods.
Table 2.1
Summary Statistics - Export Variety: Unweighted Count, Product: HS-6 level
Country Period Num. of
Goods
Num. of
Partners
Num. of
Varieties
Export
Growth
Czech Rep. 1995-2003 4625 32 72917 [211.1]
2004-2012 4466 46 102141
[-3.4] [45.6] [40.1]
Estonia 1995-2003 3544 12 21898 [176.5]
2004-2012 3550 16 26750
[0.2] [31.1] [22.2]
Hungary 1995-2003 3274 22 41719 [171.7]
2004-2012 3534 32 53512
[7.9] [41.7] [28.3]
Latvia 1995-2003 3069 8 12263 [234.4]
2004-2012 3495 18 27196
[13.9] [116.9] [121.8]
Lithuania 1995-2003 3673 11 21735 [265.1]
2004-2012 3897 20 38558
[6.1] [77.9] [77.4]
Poland 1995-2003 3100 15 22088 [249.1]
2004-2012 4460 45 115844
[43.9] [209.6] [424.5]
Slovak Rep. 1995-2003 3849 17 30835 [282.3]
2004-2012 3487 26 43710
[-9.4] [58.8] [41.8]
Slovenia 1995-2003 3984 19 35766 [107.0]
2004-2012 4069 28 53537
[2.1] [46.1] [49.7]
Turkey 1995-2003 4347 39 85035 [200.9]
2004-2012 4312 60 143183
[-0.8] [54.6] [68.4]
Notes: Table reports the average number of products, partners and varieties over the pre- (1995-2003) and post-EU (2004-
2012) periods. Each HS-6 category defines a product and each product-partner combination is defined as a variety. The
percentage growth rates between pre- and post-EU periods are given in brackets. Numbers are based on the author’s
calculations from UN COMTRADE data.
I also include Turkey in my analysis as a placebo. Though Turkey is not a
member of the EU, the CU agreement between Turkey and the EU was finalized
71
in January 1996. By 1995, Turkey had already taken all the necessary steps to
eliminate its custom duties, quantitative restrictions and other restrictive mea-
sures on its trade with EU members. Besides, Turkey started to implement trade
liberalization policies in the 1980s. By 2004, it had already been a member of
the CU for eight years, and had been member of several FTAs with the rest of
the world. Therefore, compared to the 2004 joiners in my sample, I expect to
see weaker trends in Turkey’s export composition between the periods pre- and
post-2004.
The values reported in Table 2.1 are acquired by first, calculating each statis-
tic for every year, then averaging over the two time periods. The growth rates of
average values between pre- and post-EU periods are given in squared brackets.
Table 2.2 and Table 2.3 repeats the exercise for higher aggregation levels i.e. HS
of 4-digit and 2-digit commodity classification codes, respectively. Overall sam-
ple statistics are presented in Table 2.4. The values in Table 2.4 are calculated
by taking the average of each statistic for the two time frames, over the sample
countries except Turkey.
The third, fourth and fifth columns of Table 2.1 present a crude measure of
variety in the exports of 2004 joiners of EU, for the pre- and post- EU periods
at the 6-digit level. The last column reports the growth rate of average exports
between the pre- and post-2004 periods per country.
These data reveal that the increase in the average value of exports ranges
from 107% to 282% between 1995-2003 and 2004-2012. Slovak Republic ranks
the first with an increase of 282.3%. It is followed by Lithuania and Poland by
265.1% and 249.1% increase, respectively. Slovenia ranks the last with an increase
72
of 107%. Overall, the average value of exports increased by 212.2% from 1995-
2003 to 2004-2012 for the CEECs that joined EU in 2004.
Table 2.2
Summary Statistics - Export Variety: Unweighted Count, Product: HS-4 level
Country Period Num. of
Goods
Num. of
Partners
Num. of
Varieties
Export
Growth
Czech Rep. 1995-2003 1213 46 33112 [211.1]
2004-2012 1195 63 44614
[-1.5] [35.0] [34.7]
Estonia 1995-2003 1066 19 11294 [176.5]
2004-2012 1074 25 13818
[0.8] [32.1] [22.3]
Hungary 1995-2003 982 33 20637 [171.7]
2004-2012 1036 44 25752
[5.5] [33.7] [24.8]
Latvia 1995-2003 991 13 6814 [234.4]
2004-2012 1051 29 14554
[6.1] [121.5] [113.6]
Lithuania 1995-2003 1085 18 11070 [265.1]
2004-2012 1119 31 18840
[3.1] [72.1] [70.2]
Poland 1995-2003 1019 22 12213 [249.1]
2004-2012 1199 62 49253
[17.7] [181.0] [303.3]
Slovak Rep. 1995-2003 1116 25 15026 [282.3]
2004-2012 1056 39 21425
[-5.4] [52.2] [42.6]
Slovenia 1995-2003 1129 29 16734 [107.0]
2004-2012 1135 41 24243
[0.5] [40.6] [44.9]
Turkey 1995-2003 1188 55 37730 [200.9]
2004-2012 1182 78 58491
[-0.5] [43.7] [55.0]
Notes: Table reports the average number of products, partners and varieties over the pre- (1995-2003) and post-EU (2004-
2012) periods. Each HS-4 category defines a product and each product-partner combination is defined as a variety. The
percentage growth rates between pre- and post-EU periods are given in squared brackets. Numbers are based on the
author’s calculations from UN COMTRADE data.
The third column of Table 2.1 reports the average number of goods exported
in a given year. It can be seen that there is no common trend among the sam-
ple countries. The growth rate of number of good categories range from -9.4%
73
(Slovak Rep.) to 43.9% (Poland), where six of the eight countries in consideration
experienced positive growth rates. Between 1995 and 2003, the average number
of goods exported per country was 3,640, while this number rose to 3,870 during
the period of 2004-2012, showing a 6.3% increase along the new goods margin.
However, looking solely at the growth in the number of goods does not
reveal much about the variety growth. In order to better understand the trends
in the export variety, one needs to take into account the growth in the number of
partner countries as well. Column four of Table 2.1 reports the average number
of countries per year, importing each individual good. These data reveal a sub-
stantial increase in the average number of partner countries that import a good
during the post-EU period. The range of average number of partners increased
from 8-to-32, to 16-to-46 after joining the EU. Poland experienced the highest
growth in partners (209.6%), whereas Estonia ranked last with a growth rate of
31.1%. Overall, it is observed that the average number of partners per country
increased from 17 during 1995-2003 to 20 during 2004-2012, showing a 70.6% rise.
Thus, even though the number of goods for each country does not increase sub-
stantially, the data reveal that there has been a dramatic increase in the number of
trading partners importing each individual good between the pre- and post-EU
periods, for every country in the sample.
The growth rates are even more dramatic when the number of varieties
(country-good pairs) is considered. The fifth column of Table 2.1 displays the
average number of varieties that each country exports, in a given year. Taken
together, the data suggest that the average number of varieties per country rose
from 32,403 to 57,656 between 1995-2003 and 2004-2012. This increase indicates
that after joining the EU in 2004, the countries experienced a 77.9% increase in
74
the number of varieties they export, most of which is due to the increase in the
average number of countries importing each good. Looking at the data coun-
try by country, one can see that Poland experienced the highest growth in the
extensive margin, with a 424.5% increase in the number of varieties. This means
that after joining the EU, the number of varieties more than quadrupled. Follow-
ing Poland, the number of varieties in Latvia grew by 121.8% after joining EU,
indicating a more than twofold increase in the extensive margin.
The fact that average number of varieties over the sample countries almost
doubled between the pre- and post-EU periods, coupled with the 70% increase in
the number of partners serves as an evidence that joining the EU had a significant
effect not only on the total exports, but also on the diversification of trade among
these countries. One of the explanations of this rise is that reductions in tariff
rates applied to the exports of these countries may have made it possible to
export new varieties to different countries (Broda and Weinstein, 2006). On the
other hand, these countries go under extensive changes in their trade policies
in order to join the EU. The evidence that the average number of goods grew
only moderately while the average number of partners almost doubled suggest
that joining the EU helped them to build and sustain trading relationships with
a larger set of partner countries among the EU members as well as the rest of the
world.
Although the large growth rates of the number of varieties at the 6-digit
level gives us an idea about the variety gains from joining the EU, it does not
necessarily indicate a significant change in the sectoral structure of the exports. It
may be the case that these newly traded varieties come from the existing export
sectors. To be more specific, the variety growth observed in the data may be
75
a result of the enlargement of the product basket within the sectors that were
already exporting to a specific market. Alternatively, the existing sectors may be
exporting the same set of goods to a larger set of markets. When a variety is
defined as a 6-digit good-market category, both scenarios would result in higher
number of varieties. Though one might ask: Is the variety growth after joining
the EU related to the creation of the new export sectors?
Table 2.3
Summary Statistics - Export Variety: Unweighted Count, Product: HS-2 level
Country Period Num. of
Goods
Num. of
Partners
Num. of
Varieties
Export
Growth
Czech Rep. 1995-2003 96 85 6402 [211.1]
2004-2012 96 101 7843
[0.0] [18.8] [22.5]
Estonia 1995-2003 96 42 2935 [176.5]
2004-2012 96 54 3525
[0.0] [30.6] [20.1]
Hungary 1995-2003 96 64 4699 [171.7]
2004-2012 96 77 5531
[0.0] [20.3] [17.7]
Latvia 1995-2003 97 31 2020 [234.4]
2004-2012 96 60 3902
[-1.0] [94.8] [93.2]
Lithuania 1995-2003 97 40 2841 [265.1]
2004-2012 96 63 4468
[-1.0] [56.4] [57.3]
Poland 1995-2003 96 49 3358 [249.1]
2004-2012 96 106 8550
[0.0] [114.4] [154.6]
Slovak Rep. 1995-2003 96 53 3447 [282.3]
2004-2012 96 71 4735
[0.0] [33.5] [37.4]
Slovenia 1995-2003 96 57 3725 [107.0]
2004-2012 95 73 4930
[-1.0] [27.5] [32.3]
Turkey 1995-2003 96 95 7446 [200.9]
2004-2012 96 121 9914
[0.0] [27.1] [33.1]
Notes: Table reports the average number of products, partners and varieties over the pre- (1995-2003) and post-EU (2004-
2012) periods. Each HS-2 category defines a product and each product-partner combination is defined as a variety. The
percentage growth rates between pre- and post-EU periods are given in brackets. Numbers are based on the author’s
calculations from UN COMTRADE data.
76
In order to look into the effect of accession to the EU on the number of sectors
these countries export in, I repeat the exercise at higher levels of aggregation. The
idea is that if similar magnitudes of growth in the number of goods and number
of good-market categories are observed at the higher aggregation levels, one may
conclude that joining the EU not only increased the product differentiation within
the existing export sectors, but also it made it possible for new sectors to engage
in international trade.
This hypothesis is in line with the Helpman et al. (2008) model of interna-
tional trade with heterogeneous firms. In a setting where firms vary by produc-
tivity and there are fixed exporting costs, only relatively more productive firms
find it profitable to export. This means that there are less productive firms that
only produce for the domestic market. However, as the fixed costs to export are
reduced or eliminated, it will be profitable for these lower productivity firms to
enter the export market. Moreover, Helpman et al. (2008) allows the profitability
of exports vary by destination markets. If this is the case, there should be some
firms that find new markets more profitable after a trade liberalization episode.
Same logic is applicable at the sectoral level. There should be some sectors, that
were producing only domestically, become capable of exporting after trade lib-
