International Linkages, Value Added Trade and

International Linkages, Value Added Trade
and LAC Firms’ Productivity
(DRAFT VERSION PREPARED FOR THE ETSG CONFERENCE 2014 )
Pierluigi Montalbano∗
Silvia Nenci†
Carlo Pietrobelli‡
This version: August, 2014
Abstract
This chapter addresses the following research questions: i) are firms characterized by international linkages more productive than other firms? ii) and,
eventually, are those belonging to industries more involved in global production networks even more productive? To answer these questions we combine
the World Bank Enterprise Survey data with the new OECD-WTO Trade
in Value Added (TiVA) data and present three main empirical exercises: i)
an analysis of the productivity premia associated to participation in international trade and presence of inward fdi; ii) a Cobb-Douglas output function
expanded to firms international linkages; iii) a further expanded version of
the above relationship including the TIVA-based indicators of value added
trade and industry participation and position in global production networks.
Our empirical outcomes confirm the presence of a positive causal relationship between participation in international activities and firm performance
in the LAC region. Focusing on four big Latin American countries (Argentina, Brazil, Chile and Mexico) we show that the actual level of position
into global value chains matters as well. These empirical results are relevant
for policy-making. They contribute to help institutions to shed light on the
importance of the involvement in global production networks in increasing
firms performance in LAC countries.
Keywords: International Trade; Trade in Value added; Global value
chains; Firm productivity
JEL: F14; F61; D24; L22; O54
Acknowledgements: We are grateful to Juan Blyde, Christian Volpe
Martincus and Adam Szirmai for insightful comments and suggestions and
participants to the IDB Workshop ”‘Determinants of Firm Performance in
LAC: What Does the Micro Evidence Tell Us?”’, Washington, DC, June 5-6,
2014. The usual disclaimer applies.
∗
Corresponding Author. University of Sussex (UK) and Sapienza University (Italy)
[email protected]
†
University of Roma Tre (Italy). [email protected]
‡
Inter-American Development Bank, USA. [email protected]
1
1
Introduction
One of the key issues in the current empirical debate on the determinants of
firm performance is the influence of international linkages.
The aim of this paper is to study the causal relationship between international linkages and firm performance in the LAC region. The notion of firm
international linkages adopted in this analysis includes two different dimensions: (i) participation in international trade; (ii) presence of inward foreign
direct investments (fdi). To this end, we take advantage of the new firm
level data provided by the World Bank Enterprise Survey (ES). By matching ES firm level data with the new OECD-WTO Trade in Value Added
(TiVA) data set, we are able to combine different levels of data aggregation
to provide a richer picture of the relationship between firm performance and
country/industry actual involvement in international production networks in
the LAC region.
In particular, this paper addresses the following research questions:
• Are firms characterized by international linkages more productive than
other firms?
• And, eventually, are those belonging to industries more involved in
global production networks even more productive?
To derive empirically the causal relationship between firms performance
and their international linkages we provide: i) a static analysis of the productivity premia associated to participation in international trade and presence
of inward fdi; ii) a version of the standard Cobb-Douglas output function
expanded to firms international linkages; iii) a further expanded version of
the above relationship including the TIVA-based indicators of value added
trade as well as the degree and the typology of industry involvement in
global production networks. Moreover, to control for firms’ heterogeneity by
country and industry we add a full set of country (year), industry, and subnational region fixed effects, while to avoid endogeneity bias we present also
instrumental variables and control function versions of the same empirical
estimates.
Our empirical outcomes confirm the presence of a positive causal relationship between participation in international activities and firm performance
in the LAC region. Focusing on four big LAC countries (Argentina, Brazil,
Chile and Mexico) we show that the actual level of position into global value
2
chains matters as well. These empirical results appear relevant for policymaking. They contribute to help institutions to shed light on the importance
of the involvement in global production networks in increasing firms performance in LAC countries and determine how reforms can boost internal and
external competitiveness to take advantage of a dynamic global economy.
The work is organized as follows: Section 2 reviews the literature on
international linkages and firm productivity; Section 3 describes how to trace
the competitiveness of countries by looking at their production of value added
as well as their level of integration in global markets. Section 4 reports some
stylized facts on the main LAC firm characteristics and the relevant TiVA
indicators; Section 5 presents the empirical analysis; Section 6 concludes.
2
International linkages and firm productivity: review
of the literature
Participation in international trade can be an important source of information, knowledge spillover, technology transfers, technical assistance, competitive pressures and other productivity advantages for firms, leading to significant performance improvements (Grossman and Helpman, 1991; Clerides et
al.,1998; Verhoogen, 2007; Fafchamps et al., 2008; Bernard et al. 2003). At
the same time, the presence of FDI and /or multinational firms may achieve
total cost reduction by utilizing low-priced production factors. These positive
effects are called the “learning effects” in global activities.
Looking at the trade side of the international linkages, the “learning-byexporting hypothesis” spurred a large number of empirical studies that seek
to assess the causal effect of exporting at the firm level.1 However, there is
no consensus among scholars on whether such learning effect exists or what
specific factors may be behind it. A comprehensive survey by Wagner (2007)
indicates that the evidence on this learning effect is mixed and unclear, while
a significant positive effect of the export experience on firms productivity has
been found in several studies such as Girma et al.(2004) for UK firms, Van
Biesebroeck (2005) for sub-Saharan African countries, Fernandez and Isgut
(2005) for Colombia, Alvarez and L´opez (2005) for Chile, De Loecker (2007)
for Slovenia, Lileeva and Treer (2007) and Serti and Tomasi (2008) for Italy,
1
The learning effect has not been fully examined theoretically in the literature. The
major exception is Clerides et al. (1998).
3
and Park et al. (2010) for China. Recently, the meta analysis conducted
by Martins and Yang (2009) indicates that the impact of exporting upon
productivity is higher for developing than for developed economies.
Moreover, the direction of causality between openness and firm performance is controversial (see Greenaway and Kneller, 2007). Firms productivity and sunk cost play important roles in the selection mechanism of firms
international activities. These costs discourage less productive firms from
setting their international linkages. Therefore, firms are self-selected to participate in global markets. Such a selection mechanism according to the
level of productivity is called the ’selection effect’ in exporting. The work of
Melitz (2003) - in which he shows that exporting firms have relatively high
productivity - is the theoretical benchmark on the selection mechanism in
firms exporting, while the study of Bernard and Jensen (1999) on US firms
is one of the pioneering empirical works of a vast series of subsequent analyses (L´opez, 2005, Greenaway and Kneller, 2007, and Wagner, 2007 provide
surveys on the topic). Most of the surveyed studies (included Clerides et
´
al., 1998 for Mexico and Colombia and Alvarez
and L´opez, 2005 for Chile)
find that the more productive producers self-select into the export market
(Hayakawa et al., 2012).
The identification of learning effects of FDI is an important issue as well.
The presence of FDI may improve the performance of domestic firms, particularly in the case of inward FDI in the form of cross border mergers and
acquisitions (M&A). Integrated with the superior know-how, human capital and organization of foreign firms, the local advantages (i.e., geographic
advantage, experience in the local market and knowledge of the local institutional environment) of the target domestic firm could translate into enhanced
productivity (Unctad, WIR various years). Helpman et al. (2004) theoretically show that investing firms have relatively high productivity. Several
studies have empirically tested this proposition (see Greenaway and Kneller,
2007, for a survey within this literature). Papers analyzing the learning effect
in investing include, among others, Aitken and Harrison (1999) for MNEs in
Venezuela, Murakami (2005), Kimura and Kiyota (2006), Hijzen et al. (2007)
and Ito (2007) for Japanese MNEs, Navaretti and Castellani (2004) for Italian MNEs, and Hijzen et al. (2006) and Navaretti et al. (2006) for French
MNEs. Studies in this literature do not necessarily succeed in detecting a
positive causal effect of investing on firms productivity. While Navaretti and
Castellani (2004) and Kimura and Kiyota (2006) find significantly positive
4
impacts, Aitken and Harrison (1999), Hijzen et al. (2007) and Ito (2007)
detect a quite small or no positive effect.
Hijzen et al. (2006) and Navaretti et al. (2006) take also into consideration a possible qualitative difference in learning effect due to the presence
of two types of FDI: horizontal and vertical. They find positively significant
enhancements in productivity in the French horizontal FDI but not in its
vertical one.
Other papers propose a specific focus on the impact of MA on firms
performance. They are, among others: Arnold and Javorcik (2005) and
Petkova (2008) for Indonesia, Conyon et al. (2002), Girma (2005), Girma
et al. (2007) and Harris and Robinson (2002) for the United Kingdom,
Bertrand and Zitouna (2008) for France, Salis (2008) for Slovenia, Piscitello
and Rabbiosi (2005) for Italy, Fukao et al. (2006) for Japan and Chen (2011)
for the USA. Most of the above studies find significantly positive impacts
(see Hayakawa et al., 2012 for a survey).
3
Trade in valued added and global value chain (GVC):
definition and measurement
The increasing international fragmentation of production that has occurred in
recent decades has challenged the conventional wisdom on how we look at and
interpret trade. Traditional measures of trade record gross flows of goods and
services each and every time they cross borders leading to a multiple counting
of trade, which may lead to misguided empirical analyses. Furthermore, since
nowadays a wide number of countries has developed comparative advantages
in specific parts of the value chains, standard trade statistics are becoming
much less informative (e.g. they are not able to reveal the foreign added value
of exports producing biased assessments of revealed comparative advantages
for policy-making).
3.1
Tracing trade in value added
Several initiatives and efforts have tried to address the issue of the measurement of trade flows in the context of the fragmentation of world production
by estimating the so called trade in value-added. Value-added reflects the
value that is added by industries in producing goods and services. It is
5
equivalent to the difference between industry output and the sum of its intermediate inputs. Looking at trade from a value-added perspective is also
able to better reveal how upstream domestic industries contribute to exports
as well as the importance of firm participation in global value chains (OECDWTO, 2012). Furthermore, it demonstrates that access to efficient imports
in a world of international fragmentation matters as much as does access to
markets (Ahmad, 2013). (Ahmad, 2013).
A new literature has emerged recently with the idea of tracing the value
added of a country’s trade flows by combining input-output tables with bilateral trade statistics and proposing new indicators (e.g. Hummels et al.,
2001; Johnson and Noguera, 2012a, 2012b; Miroudot and Ragousssis, 2009;
Koopman et al., 2011 and 2014; De La Cruz et al., 2011; Stehrer, 2013).
Interpretation of these indicators and results for individual countries in the
temporal, geographic and industry dimensions are still in progress. In addition, advanced research on constructing appropriate databases has been
recently conducted by the World Trade Organization (WTO) and the Organization for Economic Co-operation and Development (OECD)2 . In this
paper we use data from the new OECD-WTO Trade in Value Added (TiVA)
database. The TiVA database aims at better tracking global production networks and supply chains. It is derived from official national input-output
tables linked together using bilateral trade statistics in goods by industry
and end-use category (BTDIxE) and estimates of bilateral trade flows in
services.3
3.2
Trade in value added and GVC indicators
Our aim is going beyond the information set provided by standard trade
statistics. Specifically, we gather a set of TiVA indicators able to map out
country trade relations and describe the competitiveness of country indus2
A related but separate data initiative is the World Input-Output Database (WIOD)
funded by the European Commission and developed by the University of Groningen, based
on individual countries’ supply-and-use tables (Timmer et al. 2014). Since the coverage
of Latin American countries in this database is limited to two countries (namely Brazil
and Mexico), it was considered unsuitable for this work.
3
The current TiVA version provides 39 indicators for 57 countries (34 OECD countries
and other 23 economies including Argentina, Brazil, China, India, Indonesia, the Russian
Federation, and South Africa) with a breakdown into 18 industries. As in ES the industry
classification adopted is based on the ISIC Rev. 3.1. The time coverage includes the years
1995, 2000 2005, 2008, 2009.
6
tries by looking at their production of value added as well as their level of
integration in global markets. These indicators are the following:
1. The decomposition of the value added embodied in national exports;
2. The participation into GVCs;
3. The position in GVCs.
Concerning the first point, we follow the decomposition of the value added
embodied in national gross exports proposed by Koopman et al. (2011). According to this methodology, gross exports can be decomposed in the following components (see Figure 1):
• Direct domestic value-added embodied in exports of goods and services
(DVA). This reflect the direct contribution made by an industry in
producing a final or intermediate good or service for export (i.e., value
added exported in final goods or in intermediates absorbed by direct
importers) (1a);
• Indirect domestic value added embodied in intermediate exports (IVA).
This reflects the indirect contribution of domestic supplier industries
of intermediate goods or services used in other countries exports (i.e.,
value added exported in intermediates re-exported to third countries).
(1b).
• Re-imported domestic value added embodied in gross exports (RVA).
This reflects the domestic value added that was exported in goods
and services used to produce the intermediate imports of goods and
services used by the industry (i.e., exported in intermediates that return
home)(1c);
• Foreign value-added embodied in gross exports (FVA). This reflects the
foreign value added content of intermediate imports embodied in gross
exports (i.e., other countries domestic value added in intermediates
used in exports) (2 ).
Components 1a, 1b, 1c represent the value of exports that is created
domestically (i.e., the domestic value added - DoVA, see Figure 1), while
component 2 shows the value of exports created abroad. Only components
7
Figure 1:
Gross export decomposition in value added
Gross exports
Domestic Value added
(DoVA)
Foreign Value added
(FVA)
(1)
(other countries DVA in intermediates)
Direct value‐
added exports
(DVA)
Indirect value‐
added exports
(IVA)
(exported final good
and intermediates
absorbed by direct
importers)
(1a)
(exported
intermediates re‐
exported to third
countries)
(1b)
(2)
Re‐imported
value‐added
exports
(RVA)
(exported
intermediates
that return home)
(1c)
Source: adapted from Koopman al. (2011)
1b, 1c and 2 can be thus considered as part of the global value chain framework.
By combining these value-added components it is possible to assess both
the level of participation and whether a country (or industry) is likely located
upstream or downstream in the global production chain. For instance, the
GVC participation index takes into account the indirect domestic value added
exports (IVA) and the foreign value-added exports (FVA) to summarize the
importance of global production chains in country (or industry) exports by
measuring the degrees of participation in GVCs. This index is given by the
sum of the IVA as a share of gross exports and the FVA as a share of gross
8
exports. The higher (lower) the value of the index, the wider (lower) is the
participation of a country in GVCs. To be noted that a high IVA component
shows the importance of domestic production in global value chains while a
high FVA component reveals that the country/industry is deeply embedded
in global value chains but only captures a small part of value added.
To complete information on international integration into global markets,
we present a third index that characterizes the position of country (or industry) exporters in GVCs: the GVC position indicator. It measures the level of
involvement of a country (or industry) in vertically fragmented production.
It is determined by the extent to which the country (or industry) is upstream
or downstream depending on its specialization (Koopman et al., 2011). The
indicator is given by the ratio of the IVA exports and the FVA exports.