eralization. However, if the large growth rates are only abserved at the 6-digit
level, but not at the 4- or 2-digit levels, one may conclude that the extensive mar-
gin growth was limited to the increased variety within the existing export sectors,
but was not due to an increase in the number of sectors that the countries export
in.
77
Table 2.2 and Table 2.3 report the change in the extensive margin of exports
in each country after joining the EU based on the HS of 4-digit and 2-digit com-
modity classification codes, respectively. The empirical evidence suggests that at
the 4-digit level, the growth of the number of goods remains below its counter-
part at the 6-digit level for every country in the sample, with the exception of
Estonia. Table 2.4 reports that on average, the average number of goods grew by
3.1%, rising from 1,075 to 1,108 after joining the EU. This is roughly the half of
the growth that have been seen at the 6-digit level. However, the fact that the
growth of the number of partner countries remained quite stable across differ-
ent HS levels results in the similar growth rates of number of varieties at 6- and
4-digit levels of aggregation. On the other hand, it is observed that the average
number of goods remained almost the same between pre- and post-EU periods
at the 2-digit level, with a growth rate of -0.4%.
Hummels and Klenow (2005) define a variety as a 6-digit HS category and a
sector as a 2-digit HS category. Based on this definition and my findings concern-
ing higher aggregation levels, I can say that the increased variety after joining the
EU was dominated by product-market differentiation within the existing export
sectors. The low growth rate of the number of goods at 4- and 2-digit levels indi-
cate that the EU accession did not create new export sectors, but it increased the
variety of the existing export sectors.
Overall, the empirical evidence presented in Table 2.1 through Table 2.4 show
that the new members of the EU experienced a significant growth in bilateral
trade. Their total exports grew by 212.2% on average between the periods 1995-
2003 and 2004-2012. More interestingly, the growth in total exports was accom-
panied by almost 80% growth in the export variety, which confirms the findings
78
Table 2.4
Summary Statistics - Overall Export Variety: Unweighted Count
Aggregation Measure Period Mean Median St. Dev Growth
HS-6 Num. of Goods 1995-2003 3640 3609 520.2 [6.3]
2004-2012 3870 3724 421.9
Num. of Partners 1995-2003 17 16 7.4 [70.6]
2004-2012 29 27 11.6
Num. of Varieties 1995-2003 32403 26462 18815.3 [77.9]
2004-2012 57656 48611 33468.8
HS-4 Num. of Goods 1995-2003 1075 1076 78.1 [3.1]
2004-2012 1108 1097 64.4
Num. of Partners 1995-2003 26 24 10.5 [61.9]
2004-2012 41 40 14.3
Num. of Varieties 1995-2003 15863 13620 8113.8 [67.5]
2004-2012 26562 22834 13306.3
HS-2 Num. of Goods 1995-2003 96 96 0.5 [-0.4]
2004-2012 96 96 0.4
Num. of Partners 1995-2003 53 51 16.9 [43.5]
2004-2012 76 72 18.7
Num. of Varieties 1995-2003 3678 3403 1342.8 [47.8]
2004-2012 5436 4833 1820.1
Export Growth [212.2] [222.8] [54.3]
Notes: Table reports the overall sample statistics for the pre- (1995-2003) and post-EU (2004-2012) periods. The values
are calculated by taking the average of each statistic for the two time frames, over the sample countries except Turkey.
Each HS-6 (HS-4 and HS-2) category defines a product and each product-partner combination is defined as a variety. The
percentage growth rates between pre- and post-EU periods are given in brackets. Numbers are based on the author’s
calculations from UN COMTRADE data.
of Hummels and Klenow (2005) suggesting that countries that export more, do
so by exporting a wider range of varieties. However, as mentioned before the
crude measure of export diversification tends to overestimate the variety growth
of exports because it assigns equal weights to all varieties in the export basket.
Thus, in order to get a more robust picture, one needs to examine the trends by
using the weighted measures.
Turkey is included in this analysis as a placebo. The expectation is that if
trade liberalization and joining the EU are the major factors behind the extensive
margin growth, one should see weaker trends in Turkey’s statistics since it started
its trade liberalization policies in the 1980s and has already been a member of the
79
CU for eight years in 2004. The data reveals that the average number of goods
exported by Turkey decreased at both 6-digit and 4-digit levels, by 0.8% and 0.5%
respectively between the periods 1995-2003 and 2004-2012. However, similar to
the CEECs, there has been a 55% (44% at the 4-digit level) increase in the average
number of varieties, which is dominated by the increase in the number of export
markets. It seems that, as expected, the effect of the CU on the product variety
has faded away, but the impact on the number of trading partners has lasted for
a longer period of time.
This pattern is consistent with the findings of Eaton et al. (2007). By track-
ing Colombian firms’ entry and exit into and out of destination markets, they
find that the firms who survive the first year in a new market, expand into
additional destinations gradually. They relate their results to a two-tiered entry
cost structure, where the first entry costs remain relatively small, but the cost of
establishing lucrative long-term export contracts that depends on the learning of
processes at work, is quite large. Once firms overcome these larger costs, they
become more likely to penetrate other markets, which has amplifying effects on
the market variety in the long-run.
It should be noted that the simple count measure of the export diversification
applied in this analysis is closely related to the definition of a non-traded good. The
values presented in Table 2.1 through Table 2.4 take a good as non-traded if the
exports of that good is registered zero for the given year. A good is treated
as traded, as long as the registered value of exports exceeds zero U.S. dollars.
Two concerns arise with this approach. First, small-value shipments tend to go
unreported, so even though there is a trade relationship between the countries,
this can not be observed in the data. Second, export relationships can be observed
80
even if the good is traded in small values, for a short period of time (Kehoe and
Ruhl, 2013). Following Kehoe and Ruhl (2013), I define a second version of
simple count measure. I order the goods by their average value of exports over
the first three years of the sample (1995-1997). Then, I take the cumulative sum
of the exports, forming 10 sets, each representing 10% of the total exports of
the country in a given year. The first set is named as the least traded goods.
My objective is to analyze the changes in the new goods margin by looking at
the share of the least traded goods of 1995-1997 over the sample period. More
specifically, I will compare the share of the 1995-1997 least traded goods in 2004
and 2012, respectively. An increase in the share of these least traded goods would
indicate a rise in the importance of the new goods margin in the bilateral trade
of the sample countries.
The evolution of the exports of each country in the dataset is presented in
Figure 2.1 through Figure 2.8. The bars represent the 10 sets of cumulative frac-
tions of the country’s exports. By construction, the first 10% is the least traded
goods based on their average export value over the period 1995-1997. On the
y-axis, there is the share of the 1995-1997 least traded goods in 2004 and 2012,
respectively. Note that, if the growth in exports were dominated by the intensive
margin only, the shares of each percentile would remain the same over the years.
However, if the share of the least exported goods increase over time, then I would
conclude that there is extensive margin growth Kehoe and Ruhl (2013).
Figure 2.1 through Figure 2.8 show that there is an upsurge in the share of
the least traded goods. For all the countries in the sample, by 2004, the least
exported goods that accounted for 10% of the total exports, rise to at least 30%.
By 2012, they account for at least 41% of the exports. On average, the bottom
81
Figure 2.1
Fraction of the least traded goods in total exports: Czech Republique
5 10 15 20 25 30
% of 2004 Export Value
0 20 40 60 80 100
Cumulative % of 1995 Export Value
Czech Rep. 1995 Least Traded Goods in 2004
0 10 20 30 40
% of 2012 Export Value
0 20 40 60 80 100
Cumulative % of 1995 Export Value
Czech Rep. 1995 Least Traded Goods in 2012
Figure 2.2
Fraction of the least traded goods in total exports: Estonia
0 10 20 30 40
% of 2004 Export Value
0 20 40 60 80 100
Cumulative % of 1995 Export Value
Estonia 1995 Least Traded Goods in 2004
0 20 40 60
% of 2012 Export Value
0 20 40 60 80 100
Cumulative % of 1995 Export Value
Estonia 1995 Least Traded Goods in 2012
10% of the exports, account for 41% of the total exports in 2004, and 46.4% of
the exports in 2012. Two conclusions may be drawn from these results. First, the
new goods margin clearly plays a significant role in changing the trade patters of
the CEECs following their accession to the EU. Second, the experience of the EU
joiners differs from the other trade liberalization cases in terms of the magnitudes
of the growth rates.
Kehoe and Ruhl (2013) finds that the least traded goods account for 25% of
Mexican exports to Canada and 30% of the Canadian exports to Mexico after
82
Figure 2.3
Fraction of the least traded goods in total exports: Hungary
0 10 20 30 40 50
% of 2004 Export Value
0 20 40 60 80 100
Cumulative % of 1995 Export Value
Hungary 1995 Least Traded Goods in 2004
0 10 20 30 40 50
% of 2012 Export Value
0 20 40 60 80 100
Cumulative % of 1995 Export Value
Hungary 1995 Least Traded Goods in 2012
Figure 2.4
Fraction of the least traded goods in total exports: Latvia
0 10 20 30 40 50
% of 2004 Export Value
0 20 40 60 80 100
Cumulative % of 1995 Export Value
Latvia 1995 Least Traded Goods in 2004
0 20 40 60
% of 2012 Export Value
0 20 40 60 80 100
Cumulative % of 1995 Export Value
Latvia 1995 Least Traded Goods in 2012
the implementation of NAFTA. The numbers drop down to 12% and 14% when
U.S. exports to Canada and Mexico are considered. Mukerji (2009) finds that the
new goods account for 26.5% of Indian exports, following the unilateral trade
liberalization of India. On the other hand, my findings closely match the findings
of Dalton (2013). He shows that, following the 2004 enlargement, the share of the
new goods in Austrian imports from the CEECs used in this study increases from
10% to 47%.
83
Figure 2.5
Fraction of the least traded goods in total exports: Lithuania
0 10 20 30 40 50
% of 2004 Export Value
0 20 40 60 80 100
Cumulative % of 1995 Export Value
Lithuania 1995 Least Traded Goods in 2004
0 20 40 60
% of 2012 Export Value
0 20 40 60 80 100
Cumulative % of 1995 Export Value
Lithuania 1995 Least Traded Goods in 2012
Figure 2.6
Fraction of the least traded goods in total exports: Poland
0 10 20 30 40
% of 2004 Export Value
0 20 40 60 80 100
Cumulative % of 1995 Export Value
Poland 1995 Least Traded Goods in 2004
0 10 20 30 40 50
% of 2012 Export Value
0 20 40 60 80 100
Cumulative % of 1995 Export Value
Poland 1995 Least Traded Goods in 2012
Though not reported here, I calculated the share of the least traded goods
for each year in the sample. I find that, although their fraction mostly remains
below 15% until 1998, by 1999, they account for at least 20% of the total exports
of each country in the sample. On average, the goods that account for 10% of the
exports in 1995, account for 24% of the exports in 1999. Therefore, it is safe to
say that the growth in the new goods margin becomes significant around 5 years
prior to the actual EU accession. Bergin and Lin (2012) find similar results that
84
Figure 2.7
Fraction of the least traded goods in total exports: Slovakia