At the global level, IVA and FVA equal each other, therefore, the average
IVA/FVA ratio is equal to 1. A ratio larger than 1 indicates the country lies
upstream in the GVC, either by producing inputs for others, either by providing raw materials, or by providing manufactured intermediates or both. A
ratio lower than 1 means the country lies downstream in the GVC, i.e., it will
use a large portion of other countries intermediates to produce final goods
for exports (i.e., it is a downstream processor or assembler adding inputs and
value towards the end of the production process). 4 It follows that the economic significance of this indicator is not unambiguous since both upstream
and downstream positions are associated with heterogeneous situations.
Since two countries can have identical values of the GVC position index
in a given sector while having very different degrees of participation in GVCs,
it is important to look at both these two indicators in order to have a correct
picture of the degree of international integration of a country (or industry)
(Koopman et al., 2011).
4
LAC firms characteristics and trade in valued added
performance: a descriptive analysis
Micro data offer crucial information for understanding the drivers of competitiveness, as aggregate performance depends strongly on firm-level factors
4
To be noted the presence of a caveat in this decomposition at the industry level since
while the value added embedded in a given imported intermediate could travel across
many sectors before it is exported, the adopted decomposition traces only the direct and
the indirect effects.
9
(such as size, ownership and technological capacity). In particular, the micro
dimension provides the necessary tool to analyze determinants of productivity, its distribution within and across sectors, the role of resource misallocation and the relationship with exports (ECB, 2013).
In conducting our empirical exercise we use a subset of the ES database
specifically focused on LAC countries’ firms developed by the World Bank in
collaboration with the Inter-American Development Bank (IDB). It provides
information on the characteristics of firms across various dimensions, including size, ownership, trading status, and performances, and collects data for
14,657 firms and 31 LAC countries. Tables 1A and 2A in the Appendix
present a synthetic view of the international linkages under analysis (i.e. exporting, importing and foreign owned firms) by country and by industry for
the whole LAC sample we use in the empirical exercises.5 In addition, in
order to provide a richer picture of the phenomena under analysis and combine different level of aggregation to map out sources and components of
trade in value added, we use the new OECD-WTO Trade in Value Added
(TiVA) data set by industries (see Section 3). To take advantage of the
joint availability of ES and TiVA data, we focus specifically on the following
countries for which TiVA and ES data are both available for the same fiscal
year: Argentina, Brazil, Chile, and Mexico.6 Looking at both data on firms
and industries from ES and TiVA we can draw a synthetic picture of the
current international linkages of the four LAC countries as well as trade in
value added components and GVC characteristics.
Tables 1-4 present a descriptive analysis of the firms international linkages’ characteristics and TiVA indicators by industries for the above four
LAC countries. This ES LAC sub-sample includes overall 5,120 firms split
up quite homogeneously in the four LACs. Looking at the first five columns
of the tables (ES data section) we detect that overall almost 15% of these
5
The ES uses a stratified random sampling method. The strata are business sector,
location, and firm size. Indicators are representative at the country level but more care
is necessary when interpreting indicators by subgroups since this sampling method does
not stratify by gender of the top manager, exporter status, or ownership. We take this
into account in our empirical exercise by controlling for a full set of industry, country and
region fixed effects.
6
Since the information collected in the surveys refers to characteristics of the firm at
the moment of the survey or to the last completed fiscal year, we use in this empirical
analysis the ES survey for the year 2010 for Argentina, Chile and Mexico and 2009 for
Brazil.
10
firms declare to be exporters7 and their export intensity is on average over
1/3 of their total sales.8 Only 8.5% are foreign owned firms but foreign
investors own on average a significant share (85.4%). Looking at country
level, the four LAC countries present a quite heterogeneous level of firms
internationalization. With regards to international trade, Argentina holds
the highest number of exporting firms (over 27%), while Chile, Mexico and
Brazil lag behind (they show, respectively, 16.7%, 15.2% and 7% of exporting firms). Chile shows the highest export intensity (41.8%), coming before
Mexico (35.5%), Argentina (33%) and Brazil (30.3%). With regards to FDI,
Chile and Argentina present the highest number of foreign owned firms (both
around 13%) while Brazil registers the weakest presence (3.7%). The foreign
ownerships share of these firms is quite high, ranging on average from about
83% in Mexico to nearly 90% in Argentina. As already underlined these
percentages are not necessarily representative of the firms’ population by
country.9
For what concerns trade in value added components, columns from 6 to
10 of Tables 1-4 (TiVA data section) present the complete decomposition
of the overall gross exports by industry depicted in Figure 1 (see Section
3). These columns sum up to 100% of gross exports, thus verifying that
the decomposition is complete. This decomposition provides a more detailed
breakout of domestic value added in exports than are usually available in the
literature and shows heterogeneity across countries and industries in value
added components. Figures 2 and 3 present an international comparison of
the same components with some selected countries that are representative of
industrialized, developing and transition economies (namely the USA, Japan,
Germany, China, India, South Korea, Poland, Turkey, and South Africa).
The two last columns in the TiVA section provides a synthetic view of the role
and position in global production networks by industry, using the indicators
of GVC participation and position illustrated in Section 3.
7
Only direct exporters with exports above 10% of total sales are considered as exporters.
Because of the adopted threshold of 10% of exports on total sales the registered export
intensity is slightly higher than that reported in similar analyses (see, among others,
Lederman, 2010, 2013).
9
This likely source of bias in our sample does not affect by any means the subsequent
empirical estimates. Since the focus of our analysis is on assessing the average causal relationship between firms performance and their international linkages, we use the available
information on firms’ heterogeneity by country and industry purely as control factors.
8
11
12
22
14
174
112
8
42
165
12
13
164
37
1010
Construction
Electrical Equip.
Food Products
Machinery
Other Manufactur.
Business Services
Textiles & Appar.
Transport & Tele.
Transport Equipm.
Wholesale & Reta.
Wood & Paper
Total
276
4
8
4
22
23
2
50
60
5
77
# exporters*
21
30.00
16.33
37.75
30.03
48.75
40.08
33.30
24.30
43.00
23.35
40.80
27.52
29.38
20.19
12.50
3.54
31.61
24.62
41.82
32.68
exp.intensity**
23.24
18.33
29.65
19.76
129
4
11
4
1
12
8
1
9
27
2
1
39
# foreign owned
10
% foreign ownership
91.70
25.20
92.56
20.92
50.00
.
99.50
0.71
94.85
15.63
91.78
18.40
100.00
.
77.50
32.40
78.83
28.89
65.00
.
77.50
26.30
92.18
23.99
87.75
24.50
89.68
22.44
45.77
45.44
74.02
37.37
67.19
38.18
66.41
45.92
41.79
28.91
44.78
51.82
37.6
DVA %
37.5
42.12
40.81
22.9
31.6
22.23
49.27
28.07
41.58
42.07
62.83
31.15
40.94
43.39
*** Indirect VA %
44.8
0.03
0.05
0.01
0.09
0.02
0.03
0.01
0.04
0.03
0.02
0.03
0.01
0.04
RVA %
0.04
12.08
13.71
3.08
30.94
10.57
12.52
5.51
12.46
16.11
8.23
24.04
7.23
18.97
FVA %
17.66
TiVA DATA
Gross export decomposition in value added
100
100
100
100
100
100
100
100
100
100
100
100
100
Total
100
34.63
0.87
1.79
3.19
4.99
2.32
1.20
0.12
0.56
4.25
0.55
0.06
4.12
GVC participation
2.07
1.87
2.30
12.33
0.11
3.35
1.67
185.80
0.60
0.98
0.86
0.55
25.94
1.05
GVC position
1.59
International integration
Note: Note: Firms values are sample means. Standard deviations are reported in italics.
* Only direct exporters (above 10% of total sales). ** Ratio of exports to total sales
*** This measure is obtained from data on the TiVA variable EXGR IDCSH (i.e., Indirect Domestic Value Added %).This reflects the value-added created in upstream
industries providing domestic inputs to the exporting industry
170
# firms
77
Chemicals & Mine
Industries
Basic Metals
ENTERPRISE SURVEY DATA
Firms internationalisation
Firms international linkages and TiVA indicators by industries (Argentina 2010)
Table 1:
13
20
19
187
110
125
72
174
23
5
165
21
1421
Construction
Electrical Equip.
Food Products
Machinery
Other Manufactur.
Business Services
Textiles & Appar.
Transport & Tele.
Transport Equipm.
Wholesale & Reta.
Wood & Paper
Total
216
3
4
2
3
22
5
13
27
16
4
2
74
# exporters*
39
exp.intensity**
41.15
31.71
29.42
22.50
50.00
35.36
38.25
12.34
30.63
25.94
34.44
29.73
45.38
30.72
23.00
15.65
46.41
29.06
50.00
26.46
70.00
14.14
17.50
9.57
15.00
5.00
35.49
27.18
126
1
16
1
1
5
6
1
14
13
2
2
40
# foreign owned
22
% foreign ownership
89.77
19.77
83.13
28.14
80.00
28.28
80.00
28.28
80.77
27.53
100.00
0.00
51.00
.
76.33
38.32
64.00
37.82
30.00
.
100.00
.
74.38
30.49
100.00
.
83.08
26.99
44.59
45.35
75.36
34.87
63.47
39.23
73.29
37.83
38.57
39.15
20.39
.
30.63
DVA %
36.90
32.17
35.42
17.54
31.44
27.09
37.72
19.28
30.69
30.10
45.09
22.65
.
51.01
*** Indirect VA %
38.05
0.15
0.10
0.04
0.24
0.13
0.16
0.05
0.18
0.20
0.11
0.28
.
0.14
RVA %
0.20
23.09
19.13
7.06
33.45
9.31
22.89
7.38
31.30
31.13
15.65
56.68
.
18.23
FVA %
24.85
TiVA DATA
Gross export decomposition in value added
100
100
100
100
100
100
100
100
100
100
100
.
100
Total
100
41.79
0.35
2.15
7.03
0.54
0.86
0.71
1.24
1.11
0.71
15.60
0.01
2.55
GVC participation
3.57
0.38
0.77
6.29
0.06
2.85
0.19
62.11
0.14
0.13
0.10
0.05
.
0.44
GVC position
0.77
International integration
Note: Note: Firms values are sample means. Standard deviations are reported in italics.
* Only direct exporters (above 10% of total sales). ** Ratio of exports to total sales
*** This measure is obtained from data on the TiVA variable EXGR IDCSH (i.e., Indirect Domestic Value Added %).This reflects the value-added created in upstream
industries providing domestic inputs to the exporting industry
328
# firms
172
Chemicals & Mine
Industries
Basic Metals
ENTERPRISE SURVEY DATA
Firms internationalisation
Firms international linkages and TiVA indicators by industries (Mexico 2010)
Table 2:
14
13
6
153
46
897
Transport & Tele.
Transport Equipm.
Wholesale & Reta.
Wood & Paper
Total
150
4
4
1
12
7
1
5
41
1
44
# exporters*
29
65.00
.
28.75
18.87
32.50
26.30
41.80
30.23
15.00
.
65.88
33.73
26.00
9.62
70.00
.
17.86
9.06
29.25
19.95
exp.intensity**
45.10
29.55
28.34
18.50
116
5
18
1
2
7
3
1
20
1
38
# foreign owned
20
74.67
43.88
83.29
23.46
59.00
57.98
50.00
.
76.61
30.94
74.00
25.10
83.13
27.83
70.00
.
84.45
29.12
100.00
.
% foreign ownership
92.95
21.47
84.24
27.89
43.03
30.76
45.88
49.28
41.04
34.85
59.95
48.19
37.32
24.16
78.79
.
28.89
DVA %
37.28
31.19
45.57
36.59
21.70
14.55
33.05
29.88
34.11
29.87
49.69
16.02
.
21.80
***Indirect VA %
41.45
0.05
0.04
0.03
0.07
0.07
0.05
0.02
0.04
0.11
0.04
0.01
.
0.06
RVA %
0.08
25.73
23.63
17.49
28.94
44.34
32.05
10.16
17.66
32.70
26.11
5.18
.
49.25
FVA %
21.19
Gross export decomposition in value added
100
100
100
100
100
100
100
100
100
100
100
.
100
Total
100
TiVA DATA
52.21
4.46
2.62
0.46
5.86
0.29
2.93
0.09
0.59
3.03
0.05
0.10
5.14
GVC participation
1.31
1.83
0.77
2.34
0.49
0.78
0.38
20.30
0.45
0.81
0.07
15.13
.
0.41
GVC position
2.10
International integration
Note: Note: Firms values are sample means. Standard deviations are reported in italics.
* Only direct exporters (above 10% of total sales). ** Ratio of exports to total sales
*** This measure is obtained from data on the TiVA variable EXGR IDCSH (i.e., Indirect Domestic Value Added %).This reflects the value-added created in upstream
industries providing domestic inputs to the exporting industry
116
Textiles & Appar.
18
Machinery
35
175
Food Products
Business Services
3
Electrical Equip.
5
8
Construction
Other Manufactur.
174
# firms
145
Chemicals & Mine
Industries
Basic Metals
Firms internationalisation
ENTERPRISE SURVEY DATA
Firms international linkages and TiVA indicators by industries (Chile 2010)
Table 3:
15
5
120
168
44
1792
Transport & Tele.
Transport Equipm.
Wholesale & Reta.
Wood & Paper
Total
126
1
1
17
32
3
13
15
14
5
19
# exporters*
6
18.29
10.52
10.00
.
10.00
.
30.33
27.01
38.00
40.87
40.71
28.14
23.53
14.27
44.31
34.15
13.33
2.89
34.56
31.31
exp.intensity**
23.17
7.41
27.21
28.53
66
1
2
16
6
4
3
16
4
1
12
# foreign owned
1
100.00
0.00
10.00
.
85.55
27.34
100.00
.
80.00
21.60
91.06
16.55
61.67
30.07
79.75
39.84
62.67
43.09
98.31
6.23
% foreign ownership
89.00
.
84.75
33.17
46.46
48.34
59.84
32.27
59.96
46.27
71.40
49.81
40.27
27.01
39.42
58.78
29.72
DVA %
40.92
44.55
43.95
36.11
53.55
33.99
45.73
24.31
42.63
48.91
66.15
45.31
35.34
55.40
*** Indirect VA %
47.83
0.05
0.05
0.02
0.12
0.03
0.05
0.02
0.05
0.06
0.04
0.06
0.03
0.07
RVA %
0.06
8.93
7.66
4.03
14.05
6.03
7.95
4.26
7.52
10.75
6.80
15.21
5.85
14.81
FVA %
11.19
Gross export decomposition in value added
100
100
100
100
100
100
100
100
100
100
100
100
100
Total
100
TiVA DATA
36.20
1.42
3.64
1.51
2.62
0.68
2.09
0.10
0.76
2.07
0.87
0.14
4.26
GVC participation
4.46
3.01
3.58
17.08
0.27
8.40
2.85
14.17
1.21
0.98
0.83
0.67
5.43
1.44
GVC position
2.96
International integration
Note: Note: Firms values are sample means. Standard deviations are reported in italics.