0 10 20 30 40
% of 2004 Export Value
0 20 40 60 80 100
Cumulative % of 1995 Export Value
Slovakia 1995 Least Traded Goods in 2004
0 20 40 60
% of 2012 Export Value
0 20 40 60 80 100
Cumulative % of 1995 Export Value
Slovakia 1995 Least Traded Goods in 2012
Figure 2.8
Fraction of the least traded goods in total exports: Slovenia
0 10 20 30 40
% of 2004 Export Value
0 20 40 60 80 100
Cumulative % of 1995 Export Value
Slovenia 1995 Least Traded Goods in 2004
0 10 20 30 40 50
% of 2012 Export Value
0 20 40 60 80 100
Cumulative % of 1995 Export Value
Slovenia 1995 Least Traded Goods in 2012
the effect of European Monetary Union sets in 4 years ahead of the actual agree-
ment. Magee (2008) also finds significant anticipatory effects of regional trade
agreements, going back 4 years leading up to the beginning of the agreement.
2.4.2 Weighted Measures of Export Variety
As mentioned before, the unweighted measures tend to overestimate the growth
of the extensive margin. The main reason of this is the fact that the simple count
85
approach treats each good and each destination market as the same. It does
not take into account the relative importance of the goods in the world market,
or different characteristics of the destination markets. Thus, exporting a small
good (in terms of its share in the world exports) to a less competitive destination
is considered the same as exporting a widely traded good to a more competi-
tive destination. In order to compute the extensive margin more accurately, one
needs a measure that gives different weights to different varieties. The weighted
measure of the extensive margin applied hereafter weights each good-market
category by its share in the rest of the world exports, minimizing the risk of
overstatement of the variety growth (Hummels and Klenow, 2005).
I calculated the extensive and intensive margins of exports of each country
and year, according to the equations (2.2) and (2.3). Then, I took the average
values for the periods 1995-2003 and 2004-2012 separately. Table 2.5 reports the
growth in the extensive and intensive margins of exports along with the growth
in total exports between the periods of pre- and post-EU for each country in
the sample, broken down by different levels of aggregation and different variety
definitions. Columns three and four take a variety as a good-market category,
letting destination markets, along with the exported goods, vary in terms of
profitability, as suggested by Helpman et al. (2008). Whereas, columns five and
six treat each HS-product category as a different variety, giving different weights
to goods, but same weights to each market. That is, if a country exports the same
set of goods in each period, but to a larger set of partners in the second period, it
will have a larger extensive margin growth in column three. The overall sample
statistics are reported in Table 2.6. The values in Table 2.6 are calculated by
computing each statistic over the average growth rates of the sample countries
except Turkey.
86
Table 2.5
Change in the Average Export Variety (in %) - Weighted Count
Variety: (Good-Market) pair Variety: Good
Country Product
Code
Extensive
Margin
Intensive
Margin
Extensive
Margin
Intensive
Margin
Exports
Czech Rep.
HS-6 16.3 45.9 1.4 67.0
HS-4 7.1 58.7 4.7 61.9 211.1
HS-2 4.2 62.8 1.8 66.9
Estonia
HS-6 30.5 9.44 -0.6 51.4
HS-4 19.7 23.5 1.9 47.3 176.5
HS-2 13.7 33.0 1.7 48.3
Hungary
HS-6 16.4 30.9 4.5 44.3
HS-4 4.5 45.6 4.5 44.5 171.7
HS-2 2.4 47.0 1.0 48.5
Latvia
HS-6 73.6 0.2 5.0 71.8
HS-4 54.4 14.1 4.8 72.1 234.4
HS-2 26.1 42.2 0.1 81.0
Lithuania
HS-6 37.0 34.4 1.6 93.7
HS-4 25.0 52.7 4.3 88.4 265.1
HS-2 17.4 67.3 0.3 96.7
Poland
HS-6 102.4 -6.6 12.2 69.8
HS-4 59.6 19.3 8.0 76.2 249.1
HS-2 17.5 61.9 1.2 88.2
Slovak Rep.
HS-6 33.4 52.7 -0.2 108.2
HS-4 18.4 75.3 2.5 102.8 282.3
HS-2 6.6 95.1 2.1 104.3
Slovenia
HS-6 29.3 -15.3 2.1 10.3
HS-4 17.3 -5.5 3.2 9.3 107.0
HS-2 2.9 9.3 0.3 12.3
Turkey
HS-6 29.6 18.6 -1.2 59.0
HS-4 15.3 35.2 0.0 57.2 200.9
HS-2 4.6 49.9 1.7 55.2
Notes: Table reports the percentage change in the average extensive and intensive margin measures between the peri-
ods pre- (1995-2003) and post-EU (2004-2012). Each HS category defines a product. A variety is defined either as a HS
category or a product-market category. Numbers are based on the author’s calculations from UN COMTRADE data.
At the first glance, it is evident that the unweighted measure significantly
overestimates the extensive margin growth, when a variety is defined as a
market-good category. However, the growth rates along the extensive margin
are still very high. The data in Table 2.5 reveal that at the 6-digit level, for every
country except Estonia, the growth in the average number of varieties exceeds
87
the growth in the extensive margin. The extensive margin growth ranges from
16.3 to 102.4%. Poland is the country that experienced the largest variety growth,
whereas Czech Republic has the lowest growth rate. On average, the 2004 join-
ers experienced a 42.4% increase in the variety of their exports. Based on this
evidence, I can say that, though not as high as 78% (as the unweighted measure
predicts), the CEECs experience a major growth in extensive margin of exports
following their accession to the EU.
On the other hand, the growth rates drop substantially, when different desti-
nation markets are not taken into account in calculation of the extensive margin.
In Czech Republic, for example, column five of Table 2.5 shows that the exten-
sive margin growth drops from 16.3% to 1.4% when a variety is defined as a
good category rather than a good-market pair. Same trend is observed for all
countries in the sample. In terms of good categories, Table 2.6 shows that the
average extensive margin growth is 3.25%, which appears quite low compared
to the 42.4% increase in the market-good categories. This pattern confirms our
prior findings that the extensive margin growth is dominated by penetration of
new destinations.
The fourth column of Table 2.5 reports the growth of average intensive mar-
gin. It shows that the intensive margin growth exceeds the extensive margin
growth only for three countries. For five countries in the sample, the extensive
margin growth dominates the growth of total exports between the eras 1995-
2003 and 2004-2012. The data presented in Table 2.6 reveal that on average, the
mean intensive margin grew by 18.9%, which is 23.5 percentage points below the
growth of extensive margin.
88
In contrast to the patterns of the extensive margin, the growth in the inten-
sive margin more than triples, rising to 64.6%, when the destination market is
ignored. Considering the evidence that the average number of partners import-
ing each good was the dominant factor behind the increase in the number of
varieties exported, this is not a surprising result. The substantial increase in
the intensive margin growth from 19% to 65%, combined with the decline in
the extensive margin growth from 42.4% to 3.25% means that the variety gains
observed after joining the EU stem mainly from accession to new export mar-
kets rather than an increase in the set of goods that are exported. These results
are parallel to the findings of Evenett and Venables (2002). Their measure of
extensive margin depends on the exports of the previously traded goods to new
markets. They find significant growth in the extensive margin due to the exports
to new destinations.
Table 2.6
Change in the Overall Export Variety (in %) - Weighted Count
Variety: (Good-Market) pair Variety: Good
Statistic Product
Code
Extensive
Margin
Intensive
Margin
Extensive
Margin
Intensive
Margin
Exports
Mean
HS-6
42.36 18.96 3.25 64.56 212.15
Median 31.95 20.17 1.85 68.40 222.75
St. Dev. 30.13 25.40 4.12 30.16 58.00
Mean
HS-4
25.75 35.46 4.24 62.81 212.15
Median 19.05 34.55 4.40 67.00 222.75
St. Dev. 20.45 26.90 1.86 29.18 58.00
Mean
HS-2
11.35 52.33 1.06 68.28 212.15
Median 10.15 54.45 1.10 73.95 222.75
St. Dev. 8.64 25.66 0.77 30.68 58.00
Notes: Table reports the percentage change in the average extensive and intensive margin measures between the periods
pre- (1995-2003) and post-EU (2004-2012), over the sample countries except Turkey. Each HS category defines a product.
A variety is defined either as a HS category or a product-market category. The overall sample statistics are calculated
by averaging the growth rates of the mean values of each margin between the two periods. Numbers are based on the
author’s calculations from UN COMTRADE data.
89
Overall, the results presented in this section provide strong evidence sup-
porting the hypothesis that the dramatic increase in exports following trade lib-
eralization is accompanied by a similar spike in the extensive margin. More
interestingly, the growth in the intensive margin falls below the growth in the
extensive margin, suggesting that the boost in the exports of the CEECs was the
result of exports of new products to new destinations, rather than an increase in
the trade of the existing products to the same markets. This result once more
underlines the importance of the extensive margin in the export growth follow-
ing trade liberalization. The fact that expanding to new destination markets
plays a more important role than expanding the product range in increasing the
extensive margin of exports suggests that in order to take full advantage of trade
liberalization in terms of export growth, policy makers should seek to generate
growth along the extensive margin by facilitating exports to different partner
countries.
My findings also coincide with the findings of Broda and Weinstein (2006).
They find a 12.5% increase in the average number of goods imported by U.S.,
during the implementation period of NAFTA. On the other hand, when they
look at the common goods imported at the start and end of the period, they find
that the average number of partners supplying each good increased by almost
32%. They explain this rise by the assumption that goods are differentiated by
country (as in Krugman (1980)), and the globalization process that enables to
source the varieties from different countries. Similar logic can be applied to the
exports. By joining the EU, these countries not only removed the barriers to
trade with the EU countries, they also adopted the tariff structure of the EU with
respect to the rest of the world. The implementation of these relatively more
relaxed trade policies may have reduced the fixed trade costs, enabling them to
90
export to different countries in the rest of the world, as well as within the EU.
On the other hand, if the goods are differentiated by country, then it is natural
to think that more countries in the world would demand the goods supplied by
these countries.
When the trend of exports in Turkey is analyzed, one sees that unlike the
2004 joiners, the extensive margin of Turkey’s exports grew only 29.6% between
1995-2003 and 2004-2012, which is roughly 12.8 percentage points below the
growth that CEECs experienced. However, a 30% increase is still significant. This
shows that the impact of the trade liberalization policies that are implemented in
line with the Turkey’s accession to the CU, have long lasting effects on the export
variety.
2.5 Regression Analysis
2.5.1 Specification
To further study the effect of EU accession on the extensive margin of CEECs’
exports, I use an augmented version of the standard gravity model, which is
considered as the standard empirical tool for explaining the trade flows. The
gravity equation is initially introduced by Tinbergen (1962) to study the impact of
the economic factors of the trade partners on their bilateral trade with each other.
Following Tinbergen, ? used the gravity model to estimate the impact of the
regional trade agreements on the member countries’ trade volume. Since then,
it has been widely used to infer trade flow effects of protection and openness,
91
regional trade agreements, international borders, linguistic identity and exchange
rate mechanisms.
Even though the gravity model was originally entirely empirical, Anderson
(1979) provided a theoretical foundation to the gravity equation based on an
expenditure system. He defined the volume of trade from country j to country i
(X
ji
) as:
X
ji
= q
j
Y
i
(2.5)
assuming that consumers have identical homothetic preferences and each coun-
try specializes completely in production of one good, where its price is normal-
ized to unity. In equation (2.5)q
j
denotes the fraction of income spent on country
i’s product and Y
i
denotes real GDP in importing country i. Since production of
every country j must equal the volume of exports and domestic consumption of
the good, market-clearing implies:
Y
j
= q
j
Y
W
(2.6)
Then,
q
j
=
Y
j
Y
W
(2.7)
where Y
W
is the world real GDP . Substituting equation (2.7) into equation (2.5)
yields the basis for the simple frictionless gravity equation:
92
X
ji
=
Y
j
Y
i
Y
W
(2.8)
The traditional gravity model is the log-linear specification of equation (2.8).
It explains bilateral trade with the joint income of countries. In order to mea-
sure the impact of regional trade agreements and other country specific factors,
dummy variables were added to the equation that takes the value of 1 if the
country possesses the interested quality (Aitken (1973), Frankel et al. (1997)). In
this case, the equation looks like as follows:
PX
j
= b
0
(Y
j
)
b
1
(Y
i
)
b
2
(Dist
ji
)
b
3
e
b
4
(Z
ji
)
#
ji
(2.9)
where, e is the natural logarithm base, Dist
ji
denotes the distance between part-
ner countries (used as a proxy for variable trade costs) and Z
ji
denotes the
dummy variables of interest.
In this study, I estimate an augmented version of the standard gravity equa-
tion. Since OLS is linear, equation (2.4) implies that the sum of the estimated
coefficients for the two margins of any independent variable will equal the coef-
ficient of that variable in a standard gravity equation with relative exports as the
dependent variable. Therefore, based on the construction of the two margins,
I can estimate the log of extensive margin instead of total exports
2
. The exact
specification I use is:
2
See Crozet and Koenig (2010) for a more explicit derivation.
93
EM
jt
= b
0
+b
1
EU+b
2
(GDP
jt
GDP
it
)+b
3
Distance
ji
+b
4
(Pop
j
Pop
i
)+b
5
Area
j
+b
6
Area
i
+b
7
Landl
ji
+b
8
Contig
ji
+b
9
ComLeg
ji
+b
10
ComLand
ji
+b
11
ComCol
ji
+b
12
Col45
ji
+b
13
Colony
ji
+
å
t
a
t
T
t
+
å
j
g
j
Exporter
+
å
i
f
i
Importer+#
jit
(2.10)
where j and i denote exporter and importer countries, respectively, and t denotes
time. All numeric variables go into the equation in natural logarithm. The vari-
ables are defined as:
EM: Natural logarithm of the extensive margin of exports.
EU: Dummy variable that takes the value 1 in 2004 and afterwards.
Y: Natural logarithm of real GDP .
Distance: Natural logarithm of distance between country i and country j in
kilometers.
Pop: Natural logarithm of population.
Area: Natural logarithm of land area in square kilometers.
Landl: Number of landlocked countries in the country-pair.
Contig: Dummy variable that takes the value 1 if the country-pair share a
border.
ComLeg: Dummy variable that takes the value 1 if the country-pair share
a common legal origin.
94
ConLang: Dummy variable that takes the value 1 if a language is spoken
by at least 9% of the population in both countries.
ComCol: Dummy variable that takes the value 1 if the country-pair have a
common colonizer post 1945.
Col45: Dummy variable that takes the value 1 if the country-pair was in a
colonial relationship post 1945.
Colony: Dummy variable that takes the value 1 if the country-pair was ever
in a colonial relationship.
fT
t
g: Set of time fixed effects.
fExporterg andfImporterg: Sets of country fixed effects.
#: Omitted other influences, assumed to be well behaved.
In the classic gravity equation, the geographical distance variable is used to
capture the variable trade costs between two country pairs. Among the wide
range of theoretical derivations of the gravity equation, a limited number of
authors model trade costs explicitly (Baier and Bergstrand, 2001, Bergstrand,
1985, 1989, Deardorff, 1995, Lim˜ ao and Venables, 2001). Based on these mod-
els, use of tariffs and transportation costs as a means to capture the effects of
trade frictions became more and more popular in the recent empirical studies.
However, detailed data on transportation costs are available for a very limited
number of countries, and though more available, tariff data is still scarce. There-
fore, following many papers in the related literature, I use geographical distance
as a crude proxy for trade costs.
My objective is to control for as many natural factors affecting trade as possi-
ble and look for the effect of the EU on the extensive margin. My main hypothesis
is that the EU accession decreases the fixed trade costs and results in an increase
in the product variety of exports as well as the range of destination markets
95
that CEECs export to. Therefore, I seek to confirm that the coefficient of the EU
dummy variable is positive and its magnitude is similar to my prior findings.
2.5.2 Estimation Results
Data analysis of the previous sections provide strong evidence that the CEECs
experienced a significant growth of extensive margin following their accession to
the EU. In response to a 212.15% increase in total exports, the average extensive
margin grew by 42.36% between the pre- and post-EU periods. To isolate the
effect of the EU accession on the product variety, I employ standard regression
analysis using the gravity equation described earlier. I estimate the impact of
the accession to the EU on the extensive margin of exports based on equation
(2.10) and the methodology described in the previous subsection and find similar
results.
The regression results are reported in Table 2.7 with the dependent variable
being log of the extensive margin. Column (1) shows the ordinary least squares
estimation of the equation (2.10). These results show that the model works rela-
tively well. As the economic mass of the countries gets larger and the geographic
distance between them gets smaller, the countries expand their exports in terms
of the range of products and partner countries. Both real GDP and the distance
variables have correct sings and they are statistically significant at 1% level.
As mentioned earlier, the coefficient of the EU dummy variable is of central
importance to this part of the study. The results presented in Table 2.7 show that
the dummy variable EU has a positive and statistically significant (at 1% level)
96
Table 2.7
Gravity Regression results for Extensive Margin of Exports - variety:
product-market pair
(1) (2) (3)
EU 0.2815*** 0.5225*** 0.5505***
[0.004] [0.011] [0.013]
Distance (Log) -0.0263*** -0.0238** -0.0015
[0.003] [0.008] [0.001]
GDP
j
x GDP
i
(Log) 0.0322*** 0.0302*** 0.0495***
[0.001] [0.004] [0.008]
Pop
j
x Pop
i
(Log) 0.0213*** 0.0242*** -0.1998***
[0.002] [0.006] [0.032]
Area
i
(Log) 0.0883*** 0.0886*** -0.1045***
[0.003] [0.009] [0.013]
Area
j
(Log) -0.0348*** -0.0344*** 0.1418***
[0.001] [0.004] [0.028]
Num. of Landlocked 0.2305*** 0.2304*** -0.1647***
[0.003] [0.010] [0.032]
Common Border 0.0619*** 0.0608 0.0004
[0.012] [0.036] [0.002]
Common Legal Origin -0.0032 -0.0005 -0.0445
[0.006] [0.019] [0.040]
Common Language -0.1506*** -0.1517* 0.0071**
[0.024] [0.070] [0.003]
Common Colonizer - post 1945 -0.2558*** -0.2556*** -0.0122***
[0.013] [0.037] [0.002]
Colonial Relationship - post 1945 -0.0722 -0.0796 -0.0142*
[0.045] [0.112] [0.006]
Ever Colony -0.1222*** -0.1171 0.0011
[0.025] [0.060] [0.004]
Time Fixed Effects No Yes Yes
Country Fixed Effects No No Yes
R
2
0.478 0.58 0.915
Adjusted-R
2
0.478 0.579 0.915
N 21259 21259 21259
Notes: All models report clustered robust standard errors, which account for panel-level heteroskedasticity and autocor-
relation of arbitrary form, in parentheses.
*
indicates significance at the 10% level,
**
indicates significance at the 5% level,
and
***
indicates significance at the 1% level.
coefficient, indicating the extensive margin have increased by 28.15% after the
EU accession.
Although it gives a general idea, there is a strong possibility that pooled
OLS suffers from omitted variable bias. In order to avoid omitting variables
97
that affect bilateral trade, I subsequently add time, exporter and importer fixed
effects, where standard errors are adjusted for clustering on country-pairs. The
results of these estimations are reported in columns (2) and (3) of Table 2.7.
The results show that, with the inclusion of the fixed effects, the explanatory
power of the model increases substantially. While regular OLS explains 48% of
he variation in the extensive margin, time fixed effects model explains 58%, and
the time and country fixed effects model explains almost 92% of the variation in
the extensive margin.
At first glance, the coefficient of the EU dummy variable is positive and sta-
tistically significant at 1% level in both models ((2) and (3)). However, compared
to the pooled OLS results, the magnitude of the EU effect increases substantially.
To be precise, when only time fixed effects are considered, it is found that EU
accession increases the extensive margin of exports by 52%. Once exporter and
importer fixed effects are included in addition to the time fixed effects, the coef-
ficient of the EU dummy variable rises to 0.55, indicating a 55% growth in the
extensive margin following the EU accession.
These findings confirm my main hypothesis that the trade liberalization sig-
nificantly increased the extensive margin of exports of CEECs. In all models the
coefficient on the EU dummy variable is positive and highly statistically signif-
icant at 1% level. Moreover, the magnitude of the coefficient is similar to my
findings in the previous section, indicating a 55% growth in extensive margin of
exports post 2004.
98
2.6 Decomposition of the Relative Export Growth
I constructed extensive and intensive margin measures so that for each country
and year the relative exports of country j with respect to the rest of the world
equals the product of two margins:
Relative Exports
row
j
= EM
j
IM
j
(2.11)
Thus, the product of the growth rates of each margin yields the growth rate
of the relative exports:
(1+ g
RE
j
)=(1+ g
EM
j
)(1+ g
IM
j
) (2.12)
Taking logarithms, we have:
g
RE
j
g
EM
j
+ g
IM
j
(2.13)
Up to this point, my findings show that after joining EU in 2004, the CEECs
experienced a startling increase in their aggregate exports, which is accompa-
nied by a large increase in their export variety. How much of this increase in
total exports is due to trading in a larger set of varieties? Does the extensive mar-
gin growth follow similar trends when the destination markets include only EU
member countries rather than the rest of the world? How does the contribution
of each margin compare when relative exports to the EU and Non-EU countries
99
considered separately? In order to answer these questions, first, I decompose the
relative export growth into its extensive and intensive margin components, using
equation (2.13) and calculate the contribution of each margin to the growth of
relative exports with respect to the rest of the world. Second, I divide the sample
into two, based on the destination markets. I compute the extensive and inten-
sive margin of exports along with their contributions to the relative exports to
the EU and Non-EU countries separately.
Table 2.8
Decomposition of Relative Export Growth with the Rest of the World (%)
Country Extensive Margin Intensive Margin Relative Exports
Czech rep. 16.3 45.9 62.2
[26.2] [73.8]
Estonia 30.5 9.4 40.0
[76.4] [23.6]
Hungary 16.4 30.9 47.3
[34.6] [65.4]