* Only direct exporters (above 10% of total sales). ** Ratio of exports to total sales
*** This measure is obtained from data on the TiVA variable EXGR IDCSH (i.e., Indirect Domestic Value Added %).This reflects the value-added created in upstream
industries providing domestic inputs to the exporting industry
461
Textiles & Appar.
161
Machinery
109
165
Food Products
Business Services
41
Electrical Equip.
176
31
Construction
Other Manufactur.
193
# firms
118
Chemicals & Mine
Industries
Basic Metals
Firms internationalisation
ENTERPRISE SURVEY DATA
Firms international linkages and TiVA indicators by industries (Brazil 2009)
Table 4:
The reported decomposition shows, on average, a high level of DVA, a
negligible role of RVA and a low level of FVA for the big LAC countries with
the relevant exception of Chile whose FVA accounts, on average, for 1/4 of
the value of its processing exports. In particular, LAC countries present a
higher level of DVA than the average level of both industrial and developing/emerging/transition countries (see Figure 2). This suggests that LAC
domestic production is comparatively less involved into global value chains,
contributing directly to exports and relying less on imported intermediates.
Consistently, the weight of foreign value added content of intermediate imports of our four LAC countries is, on average, lower than that of developing/emerging countries. In this respect Brazil is the lowest performer since
the average share of FVA of Brazilian industries is around 9%. It indicates
that most of Brazilian exports reflect their own domestic value added. The
details on comparisons with some relevant countries in the above regions are
reported in Tables 3A, 4A and 5A in the Appendix.
Regarding the GVC indicators, participation in the GVC at country level
is substantial for Chile and to a lesser extent for Mexico, while the involvement of Argentina and Brazil in the GVC is well below the selected world
competitors, except for South Africa (see Figure 3 and Table 3A-4A in the
Appendix). These outcomes are consistent with similar analyses on LAC
integration into the global production network (see Blyde et al., 2014). The
second column of the international integration section presents the GVC
position index that reflects whether countries are upstream or downstream
in GVCs for specific industries. The higher the value of the index (higher
than 1), the more upstream the country exporters are situated in global
value chains. For instance, Brazil has the highest GVCs position and it is
the most upstream among our LAC countries. In general, with a notable
exception of Mexico (which seems acting as a final producer using inputs
provided by upstream countries), our LAC countries are located upstream
(i.e., away from the final customer) in global value chains, showing a value
of the GVC position index more in line with developed economies, such as
USA and Japan (see Figure 3 and Table 3A in the Appendix) than that of
the developing/emerging and transition countries (see Figure 3 and Tables
4A and 5A in the Appendix).
16
Figure 2:
TiVA international comparison
Figure 3:
GVC international comparison
Note: The values of the GVC participation index are reported on the left hand axis, while the values of
the GVC position index are reported on the right hand axis
17
5
The empirical analysis
The aim of our empirical exercise is to investigate whether LAC firms characterized by international linkages actually perform better, i.e. tend to have
higher productivity than other LAC firms. Specifically, we would like to
look more in depth whether there is a causal relationship between the degree
and typology of involvement in international production networks and firm
performance in the LAC region.
First and foremost, we start presenting static differences in firm productivity premia between exporters and non-exporters, foreign owned enterprises
and domestic owned ones.10 This first empirical exercise is conducted pooling
data for the entire sample of LAC countries included in the ES.
Productivity premia are measured as the coefficients for export and inward fdi dummies in a regression of the form:
θi = α1 + α2 di + ηc + ηr + ηj + i
(1)
where θ is the log of labor productivity11 , di is a set of dummy for exporting firms and firms characterized by foreign ownership (i.e., our proxy
of inward fdi); ηc , ηr , and ηj are, respectively, country, sub-national region12 ,
and industry fixed effects to control for bias due to unobserved factors; is
the error term.
Table 5 shows the outcomes of the OLS estimates of Eq.1. It shows two
main findings. First, the supposed positive relation between international
linkages and firm productivity is confirmed by firm level LAC data. These
findings are in line with the theoretical predictions (Helpman et al., 2004)
that low productivity firms stay in the domestic market while firms with
higher productivity export and/or engaged in FDI stay in the international
market. Second, that the highest productivity firms export and engage in
FDI at the same time. However, if we look at the estimated coefficient
of the interaction dummy this seems not to be the case if the firms are
simultaneously exporting and foreign owned.
10
As common in the literature, we consider firms as foreign owned only if the foreign
ownership is 10 per cent or higher and as exporters only if direct exports are 10 per cent
or higher of total sales.
11
Although labor productivity is a quite imperfect measure of firm productivity, our
cross-sectional data set is not suited to calculate total factor productivity using the standard methodologies.
12
Sub-national data are available only for the most recent rounds of the ES.
18
Table 5: Export and FDI premia:
dependent variable: (ln) labor productivity
exporter
inward fdi
exporter*inward fdi
cons
FE country
FE industry
FE subnational region
No. Obs
R2
Coef.
SE (robust)
Coef.
SE (robust)
0.144***
0.020
0.158***
0.022
0.170***
0.026
0.206***
0.036
-0.079
0.053
1.654***
0.196
1.665***
0.193
yes
yes
yes
yes
no
no
11150
11150
0.05
0.06
***,**,* denote significance at the 1, 5 and 10 per cent level, respectively.
FE country includes fixed effects for different survey rounds for the same country
As already underlined the above are essentially stylized facts which cannot
provide any causal interpretation. Taking advantage of the availability of
the set of firm level co-variates provided by the ES, we can test the above
relationship by controlling for the standard output function with constant
returns to scale Cobb-Douglas technology with labor, capital, and knowledge
as follows:
θi = β1 + β2 ki + β3 zi + β4 di + ηc + ηr + ηj + i
(2)
Eq. 2 adds - with respect to Eq. 1 - the following explanatory variables (all variables are in logs): ki that stands for ”‘capital intensity”’ and zi
that stands for a bundle of firms level observables, namely ’human capital’,
’employment’, ’firm size’ and ’technological innovation’. As in Farole and
Winkler (2012), the latter variable is a dummy that controls whether firms
use technology licensed from a foreign owned company (excluded office software), own internationally recognized quality certification (e.g., iso), use own
website and/or emails to communicate with clients and suppliers.13 To avoid
bias due to unobservable factors we control, as before, for the geographical
13
Crespi et al., 2014 apply the same model but deal with the problem of selectivity bias
and endogeneity in the functions of innovation and productivity.
19
and sectoral location of the firms.14 A full description of the above variables
is provided in Table 6A in the Appendix.
Table 6 shows the empirical outcomes of the base model. It is organized
in three columns. The first column reports the estimates of Eq. 2; the second
column reports the same estimates adding as in Table 5 an interacted term
for firms that are simultaneously exporting and foreign owned; the third column reports the estimates of Eq.2 for the subsample of exporting firms by
substituting the dummy variable for exports with a continuous variable (i.e.,
the value of sales exported directly). As for the export premia, this empirical exercise uses pooled data for the entire LAC data set. The signs of the
relationship between labor productivity and the set of firm level explanatory
variables are significant and consistent with the theory. A positive coefficient
is estimated for the relation between labor productivity, capital intensity, employment and innovation while a negative coefficient is estimated for unskilled
workers (a proxy of human capital) and firm size. This latter outcome suggests the absence of economies of scale for bigger firms additional to the mere
increase in employment levels. Also in this case, our findings are consistent
with the view that exporter and/or foreign owned firms (i.e., characterized
by inward fdi) show, ceteris paribus, higher productivity.