Latvia 73.6 0.2 73.8
[99.7] [0.3]
Lithuania 37.0 34.4 71.4
[51.8] [48.2]
Poland 102.4 -6.6 95.8
[106.9] [-6.9]
Slovakia 33.4 52.7 86.2
[38.8] [61.2]
Slovenia 29.3 -15.3 14.0
[209.7] [-109.7]
Turkey 29.6 18.6 48.1
[61.4] [38.6]
Notes: Table reports the growth rates (in %) of the extensive margin, intensive margin and the relative exports in trade
with the rest of the world, between the pre- (1995-2003) and pos-EU (2004-2012) periods. The contribution of each margin
to the total relative export growth is given in the squared brackets. Relative exports are calculated as the ratio of total
exports (to the rest of the world) to the total world exports. Numbers are based on the author’s calculations from UN
COMTRADE data.
The results of this decomposition are reported in Table 2.8 through Table
2.10. Table 2.8 reports the growth rates (in %) of the extensive margin, intensive
margin and the relative exports in trade with the rest of the world, between the
100
pre- (1995-2003) and pos-EU (2004-2012) periods. The contribution of each mar-
gin to the total relative export growth is given in squared brackets. The overall
sample averages are reported in Table 2.11. The values in Table 2.11 are calculated
by averaging the growth rates of the extensive margin, intensive margin and the
relative exports over the sample countries except for Turkey. Average contribu-
tion of each margin is calculated in the same manner and reported in squared
brackets. For instance, the relative exports of Czech Republic to the rest of the
world grew by 62.2% after joining the EU, whereas the extensive and intensive
margins of its exports grew by 16.3% and 45.9%, respectively. The contribution
of the extensive margin to relative export growth was 26.2%, while the intensive
margin accounted for 73.8% of the total export growth.
I begin by looking into the relative exports with respect to the rest of the
world. According to Table 2.8, relative exports grew between 14.0% and 95.8%,
with an average of 61.3%, after joining the EU. The established patterns in the
previous sections are confirmed by the decompositions as well. The contribu-
tion of extensive margin to the relative export growth ranges between 26.2% and
209.7%. For five out of eight countries in the sample, the contribution of exten-
sive margin is greater than the one of intensive margin. Thus, it can be said that
the extensive margin is not only a significant factor, but also the major driving
force behind the relative export growth of CEECs between pre- and post-EU peri-
ods. On average the extensive margin accounted for 80.5% of the relative export
growth, while the intensive margin contributed only 19.5%.
When the relative exports to the EU and non-EU markets are studied sepa-
rately, I find that the extensive margin plays a more important role in increasing
the relative exports to non-EU countries. This does not mean that in trading
101
Table 2.9
Decomposition of Relative Export Growth with the EU Partners (%)
Country Extensive Margin Intensive Margin Relative Exports
Czech rep. 8.5 68.6 77.1
[11.0] [89.0]
Estonia 25.4 8.9 34.3
[74.1] [25.9]
Hungary 8.5 44.9 53.4
[15.8] [84.2]
Latvia 75.0 10.6 85.6
[87.6] [12.4]
Lithuania 48.9 45.9 94.8
[51.5] [48.5]
Poland 62.0 27.4 89.4
[69.4] [30.6]
Slovakia 22.6 79.2 101.8
[22.2] [77.8]
Slovenia 25.3 -7.5 17.7
[142.4] [-42.4]
Turkey 20.6 25.2 45.8
[45.0] [55.0]
Notes: Table reports the growth rates (in %) of the extensive margin, intensive margin and the relative exports in trade
with the EU member countries, between the pre- (1995-2003) and pos-EU (2004-2012) periods. The contribution of each
margin to the total relative export growth is given in the squared brackets. Relative exports are calculated as the ratio of
total exports (to the EU partners) to the total world exports. Numbers are based on the author’s calculations from UN
COMTRADE data.
with the EU members, the extensive margin is insignificant. The average growth
in extensive margin is 34.5% in response to an increase in the relative exports
to the EU members of 84.1%. In the mean time, the average intensive margin
growth is 34.8%. The empirical evidence presented in Table 2.9 through Table
2.11 reveal that on average, the extensive margin accounts for 59.3% of the rel-
ative export growth to the EU members, while intensive margin accounts for
40.7%. On the other hand, the role of the extensive margin in the relative export
growth to non-EU countries is more dominant. On average the relative exports to
non-EU countries increase by 69.5% between pre- and post-EU periods. During
this time, the extensive margin growth is 60.6%, with a median contribution of
102
Table 2.10
Decomposition of Relative Export Growth with the Non-EU Partners (%)
Country Extensive Margin Intensive Margin Relative Exports
Czech rep. 29.1 31.1 60.3
[48.4] [51.6]
Estonia 40.7 21.8 62.5
[65.1] [34.9]
Hungary 30.4 29.8 60.2
[50.6] [49.4]
Latvia 79.3 -14.2 65.1
[121.8] [-21.8]
Lithuania 29.7 14.0 43.7
[68.1] [31.9]
Poland 173.1 -35.3 137.8
[125.6] [-25.6]
Slovakia 58.3 33.9 92.1
[63.3] [36.7]
Slovenia 43.9 -19.8 24.1
[182.0] [-82.0]
Turkey 42.4 12.5 54.9
[77.2] [22.8]
Notes: Table reports the growth rates (in %) of the extensive margin, intensive margin and the relative exports in trade
with the non-EU countries, between the pre- (1995-2003) and pos-EU (2004-2012) periods. The contribution of each mar-
gin to the total relative export growth is given in the squared brackets. Relative exports are calculated as the ratio of total
exports (to the non-EU partners) to the total world exports. Numbers are based on the author’s calculations from UN
COMTRADE data.
66.6%. The reason I prefer using the median is that the distribution of extensive
margin contribution among the sample countries is quite skewed. For example,
in Slovenia, extensive margin accounts for 182.0% of the 24.1% relative export
growth to Non-EU world, while in Czech Republic, it contributes 48.4% of 60.3%
growth.
Overall, I find that, though the extensive margin contributes a significant
amount to the relative export growth to the EU partners, it contributes even more
to the growth of relative exports to the non-EU countries, during the post-EU
period. This result contradicts the findings of Kehoe and Ruhl (2013). They find
that while the extensive margin accounts on average 9.9% of the trade growth
103
Table 2.11
Summary Statistics: Decomposition of Relative Export Growth (%)
Mean Median St. Dev.
Rest of the World Relative Exports 61.3 66.8 26.6
Extensive Margin [80.5] [64.1] 60.2
Intensive Margin [19.5] [34.9] 60.2
EU Partners Relative Exports 84.1 84.2 0.3
Extensive Margin [59.3] [60.5] 44.2
Intensive Margin [40.7] [39.5] 44.2
Non-EU Partners Relative Exports 69.5 69.7 0.8
Extensive Margin [90.6] [66.6] 47.6
Intensive Margin [9.4] [33.4] 47.6
Notes: Table reports the average growth rates (in %) of the extensive margin, intensive margin and the relative exports in
trade with the rest of the world, EU-member countries and non-EU countries, between the pre- (1995-2003) and pos-EU
(2004-2012) periods, over the entire sample. The average contribution of each margin to the total relative export growth
is given in the squared brackets. Numbers are based on the author’s calculations from UN COMTRADE data.
between US and NAFTA partners, it only accounts for 2.4% of the total trade
growth between US and Germany, Japan and the United Kingdom. This is not
surprising when it is considered that the US and these countries have strong and
a long standing trade relationship. CEECs on the other hand, started to imple-
ment trade liberalization policies not too long before 1995. Therefore, one may
argue that for these countries, joining the EU does not only mean to remove trade
barriers with the EU members, but also it enables them to implement proper
trade liberalization and customs policies with respect to the rest of the world. As
a result, they experience a significant rise in their exports to the non-EU countries
as well.
Furthermore, the EU distinguishes from the other free trade agreements.
It does not only impose tariff reductions. It requires the candidate states to
adopt and implement economic, political and legal reforms that correspond to
the EU criteria. This means that each candidate state goes under major structural
change in terms of political, economic and legal institutions. Assuming that these
104
changes serve as an improvement of the overall system, the accession to the EU
not only represents more liberal trade policies, but also a better structured legal
system and standardized customs environment. Hertel et al. (2001) shows that
automating customs procedures between Japan and Singapore will increase trade
flows between these countries as well as with the rest of the world. In another
study, Wilson et al. (2003) investigate the relationship between trade facilitation
and trade flows in the Asia-Pacific region, and they find that improvements to
the customs environment significantly expand trade. In light of these findings,
it is expected that the EU accession have a greater impact on the trade of CEECs
with the rest of the world, than on the trade of NAFTA members.
2.7 Conclusion
In this paper, I present an empirical assessment of the effects of EU member-
ship on the export patterns of Central and Eastern European countries. Using
disaggregated trade data, covering the period 1995-2012 and 248 partner coun-
tries, I decompose exports of CEECs into their extensive and intensive margins.
I find that large growth rates in exports between the pre- and post-EU peri-
ods, are strongly associated with large growth rates in the extensive margin.
My results do not depend on the methodology used in measuring the extensive
margin. Both unweighted and weighted measures indicate substantial changes
in the extensive margin. However, this extensive margin growth mainly comes
from the startling increase in the number of importers, rather than the growth of
the mix of the goods exported. When different destination markets are ignored,
the extensive margin growth falls sharply.
105
My empirical analysis of the gravity model further underlines the impor-
tance of the EU accession on the extensive margin growth. The positive and
statistically significant coefficient on the EU dummy variable proves that the
observed growth of the extensive margin coincides with the accession of the
CEECs to the EU. I have also showed that the extensive margin is an important
contributor to the relative export growth experienced during the trade liberaliza-
tion episodes. The results indicate that on average, 80% of the relative export
growth is attributed to the growth in the extensive margin, which makes it the
driving force behind the export growth of the CEECs.
Recent trade theory makes use of models with firm heterogeneity and mar-
ket penetration costs such as Krugman (1979), Melitz (2003) and Helpman et al.
(2008). These models predict large variety gains from trade liberalization through
reductions of fixed trade costs. My results support the predictions of these mod-
els and suggest that economists should strongly consider the extensive margin
when analyzing the effects of free trade agreements. Also, I believe that the
policy-makers seeking to improve trade performance can benefit from the results
presented in this paper. I provide strong evidence that free trade agreements
generate large gains along the extensive margin, which in return increases the
aggregate exports.
In assessing the effects of trade liberalization, one should also consider that
the underlying motivation behind the formation of free trade areas and imple-
mentation of trade liberalization policies. These policies are dictated by the pre-
diction of the standard endogenous growth models that new product creation
results in higher productivity, income and economic growth.
106
Although we can test the variety gains from trade liberalization empirically,
it is difficult to relate these variety gains to the economic growth. The main rea-
son behind this adversity is data limitations. Though highly disaggregated trade
data became available for many countries, disaggregated domestic production
data is still scarce. Therefore, documenting the correlation between the exten-
sive margin growth and the changes in the production patterns of the countries
remains a challenge. Still, one cannot help, but asks: how did the domestic
production respond to the significant increase in the extensive margin? Did the
change in the export patterns cause any significant shifts in the production sec-
tors? If so, which sectors in the economy are affected the most and why? These
questions are left for future research efforts.
107
Bibliography
Norman D. Aitken. The effect of the eec and efta on european trade: A temporal
cross-section analysis. The American Economic Review, 63(5):pp. 881–892, 1973.
ISSN 00028282.
Joshua Aizenman and Nancy Marion. Volatility and investment: Interpreting evi-
dence from developing countries. Economica, 66(262):pp. 157–179, 1999. ISSN
00130427.
George Alessandria and Horag Choi. Do sunk costs of exporting matter for net
export dynamics? The Quarterly Journal of Economics, 122(1):pp. 289–336, 2007.
ISSN 00335533.
James E. Anderson. A theoretical foundation for the gravity equation. The Amer-
ican Economic Review, 69(1):pp. 106–116, 1979. ISSN 00028282.
James E. Anderson and Eric van Wincoop. Trade costs. Journal of Economic Liter-
ature, 42(3):pp. 691–751, 2004. ISSN 00220515.
Costas Arkolakis, Svetlana Demidova, Peter J. Klenow, and Andr´ es Rodr´ ıguez-
Clare. Endogenous variety and the gains from trade. The American Economic
Review, 98(2):pp. 444–450, 2008. ISSN 00028282.
Marc Bacchetta, Marion Jansen, Carolina Lennon, and Roberta Piermartini. Expo-
sure to External Shocks and the Geographical Diversification of Exports, chapter 4,
pages 81 – 100. Breaking Into New Markets / The World Bank, 2009.
Scott L. Baier and Jeffrey H. Bergstrand. The growth of world trade: tariffs,
transport costs, and income similarity. Journal of International Economics, 53(1):
1 – 27, 2001. ISSN 0022-1996. doi: http://dx.doi.org/10.1016/S0022-1996(00)
00060-X.
Maria Bejan. Trade openness and output volatility. MPRA Paper 2759, University
Library of Munich, Germany, February 2006.
108
Paul R. Bergin and Ching-Yi Lin. The dynamic effects of a currency union on
trade. Journal of International Economics, 87(2):191 – 204, 2012. ISSN 0022-1996.
doi: http://dx.doi.org/10.1016/j.jinteco.2012.01.005.
Jeffrey H. Bergstrand. The gravity equation in international trade: Some microe-
conomic foundations and empirical evidence. The Review of Economics and
Statistics, 67(3):pp. 474–481, 1985. ISSN 00346535.
Jeffrey H. Bergstrand. The generalized gravity equation, monopolistic compe-
tition, and the factor-proportions theory in international trade. The Review of
Economics and Statistics, 71(1):pp. 143–153, 1989. ISSN 00346535.
Christian Broda and David E. Weinstein. Globalization and the gains from vari-
ety. The Quarterly Journal of Economics, 121(2):pp. 541–585, 2006. ISSN 00335533.
Claudia M. Buch, J¨ org D¨ opke, and Harald Strotmann. Does export openness
increase firm-level output volatility? World Economy, 32(4):531–551, 2009. ISSN
1467-9701. doi: 10.1111/j.1467-9701.2009.01168.x.
Ines Buono and Guy Lalanne. The effect of the uruguay round on the intensive
and extensive margins of trade. Journal of International Economics, 86(2):269 –
283, 2012. ISSN 0022-1996. doi: http://dx.doi.org/10.1016/j.jinteco.2011.11.
003.
Robin Burgess and Dave Donaldson. Can openness to trade reduce income
volatility? Evidence from colonial India’s famine era. LSE Research Online
Documents on Economics 54255, London School of Economics and Political
Science, LSE Library, September 2012.
Olivier Cadot, C´ eline Carr` ere, and Vanessa Strauss-Kahn. Export diversification:
What’s behind the hump? Review of Economics and Statistics, 93(2):pp. 590–605,
May 2011.
Cesar Calderon, Norman Loayza, and Klaus Schmidt-Hebbel. Does openness
imply greater exposure ? Policy Research Working Paper Series 3733, The
World Bank, October 2005.
Eduardo A. Cavallo, Jos´ e De Gregorio, and Norman V . Loayza. Output volatility
and openness to trade: A reassessment [with comments]. Econom´ ıa, 9(1):pp.
105–152, 2008. ISSN 15297470.
Menzie D. Chinn and Hiro Ito. What matters for financial development? capital
controls, institutions, and interactions. Journal of Development Economics, 81(1):
163–192, October 2006.
109
Matthieu Crozet and Pamina Koenig. Equations structurelles de gravit´ e avec
marges intensives et extensives.. Canadian Journal of Economics/Revue canadienne
d’´ economique, 43(1):41–62, 2010. ISSN 1540-5982. doi: 10.1111/j.1540-5982.2009.
01563.x.
John T. Dalton. Eu enlargement and the new goods margin in aus-
trian trade. Available at SSRN: http://ssrn.com/abstract=2335596 or
http://dx.doi.org/10.2139/ssrn.2335596, September 2013.
Alan V . Deardorff. Determinants of bilateral trade: Does gravity work in a neo-
classical world? Working Paper 5377, National Bureau of Economic Research,
December 1995.
Peter Debaere and Shalah Mostashari. Do tariffs matter for the extensive margin
of international trade? an empirical analysis. Journal of International Economics,
81(2):163 – 169, 2010. ISSN 0022-1996. doi: http://dx.doi.org/10.1016/j.jinteco.
2010.03.005.
Julian di Giovanni and Andrei A. Levchenko. Trade openness and volatility. The
Review of Economics and Statistics, 91(3):558–585, August 2009.
William Easterly and Aart Kraay. Small states, small problems? income, growth,
and volatility in small states. World Development, 28(11):2013 – 2027, 2000. ISSN
0305-750X. doi: http://dx.doi.org/10.1016/S0305-750X(00)00068-1.
Jonathan Eaton and Samuel Kortum. Technology, geography, and trade. Econo-
metrica, 70(5):pp. 1741–1779, 2002. ISSN 00129682.
Jonathan Eaton, Marcela Eslava, Maurice Kugler, and James Tybout. Export
dynamics in colombia: Firm-level evidence. Working Paper 13531, National
Bureau of Economic Research, October 2007.
Simon J. Evenett and Anthony J. Venables. Export growth by devel-
oping countries: Market entry and bilateral trade. Available at
http://dev3.cepr.org/meets/wkcn/2/2315/papers/Evenett.pdf, 2002.
Robert C. Feenstra. New product varieties and the measurement of international
prices. The American Economic Review, 84(1):pp. 157–177, 1994. ISSN 00028282.
Jeffrey A Frankel, Ernesto Stein, and Shang-Jin Wei. Regional Trading Blocs in the
World Economic System. Washington, DC : Institute for International Economics,
1997.
Jordi Gal´ ı and Tommaso Monacelli. Monetary policy and exchange rate volatility
in a small open economy. The Review of Economic Studies, 72(3):pp. 707–734,
2005. ISSN 00346527.
110
Alfred V . Guender. Stabilising properties of discretionary monetary policies in a
small open economy. The Economic Journal, 116(508):pp. 309–326, 2006. ISSN
00130133.
Mona Haddad, Jamus Jerome Lim, Cosimo Pancaro, and Christian Saborowski.
Trade openness reduces growth volatility when countries are well diversi-
fied. Canadian Journal of Economics/Revue canadienne d’´ economique, 46(2):765–790,
2013. ISSN 1540-5982. doi: 10.1111/caje.12031.
Gordon Hanson and Chong Xiang. Trade barriers and trade flows with prod-
uct heterogeneity: An application to fUSg motion picture exports. Jour-
nal of International Economics, 83(1):14 – 26, 2011. ISSN 0022-1996. doi:
http://dx.doi.org/10.1016/j.jinteco.2010.10.007.
Elhanan Helpman, Marc Melitz, and Yona Rubinstein. Estimating trade flows:
Trading partners and trading volumes. The Quarterly Journal of Economics, 123
(2):441–487, 2008. doi: 10.1162/qjec.2008.123.2.441.
Thomas W. Hertel, Terrie Walmsley, and Ken Itakura. Dynamic effects of the
”new age” free trade agreement between japan and singapore. Journal of Eco-
nomic Integration, 16(4):pp. 446–484, 2001. ISSN 1225651X.
David Hummels and Peter J. Klenow. The variety and quality of a nation’s
exports. The American Economic Review, 95(3):pp. 704–723, 2005. ISSN 00028282.
Marion Jansen. Income volatility in small and developing economies: Export
concentration matters. WTO Discussion Papers, (2005 IIS 4790-S9.2), 2004.
David Romer Jeffrey A. Frankel. Does trade cause growth? The American Economic
Review, 89(3):379–399, 1999. ISSN 00028282.
Georgios Karras. Trade openness, economic size, and macroeconomic volatility:
Theory and empirical evidence. Journal of Economic Integration, 21(2):pp. 254–
272, 2006. ISSN 1225651X.
Timothy J. Kehoe and Kim J. Ruhl. How important is the new goods margin in
international trade? Journal of Political Economy, 121(2):pp. 358–392, 2013. ISSN
00223808.
Katherine A. Kiel and Katherine T. McClain. House prices during siting decision
stages: The case of an incinerator from rumor through operation. Journal of
Environmental Economics and Management, 28(2):241 – 255, 1995. ISSN 0095-
0696. doi: http://dx.doi.org/10.1006/jeem.1995.1016.
111
So Young Kim. Openness, external risk, and volatility: Implications for the com-
pensation hypothesis. International Organization, 61(1):pp. 181–216, 2007. ISSN
00208183.
Pravin Krishna and Andrei A. Levchenko. Comparative advantage, complexity,
and volatility. Journal of Economic Behavior & Organization, 94(0):314 – 329, 2013.
ISSN 0167-2681.
Paul R. Krugman. Increasing returns, monopolistic competition, and interna-
tional trade. Journal of International Economics, 9(4):469 – 479, 1979. ISSN 0022-
1996. doi: http://dx.doi.org/10.1016/0022-1996(79)90017-5.
Paul R. Krugman. Scale economies, product differentiation, and the pattern of
trade. The American Economic Review, 70(5):pp. 950–959, 1980. ISSN 00028282.
Nuno Lim˜ ao and Anthony J. Venables. Infrastructure, geographical disadvan-
tage, transport costs, and trade. The World Bank Economic Review, 15(3):pp.
451–479, 2001. ISSN 02586770.
Christopher S.P . Magee. New measures of trade creation and trade diversion.
Journal of International Economics, 75(2):349 – 362, 2008. ISSN 0022-1996. doi:
http://dx.doi.org/10.1016/j.jinteco.2008.03.006.
Adeel Malik and Jonathan R.W. Temple. The geography of output volatility.
Journal of Development Economics, 90(2):163 – 178, 2009. ISSN 0304-3878. doi:
http://dx.doi.org/10.1016/j.jdeveco.2008.10.003.
Marc J. Melitz. The impact of trade on intra-industry reallocations and aggregate
industry productivity. Econometrica, 71(6):pp. 1695–1725, 2003. ISSN 00129682.
Purba Mukerji. Trade liberalization and the extensive margin. Scottish Jour-
nal of Political Economy, 56(2):141–166, 2009. ISSN 1467-9485. doi: 10.1111/j.
1467-9485.2009.00478.x.
Garey Ramey and Valerie A. Ramey. Cross-country evidence on the link between
volatility and growth. The American Economic Review, 85(5):pp. 1138–1151, 1995.
ISSN 00028282.
Assaf Razin, Efraim Sadka, and Tarek Coury. Trade openness and investment
instability. Working Paper 8827, National Bureau of Economic Research, March
2002.
Assaf Razin, Efraim Sadka, and Tarek Coury. Trade openness, investment insta-
bility and terms-of-trade volatility. Journal of International Economics, 61(2):285
– 306, 2003. ISSN 0022-1996. doi: http://dx.doi.org/10.1016/S0022-1996(03)
00014-X.
112
Dani Rodrik. Why do more open economies have bigger governments? Journal
of Political Economy, 106(5):pp. 997–1032, 1998. ISSN 00223808.
Kim J. Ruhl. The International Elasticity Puzzle. Working Papers 08-30, New York
University, Leonard N. Stern School of Business, Department of Economics,
2008.
Silvana Tenreyro, Miklos Koren, Milan Lisicky, and Francesco Caselli. Diversifi-
cation through trade, 2014.
J. Tinbergen. Shaping the World Economy: Suggestions for an International Economic
Policy. A Twentieth Century Fund Study. Twentieth Century Fund, 1962.