Because of the lack of panel data, our base model cannot avoid further
bias due to unobserved characteristics that are correlated with both firms’
characteristics and firms’ productivity. To this end, we provide additional
empirical estimates for the sub-sample of exporting firms located in the LAC
region by controlling for endogeneity bias in the relation between firms productivity levels and the value of their gross exports. More specifically, in the
ES data set we select some additional explanatory variables (i.e., excluded
instruments) that are supposed to be correlated with LAC firms’ gross exports but not with domestic firms productivity: i.e., ’Average time to clear
imports from customs (days)’ and ’Days to Obtain Import License’. They
can be considered as proxies of international trade obstacles that are negatively correlated with export flows but do not depend on firms productivity.
15
14
For instance, country fixed effects capture also the heterogeneity in prices differences
across countries
15
One can argue that better performing firms are more likely to better prepare trade
documents and shipments and thereby spend less time in customs or in getting a license.
However, in our case, the weak correlation between firm labor productivity and the above
20
Table 6: Base Model:
dep: (ln) labor productivity
ln K intensity
ln Human K
ln Employment
firm size
tech
exporter
inward fdi
(1)
dummies
0.119***
(15.43)
-0.174***
(-13.89)
0.646***
(48.55)
-0.561***
(-21.85)
0.189***
(4.96)
0.0975***
(3.74)
0.108***
(2.82)
exporter*inward fdi
(2)
interaction
0.119***
(15.42)
-0.174***
(-13.89)
0.646***
(48.52)
-0.562***
(-21.87)
0.189***
(4.94)
0.104***
(3.72)
0.127**
(2.46)
-0.0423
(-0.58)
ln export value
cons
FE country
FE industry
FE subnational region
N
R2
3.995***
(11.59)
yes
yes
yes
6438
0.600
4.003***
(11.47)
yes
yes
yes
6438
0.600
(3)
exports
0.0569***
(4.88)
-0.226***
(-11.94)
0.225***
(9.41)
-0.702***
(-15.64)
-0.242
(-1.12)
0.471***
(29.05)
2.990***
(7.98)
yes
yes
yes
1422
0.744
***,**,* denote significance at the 1, 5 and 10 per cent level, respectively.
FE country includes fixed effects for different survey rounds for the same country.
21
Table 7 provides both IV-2SLS and control function (CF) estimates again
for the pooled data (for brevity the first stage estimates are not reported in
the Table). The IV outcomes are robust and significant. Moreover, the
Hansen’s J statistics of over-identifying restrictions - which is consistent in
the presence of heteroskedasticity - does not reject the null hypothesis that
our instruments are valid. However, the Angrist-Pischke (AP) F-statistics
of weak identification is significant only at the 5% level. Since the IV inconsistency actually increases with the number of instruments used, we opt
for a more parsimonious behavior by using only one instrument, namely
’Average time to clear imports from customs’. We further apply the CF
approach which controls for the endogeneity bias by adding directly the estimated residual of the first stage equation to the main regression providing
a CF unbiased estimator which is generally more precise than the IV one
(Wooldridge, 2010). The significance of the CF estimates confirms the above
evidence of the presence of a relationship between trade and firm level productivity as well as the absence of reverse causality.16 This allows us to
assume the absence of reverse causality even when it is not directly testable.
e.g., in the case of our further TiVA estimates where the restricted sample of
4 main LAC countries, due to TiVA data set constraints, would restrict the
number of available observations to a level not consistent with 2SLS testing.
Finally, to provide a more detailed investigation of the linkages between
firm level exports and productivity and specifically address our second research question, we present a further empirical test of Eq. 2 for the subsample of exporting firms by controlling for the decomposition of the value
added embodied in national exports at the industry level as well as the GVC
indicators (see Section 3). It is worth recalling that FVA and IVA are the
key value added components of total exports since they indicate, respectively,
the foreign value added embodied in total exports and the indirect domestic
value added embodied in intermediate exports used in other countries exports. Moreover the ratio between these two components provide a measure
of country/industry relative upstreamness/downstreamness (i.e., the GVC
position index). Since the GVC index of participation is a linear combination of IVA and FVA, the parameters associated to these components of gross
exports are jointly assumed also as indicators of GVC participation.
instruments confirms that these trade obstacles are more related to causes that are external
to firms (i.g., red tape procedures, institutional efficiency, etc.).
16
A lack of significance of the ρ coefficient is normally considered as a reliable test for
the absence of endogeneity bias.
22
Table 7: Instrumental Variables 2SLS and CF (sample restricted to
exporting countries only)
dep: (ln) labor productivity
(1)
(2)
IV
CF
-0.0124 0.0896***
(-0.25)
(5.47)
-0.302*** -0.237***
(-6.37)
(-9.39)
0.0617
0.402***
(0.32)
(14.80)
-0.626*** -0.586***
(-6.90)
(-10.34)
-1.040***
0.177
(-4.12)
(0.61)
0.641***
0.408*
(3.11)
(1.82)
-0.197
(-0.87)
3.231***
0.560
(4.02)
(0.14)
yes
yes
yes
yes
yes
yes
518
1389
0.52
0.67
2
1
0.14
0.05
ln K intensity
ln Human K
ln Employment
firm size
tech
ln export value
ρ
cons
FE country
FE industry
FE subnational region
N
R2
instruments
Hansen J (prob≥ z)
AP (prob≥ F )
***,**,* denote significance at the 1, 5 and 10 per cent level, respectively.
FE country includes fixed effects for different survey rounds for the same country.
23
Table 8: Value added and GVC estimates (sample restricted to exporting countries and four LACs: Argentina, Mexico, Chile and
Brazil)
dep: ntl log labor productivity
(1)
Gross
0.0828***
(3.34)
-0.178***
(-4.92)
0.317***
(10.86)
-0.881***
(-16.08)
.
.
0.461***
(20.97)
ln K intensity
ln Human K
ln Employment
firm size
tech
ln export value
IVA
FVA
Pos
cons
1.201***
(3.22)
yes
yes
yes
392
0.776
FE country
FE industry
FE subnational region
N
R-sq
(2)
GVC
0.0843***
(3.44)
-0.180***
(-5.03)
0.321***
(10.91)
-0.886***
(-15.39)
.
.
0.460***
(21.43)
0.0107
(0.13)
0.0195
(1.03)
0.0449**
(2.62)
1.922***
(6.35)
yes
yes
yes
390
0.777
***,**,* denote significance at the 1, 5 and 10 per cent level, respectively.
24
Table 8 presents the outcomes of the value added and GVC estimates.
Unfortunately, due to data constraints we can run this last test only for a
restricted sample of exporting firms located in the four main LAC countries
for which TiVA data are available (Argentina, Brazil, Chile and Mexico). The
results are fully consistent with the theory and with the outcomes derived
in the previous empirical exercises (the coefficients of the base model are all
significant and show the expected signs). However, they show the absence
of an additional impact of the various components of value added exports
on firm productivity (both in terms of value added embodied in domestic
exports and foreign intermediate imports), once controlled for the causal
impact of gross exports. In other words, these estimates confirm the positive
impact of the international trade participation on productivity at the firm
level but suggest this to be independent from the actual decomposition of
the added value of gross exports by industries. On the contrary, the robust
and positive relationship between firm level productivity and the industry
GVC position suggests that the position of the industry in the global value
chain matters: the higher the upstreamness of its industry, the greater the
impact of its international linkages on the firm productivity performance.
6
Conclusions
This paper addresses two key research questions; i) are firms characterized
by international linkages more productive than other firms? ii) and, eventually, are those belonging to industries more involved in global production
networks even more productive? This empirical analysis provides a richer
picture of the relationship between firms performance and country/industry
actual involvement in international production networks in the LAC region
by combining the new World Bank ES firm level data and the new OECDWTO TiVA data.
Specifically, we first estimate the productivity premia associated to the
participation in trade and the presence of inward fdi, while controlling for
country (year), sector, and sub-national region fixed effects. Second, we analyze the relationship between firm international linkages and productivity
by using a standard output function with constant returns to scale CobbDouglas technology with labor, capital, and knowledge, presenting both OLS,
IV and CF estimates. Third, we run a final test of the same equation ex-
25
panded to account for TIVA-based indicators of value added trade and industry involvement in global production network.
Our empirical outcomes confirm the presence of a positive causal relationship between participation in international activities and firm performance in
the LAC region. Focusing on four big Latin American countries (Argentina,
Brazil, Chile and Mexico) we show that the actual level of position into
global value chains matters as well. These findings are relevant for policymaking since they shed light on the importance of the involvement in global
production networks in increasing firms performance in LAC countries and
determine how reforms can boost internal and external competitiveness to
take advantage of a dynamic global economy. Natural improvements of this
analysis are: to add information on the actual constraints preventing a country from fuller engagement in GVCs; to propose adequate criteria for the
prioritization of different constraints depending on whether a country tries
to go upstream and/or to integrate downstream; to broaden the variety of its
exports and opportunities to attract greater GVC participation by feasible
changes in the business or policy environment.
References
[1] Ahmad, N. (2013) Measuring Trade in Value Added, and Beyond, paper
prepared for the Conference on Measuring the Effects of Globalization
Washington DC, February 28-1 March 2013, mimeo.
[2] Aitken, B. J., and A. E. Harrison. (1999) ”Do domestic firms benefit
from direct foreign investment? Evidence from Venezuela.” American
economic review, 605-618.
[3] Alvarez, R. and R. L´opez (2005) ”Exporting and Performance: Evidence from Chilean Plants”. Canadian Journal of Economics, Canadian
Economics Association, vol.38(4), 1384-1400, November.
[4] Arnold, J.M. and B. Javorcik (2005) Gifted kids or pushy parents? Foreign acquisitions and plant performance in Indonesia. CEPR Discussion
Paper 5065.
[5] Bernard, A. B., and J. B. Jensen (1999). Exceptional Exporter Performance: Cause, Effect, or Both?, Journal of International Economics 47:
1-25.
26
[6] Bernard, A., J. Eaton, B. Jensen and S. Kortum (2003) Plants and
Productivity in International Trade, American Economic Review, Vol.
93, N. 4, pp.1268- 1290.
[7] Bertrand, O. and H. Zitouna (2008) Domestic versus cross-border acquisitions: which impact on the target firms performance? Applied
Economics 40(17): 22212238.
[8] Blyde et al. (2014). The International Fragmentation of Production:
Latin America and the Caribbean in the Era of Global Value Chains.
IDB forthcoming.
[9] Chen, W.(2011) ”The effect of investor origin on firm performance: Domestic and foreign direct investment in the United States.” Journal of
International Economics 83.2, 219-228.
[10] Clerides, S. K., S. Lach, and J. Tybout (1998) Is learning by exporting important? Micro-dynamic evidence from Colombia, Mexico, and
Morocco, Quarterly Journal of Economics 113(3): 903-947.
[11] Conyon, M.J., Girma, S., Thompson, S. and P.W. Wright (2002) The
productivity and wage effects of foreign acquisition in the United Kingdom. Journal of Industrial Economics 50(1): 85102.
[12] Crespi, G., Tacsir E., and F. Vargas (2014). Innovation Dynamics and
Productivity: Evidence for Latin America in Pietrobelli C., Grazzi M.,
and G. Crespi. Determinants of Firm Performance in LAC: What Does
the Micro Evidence Tell Us?. IDB forthcoming.
[13] De La Cruz, J., R. Koopman, and Z. Wang (2011) Estimating Foreign
Value-added in Mexicos Manufacturing Exports US International Trade
Commission Working Paper No. 2011-04.
[14] De Loecker, J. (2007) Do exports generate higher productivity? Evidence from Slovenia. Journal of International Economics, 73(1), 6998.
[15] ECB (2013) Competitiveness research network: first year results, CompNet, June.
[16] Fafchamps, M., S. El Hamine, and A. Zeufack (2008) ”Learning to
export: Evidence from Moroccan manufacturing.” Journal of African
Economies 17.2: 305-355.
27
[17] Farole, T. and D. Winkler (2012) Foreign Firm Characteristics, Absorptive Capacity and the Institutional Framework. The Role of Mediating
Factors for FDI Spillovers in Low- and Middle-Income Countries. The
World Bank Policy Research Working Paper 6265.
[18] Fernandez, A., and A. Isgut (2005) Learning-by-doing, learning-byexporting, and productivity: evidence from Colombia, World Bank policy research working paper.
[19] Fukao, K., Ito, K., Kwon, H.U. and M. Takizawa (2006) Cross-border
acquisitions and target firms performance: evidence from Japanese firmlevel data. NBER Working Paper 12422.
[20] Girma, S. (2005) Technology transfer from acquisition FDI and the absorptive capacity of domestic firms: an empirical investigation. Open
Economies Review 16(2): 175187.
[21] Girma, S., D. Greenaway, and R. Kneller (2004) Does exporting increase
productivity?: a microeconometric analysis of matched firms, Review of
International Economics 12(5): 855-866.
[22] Girma, S., Kneller, R. and M. Pisu (2007) Do exporters have anything
to learn from foreign multinationals? European Economic Review 51(4):
9931010.
[23] Greenaway, D. and R. Kneller (2007) Firm heterogeneity, exporting and
foreign direct investment. Economic Journal 117(517): 134161.
[24] Grossman, G. and E. Helpman (1991) Innovation and Growth in the
World Economy. Cambridge, MA: MIT Press.
[25] Harris, R. and C. Robinson (2002) Note. The Review of Economics and
Statistics 84(3): 562568.
[26] Hayakawa, K., Machikita, T. and F. Kimura (2012) Globalization and
productivity: A survey of firm-level analysis, Journal of Economic Surveys Vol. 26, No. 2, pp. 332350.
[27] Helpman, E., M. J. Melitz, and S. R. Yeaple (2004) Export versus FDI
with Heterogeneous Firms. American Economic Review 94 (1): 300316.
28
[28] Hijzen, A., Inui, T. and Y. Todo (2007) The effects of multinational
production on domestic performance: evidence from Japanese firms.
RIETI Discussion Paper 07-E-006
[29] Hijzen, A., Jean, S. and T. Mayer (2006) The effects at home of initiating production abroad: evidence from matched French firms. Mimeo,
CEPII.
[30] Hummels, D., Ishii J., and K. Yi (2001) The Nature and Growth of Vertical Specialization in World Trade, Journal of International Economics,
54:1.
[31] Ito, Y. (2007) Choice for FDI and post-FDI productivity. RIETI Discussion Paper 07-E-049.
[32] Johnson, R. C. and G. Noguera (2012a) ”Accounting for Intermediates:
Production Sharing and Trade in Value Added” Journal of International
Economics, 86: 224-236.
[33] Johnson, R. C. and G. Noguera (2012b) ”Fragmentation and Trade in
Value Added over Four Decades” NBER Working paper 18186.
[34] Kimura, F. and K. Kiyota (2006) Exports, FDI, and Productivity: Dynamic Evidence from Japanese Firms. Review of World Economics 2006,
Vol. 142 (4).
[35] Koopman, R., W. Powers, Z. Wang and S.-J. Wei (2011) Give credit
to where credit is due: tracing value added in global production chains,
NBER Working Papers Series 16426, September 2010, revised September 2011.
[36] Koopman, R., Wang, Z. and S.-J. Wei (2014) Tracing Value-Added and
Double Counting in Gross Exports American Economic Review, 104(2):
459-94.
[37] Lederman, D. (2010) An International Multilevel Analysis of Product
Innovation. Journal of International Business Studies 41 (4): 60619.
[38] Lederman, D. (2013) International Trade and Inclusive Growth: A
Primer. Indian Growth and Development Review 6 (1): 88112
29
[39] Lileeva, A. and D. Trefler (2007) Improved access to foreign markets
raises plant-level productivity for some plants. NBER Working Paper
13297.
[40] Lopez, R. (2005) Trade and growth: reconciling the macroeconomic and
microeconomic evidence. Journal of Economic Surveys 19(4): 623648.
[41] Martins, P. S. and Y. Yang (2009) The impact of exporting on firm
productivity: a meta-analysis of the learning-by-exporting hypothesis,
Review of World Economics (2009) 145:431445.
[42] Melitz, M. (2003) The impact of trade on intra-industry reallocations
and aggregate industry productivity. Econometrica 71(6): 16951725.
[43] Miroudot, S. and A. Ragousssis (2009) Vertical Trade, Trade Costs and
FDIOECD Trade Policy Working Paper No. 89.
[44] Murakami, Y. (2005) Are multinational enterprises more productive?
A test of the selection hypothesis. Journal of Asian Economics 16(2):
327339.
[45] Navaretti, B. and D. Castellani (2004) Investments abroad and performance at home: evidence from Italian multinationals. CEPR Discussion
Paper 4284.
[46] Navaretti, B., Castellani, D. and A-C. Disdier (2006) How does investing
in cheap labour countries affect performance at home? France and Italy.
CEPR Discussion Paper 5765.
[47] OECD-WTO (2012) Concept Note, http://www.oecd.org
[48] Park, A., Yang, D., Shi, X., and Y. Jiang (2010) Exporting and firm performance: Chinese exporters and the Asian financial crisis. The Review
of Economics and Statistics, 92(4), 822-842.
[49] Petkova, N. (2008) Does foreign ownership lead to higher firm productivity? Mimeo, University of Oregon, Department of Finance
[50] Piscitello, L. and L. Rabbiosi (2005) The impact of inward FDI on local
companies labour productivity: evidence from the Italian case. International Journal of the Economics of Business 12(1): 3551.
30
[51] Salis, S. (2008) Foreign acquisition and firm productivity: evidence from
Slovenia. The World Economy 31(8): 10301048.
[52] Serti, F. and Tomasi, C. (2008) Self-selection and post-entry effects of
exports: evidence from Italian manufacturing firms. Review of World
Economics 144(4): 66094.
[53] Stehrer, R. (2013) Accounting relations in bilateral value added trade,
wiiw Working Paper, The Vienna Institute for International Economic
Studies, Vienna.
[54] Timmer, M. P., E. Dietzenbacher, B.Los, R. Stehrer, and Gaaitzen
J. de Vries (2014). The World Input-Output Database (WIOD): Contents, Concepts and Applications. GGDC Research Memorandum 144,
Groningen Growth and Development Centre.
[55] Unctad, World Investment Report, (various years), Geneva.
[56] Van Biesebroeck, J. (2005) Exporting raises productivity in sub-Saharan
African manufacturing plants., Journal of International Economics,
67(2), 373391.
[57] Verhoogen, E. A. (2007) Trade, quality upgrading and wage inequality
in the Mexican manufacturing sector. No. 2913. IZA Discussion Papers.
[58] Wagner, J. (2007) Exports and productivity: A survey of the evidence
from firm level data. World Economy, 30(1), 6082.
[59] Wooldridge, J. M. (2010). Econometric analysis of cross section and
panel data. MIT press.
31
A
Appendix
32
33
Total firms
151
975
1010
148
150
149
608
340
1642
1792
984
899
980
845
343
525
150
360
599
360
679
332
153
520
547
94
40
39
72
21
158
72
10
106
119
exporting
29
281
276
21
48
31
74
33
0
126
129
150
102
151
importing
21
329
441
28
60
46
271
84
381
355
393
448
288
384
145
216
9
87
247
84
294
87
20
207
212
foreign
15
139
130
33
29
19
80
45
0
68
74
118
29
77
0
85
35
57
80
62
85
57
26
56
68
exp\&foreign
3
71
78
10
15
9
14
6
0
22
26
55
8
35
0
39
8
13
15
5
36
15
2
21
23
exp\&imp
5
122
162
6
33
7
42
17
0
41
66
99
40
100
0
63
4
28
41
13
109
47
4
63
87
***,**,* denote significance at the 1, 5 and 10 per cent level, respectively.
Country
Antiguaandbarbud
Argentina2006
Argentina2010
Bahamas2010
Barbados2010
Belize2010
Bolivia2006
Bolivia2010
Brazil2003
Brazil2009
Chile2006
Chile2010
Colombia2006
Colombia2010
Costarica2005
Costarica2010
Dominica2010
Dominicanrepubli
Ecuador2006
Ecuador2010
Elsalvador2006
Elsalvador2010
Grenada2010
Guatemala2006
Guatemala2010
Total
Country
Guyana2010
Honduras2006
Honduras2010
Jamaica2010
Mexico2006
Mexico2010
Nicaragua2006
Nicaragua2010
Panama2006
Panama2010
Paraguay2006
Paraguay2010
Peru2006
Peru2010
Stkittsandnevis2
Stlucia2010
Stvincentandtheg
Suriname2010
Trinidadandtobag
Uruguay2006
Uruguay2010
Venezuela2006
Venezuela2010
26423
Total firms
162
433
334
375
1420
1436
470
320
587
362
604
348
536
882
150
150
154
152
366
605
585
500
251
3653
exporting
37
52
25
36
133
216
42
21
77
10
73
37
101
203
26
51
26
19
61
99
110
15
1
8607
importing
51
135
86
81
269
526
212
68
169
31
292
82
217
455
28
31
36
36
88
275
261
0
41
2708
foreign
41
62
38
52
123
127
45
36
71
69
68
38
65
100
31
28
24
9
47
77
63
0
27
Table 1: The LAC sample: exporting, importing and foreign owned firms by country
921
exp\&foreign
16
17
8
9
50
58
10
8
18
5
20
13
24
45
8
13
9
2
14
20
25
0
0
2002
exp\&imp
19
28
15
16
84
137
24
10
24
2
41
20
55
124
11
11
12
5
33
65
67
0
0
Table 2: The LAC sample: exporting, importing and foreign owned
firms by industry
Industries (Isic rev 3.1) Total firms exporting
Basic Metals
1484
253
Business Service
896
122
Chemicals & Mine
3547
685
Construction
1032
75
Electrical Equip
394
94
Food Products
3924
711
Machinery
1082
205
Other Manufactur
1017
86
Other Services
1
0
Textiles & Appar
4400
713
Transport & Tele
547
79
Transport Equipm
354
39
Wholesale & Reta
6474
386
Wood & Paper
883
126
Total
26035
3574
foreign exp*foreign importer
136
74
878
107
32
4
466
207
2222
110
18
3
63
40
252
429
194
1543
86
46
494
22
9
352
1
0
1
196
101
2182
102
25
3
38
16
134
821
97
30
70
33
486
2647
892
8584
***,**,* denote significance at the 1, 5 and 10 per cent level, respectively.
34
exp*imp
196
533
2
82
374
154
53
488
26
4
88
2000
35
39.11
43