John S. Wilson, Catherine L. Mann, and Tsunehiro Otsuki. Trade facilitation and
economic development: A new approach to quantifying the impact. The World
Bank Economic Review, 17(3):pp. 367–389, 2003. ISSN 02586770.
113
Appendix A
Appendix to Chapter 1
Table A.1
Correlation Matrix for the Export Diversification (Concentration) Measures
Panel A: Aggregate Measures
EM-
variety
HI-
variety
Theil-
variety
EM-
product
HI-
product
Theil-
product
EM-
market
HI-
market
Theil-
market
EM-variety 1.000
HI-variety -0.569 1.000
Theil-variety -0.750 0.292 1.000
EM-product 0.568 -0.300 -0.450 1.000
HI-product -0.358 0.796 0.008 -0.157 1.000
Theil-product -0.668 0.356 0.676 -0.826 0.201 1.000
EM-market 0.475 -0.245 -0.519 0.361 -0.072 -0.348 1.000
HI-market 0.309 -0.357 0.165 0.032 -0.601 -0.114 -0.039 1.000
Theil-market -0.802 0.483 0.756 -0.397 0.169 0.492 -0.727 -0.080 1.000
Panel B: Sectoral Measures
EM
s
-
variety
HI
s
-
variety
Thei
s
l-
variety
EM
s
-
product
HI
s
-
product
Theil
s
-
product
EM
s
-
market
HI
s
-
market
Theil
s
-
market
EM
s
-variety 1.000
HI
s
-variety 0.113 1.000
Thei
s
l-variety -0.447 0.200 1.000
EM
s
-product 0.544 -0.104 -0.286 1.000
HI
s
-product 0.168 0.784 0.115 -0.150 1.000
Theil
s
-product -0.406 0.276 0.463 -0.626 0.379 1.000
EM
s
-market 0.067 -0.611 -0.415 0.283 -0.581 -0.409 1.000
HI
s
-market 0.154 0.689 0.142 0.022 0.296 0.022 -0.327 1.000
Theil
s
-market -0.226 0.478 0.776 -0.238 0.353 0.350 -0.749 0.288 1.000
Notes: Table reports the correlation matrix of the export diversification (concentration) measures for the entire sample of
countries and time frame. The data are obtained from UN COMTRADE.
114
Table A.2
List of Economic Activity Codes and Names
Code Economic activity
A Agriculture, forestery and fishing
B Mining and quarrying
C10-C12 Manufacture of food products; beverages and tobacco products
C13-C15 Manufacture of textiles, wearing apprel, leather and related products
C16-C18 Manufacture of wood, paper, printing and reproduction
C19 Manufacture of coke and refined petroleum products
C20 Manufacture of chemicals and chemical products
C21 Manufacture of basic pharmaceutical products and pharmaceutical
preparations
C22-C23 Manufacture of rubber and plastic products and other non-metallic
mineral products
C24-C25 Manufacture of basic metals and fabricated metal products, except
machinery and equipment
C26 Manufacture of computer, electronic and optical products
C27 Manufacture of electrical equipment
C28 Manufacture of machinery and equipment n.e.c.
C29-C30 Manufacture of motor vehicles, trailers, semi-trailers and of other
transport equipment
C31-C33 Manufacture of furniture; jewelry; musical instruments, toys; repair
and installation of machinery and equipment
D Electricity, gas, steam and air conditioning supply
E Water supply; sewerage, waste management and remediation
activities
J Information and communication
J58-J60 Publishing, motion picture, video, television programme production;
sound recording, programming and broadcasting activities
J61 Telecommunications
J62-J63 Computer programming, consultancy and information service
activities
M Professional, scientific and technical services
M69-M71 Legal and accounting activities; activities of head offices; management
consultancy activities; architectural and engineering activities;
technical testing and analysis
M72 Scientific research and development
M73-M75 Advertising and market research; other professional, scientific and
technical activities; veterinary activities
R-S R Arts, entertainment and recreation
S Other service activities
Note: The sectoral codes are in accordance with the General Industrial Classification of Economic Activities in the Euro-
pean Communities (NACE), Revision 2.
115
Table A.3
Summary Statistics
Variable Num.
Obs.
Mean St. Dev. Min. Max.
Panel A: Aggregate Variables
GDP/cap volatilty 136 0.050 0.027 0.004 0.118
EM-variety 136 0.412 0.147 0.116 0.705
EM-product 136 0.898 0.059 0.692 0.985
EM-market 136 0.983 0.010 0.928 0.995
HI-variety 136 0.057 0.025 0.026 0.138
HI-product 136 0.109 0.045 0.043 0.239
HI-market 136 0.268 0.052 0.174 0.401
Openness 136 0.970 0.407 0.318 2.052
TOT volatility 136 0.035 0.030 0.003 0.189
Financial openness 136 0.750 0.294 0.164 1.000
FDI volatility 136 -2.895 22.717 -153.553 3.432
Inflation volatility 136 0.628 1.107 0.043 10.555
Exchange rate volatility 136 0.047 0.037 0.002 0.178
Government spending 136 0.196 0.017 0.155 0.239
Population (log) 136 15.472 1.047 14.095 17.470
GDP/cap (log) 136 9.844 0.289 9.053 10.344
Panel B: Sectoral Variables
Output volatility 2062 0.133 0.100 0.002 1.732
EM
s
-variety 2062 0.495 0.237 0.001 1.000
EM
s
-product 2062 0.913 0.144 0.011 1.000
EM
s
-market 2062 0.869 0.177 0.015 0.996
HI
s
-variety 2062 0.179 0.137 0.036 0.997
HI
s
-product 2062 0.313 0.234 0.001 1.000
HI
s
-market 2062 0.324 0.114 0.000 0.997
Theil
s
-variety 2062 1.231 0.354 0.486 3.706
Theil
s
-product 2062 0.257 0.235 0.000 1.917
Theil
s
-market 2062 0.631 0.331 0.115 3.296
Export openness 2062 0.968 11.378 0.000 462.797
Sector share (log) 2062 -3.267 0.924 -9.982 -1.264
Output per worker (log) 2062 4.188 0.821 1.899 7.818
Capital per worker (log) 2062 1.443 0.946 -1.449 4.687
Taxes paid to government 2062 -0.002 0.021 -0.181 0.040
Note: Table reports the summary statistics of the main variables for the entire sample of countries and time frame. The
data are obtained from UN COMTRADE, World Bank, WDI and Eurostat, Economy and Finance database.
116
Table A.4
Variable Definitions and Sources - Aggregate Analysis
Variable Definition & measurement Source
GDP per capita, PPP GDP per capita based on
purchasing power parity
World Bank, WDI
EM-variety Extensive margin of exports
based on varieties
(product-market pairs)
Author’s construction with
UN COMTRADE data
EM-product Extensive margin of exports
based on products (HS-6
Digit)
Author’s construction with
UN COMTRADE data
EM-market Extensive margin of exports
based on destination markets
Author’s construction with
UN COMTRADE data
HI-variety Herfindahl Index for
varieties (product-market
pairs)
Author’s construction with
UN COMTRADE data
HI-product Herfindahl Index for
products (HS-6 Digit)
Author’s construction with
UN COMTRADE data
HI-market Herfindahl Index for
destination markets
Author’s construction with
UN COMTRADE data
Openness Total trade (exports+imports)
divided by GDP
Author’s construction with
UN COMTRADE and World
Bank WDI data
terms-of-trade (TOT) Export value index divided
by import value index
Author’s construction with
World Bank, WDI data
Financial openness Chinn-Ito financial openness
index that measures the
restrictions on cross-border
financial transactions
Chinn and Ito (2006)
Foreign direct investment
(FDI)
Net inflows of investment in
the reporting economy from
foreign investors divided by
GDP
World Bank, WDI
Inflation Annual percentage change in
the CPI
World Bank, WDI
Exchange rate Real effective exchange rate
(deflator: unit labour costs in
the total economy - 18
trading partners - Euro Area)
Eurostat
Government spending General government final
consumption expenditure
divided by GDP
World Bank, WDI
Population Total population World Bank, WDI
Note: Volatility is measured as the rolling relative standard deviation (
s
x
m
x
) over the period [t, t+2].
117
Table A.5
Variable Definitions and Sources - Sectoral Analysis
Variable Definition & measurement Source
Output Sectoral output Eurostat, Economy and
Finance database
EM-variety Extensive margin of sectoral
exports based on varieties
(product-market pairs)
Author’s construction with
UN COMTRADE data
EM-product Extensive margin of sectoral
exports based on products
(HS-6 Digit)
Author’s construction with
UN COMTRADE data
EM-market Extensive margin of sectoral
exports based on destination
markets
Author’s construction with
UN COMTRADE data
HI-variety Herfindahl Index for
varieties (product-market
pairs) exported
Author’s construction with
UN COMTRADE data
HI-product Herfindahl Index for
products (HS-6 Digit)
exported
Author’s construction with
UN COMTRADE data
HI-market Herfindahl Index for
destination markets of
exports
Author’s construction with
UN COMTRADE data
Theil-variety Theil Index between
component for varieties
(product-market pairs)
exported
Author’s construction with
UN COMTRADE data
Theil-product Theil Index between
component for products
(HS-6 Digit) exported
Author’s construction with
UN COMTRADE data
Theil-market Theil Index between
component for destination
markets of exports
Author’s construction with
UN COMTRADE data
Export openness Total sectoral exports
divided by sectoral output
Author’s construction with
UN COMTRADE and
Eurostat data
Sector share in GDP Sectoral output divided by
GDP
Author’s construction with
Eurostat and World Bank,
WDI data
Output per worker Sectoral output divided by
total sectoral employment
Author’s construction with
Eurostat data
Capital per worker Consumption of fixed capital
divided by total sectoral
employment
Author’s construction with
Eurostat data
Taxes paid to the government Taxes (less subsidies) on
production
Author’s construction with
Eurostat data
Note: Volatility is measured as the rolling relative standard deviation (
s
x
m
x
) over the period [t, t+2].
118
Appendix B
Appendix to Chapter 2
Table B.1
Partner Country List
Aruba Guinea Neutral Zone
Afghanistan Guadeloupe New Zealand
Angola Gambia, The Other Asia
Anguila Guinea-Bissau Oman
Albania Equatorial Guinea Pakistan
Andorra Greece Panama
Netherlands Antilles Grenada Pitcairn
United Arab Emirates Greenland Peru
Argentina Guatemala Philippines
Armenia French Guiana Palau
American Samoa Guam Papua New Guinea
Antarctica Guyana Poland
Fr. So. Ant. Tr Hong Kong Korea, Dem. Rep.
Antigua and Barbuda Heard and McDonald Isl. Portugal
Australia Honduras Paraguay
Austria Croatia Occ.Pal.Terr
Azerbaijan Haiti French Polynesia
Burundi Hungary Qatar
Belgium Indonesia Reunion
Benin India Romania
Burkina Faso British Indian Ocean Ter. Russian Federation
Bangladesh Ireland Rwanda
Bulgaria Iran, Islamic Rep. Saudi Arabia
Bahrain Iraq Fm Sudan
Bahamas, The Iceland Senegal
Bosnia and Herzegovina Israel Yugoslavia
Belarus Italy Singapore
Belgium-Luxembourg Jamaica South Georgia
Belize Jordan Saint Helena
Bermuda Japan Solomon Islands
Bolivia Kazakhstan Sierra Leone
Brazil Kenya El Salvador
Barbados Kyrgyz Republic San Marino
Brunei Cambodia Somalia
Bhutan Kiribati Special Categories
Bunkers St. Kitts and Nevis St. Pierre
(Continued)
119
Table B.1
Partner Country List – continued
Bouvet Island Korea, Rep. Sao Tome
Botswana Kuwait Sudan
Central African Republic Lao PDR Suriname
Canada Lebanon Slovak Republic
Cocos Islands Liberia Slovenia
Switzerland Libya Sweden
Chile St. Lucia Swaziland
China Sri Lanka Seychelles
Cote d’Ivoire Lesotho Syrian Arab Republic
Cameroon Lithuania Turks and Caicos Isl.
Congo, Rep. Luxembourg Chad
Cook Islands Latvia Togo
Colombia Macao Thailand
Comoros Morocco Tajikistan
Cape Verde Monaco Tokelau
Costa Rica Moldova Turkmenistan
Cuba Madagascar East Timor
Christmas Island Maldives Tonga
Cayman Islands Mexico Trinidad and Tobago
Cyprus Marshall Islands Tunisia
Czech Republic Macedonia Turkey
Germany Mali Tuvalu
Djibouti Malta Tanzania
Dominica Myanmar Uganda
Denmark Mongolia Ukraine
Dominican Republic Northern Mariana Islands U.S. Minor Isl.
Algeria Montenegro Unspecified
Ecuador Mozambique Uruguay
Egypt, Arab Rep. Mauritania United States
Eritrea Montserrat Us Msc.Pac.I
Western Sahara Martinique Uzbekistan
Spain Mauritius Holy See
Estonia Malawi St. Vincent
Ethiopia Malaysia Venezuela
Finland Mayotte British Virgin Islands
Fiji Namibia Vietnam
Falkland Island New Caledonia Vanuatu
France Niger Wallis and Futura Isl.
Free Zones Norfolk Island Samoa
Faeroe Islands Nigeria Yemen
Micronesia Nicaragua South Africa
Gabon Niue Congo, Dem. Rep.
United Kingdom Netherlands Zambia
Georgia Norway Zimbabwe
Ghana Nepal Nauru
Gibraltar
120

Abstract (if available)

Abstract This dissertation is composed of two essays, whose objective is to better understand the effects of trade liberalization on the export variety, and the impact of the change in the export structure on the output volatility during liberalization episodes. ❧ In the first chapter, I investigate the effects of trade liberalization on output volatility. Traditional comparative advantage theory such as Heckscher‐Ohlin and the New Trade Theory pioneered by Krugman and Melitz are at odds when predicting the effects of trade liberalization on export diversification. The traditional view is that trade openness results in greater trade specialization and generates output volatility by exposing countries to external shocks. The new trade theory on the other hand implies that trade liberalization leads to growth in the extensive margin of exports. This expansion in the extensive export margin leads to greater export diversification. More diversified markets mean that with more trade liberalization

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Essays on business cycle volatility and global trade

PDF

The impact of economic shocks on firm behavior: insights from three studies

PDF

Trade and legalization in East Asia: government-business collaboration in trade dispute settlement

PDF

Essays on commodity futures and volatility

PDF

Essays on economic modeling: spatial-temporal extensions and verification

PDF

Growth, trade and structural change in low income industrializing economies

PDF

Essays in macroeconomics and macro-finance

PDF

The impact of trade liberlaization on firm performance in developing countries -- new evidence from Pakistani manufacturing sector

PDF

Essays in macroeconomics

PDF

Financial crises and trade policy in developing countries

PDF

Empirical essays on industrial organization

PDF

Essay on monetary policy, macroprudential policy, and financial integration

PDF

Three essays in international macroeconomics and finance

PDF

Essays on effects of reputation and judicial transparency

PDF

Essays on family and labor economics

PDF

Three essays on heterogeneous responses to macroeconomic shocks

PDF

Essays on delegated portfolio management under market imperfections

PDF

Essays on improving human interactions with humans, algorithms, and technologies for better healthcare outcomes

Asset Metadata

Creator Kartalciklar, Bahar (author)

Core Title Empirical essays on trade liberalization and export diversification

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Economics

Publication Date 11/15/2016

Defense Date 09/08/2016

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag empirical macroeconomics,export diversification,extensive margin of exports,international economics,international macroeconomics,International trade,new trade theory,OAI-PMH Harvest,output volatility,trade diversification,trade growth,trade openness

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Dekle, Robert (committee chair), Betts, Caroline (committee member), Nakano, Aiichiro (committee member)

Creator Email bkartalciklar@gmail.com,kartalci@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-673258

Unique identifier UC11336597

Identifier etd-Kartalcikl-4917.pdf (filename),usctheses-c16-673258 (legacy record id)

Legacy Identifier etd-Kartalcikl-4917-0.pdf

Dmrecord 673258

Document Type Dissertation

Rights Kartalciklar, Bahar

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA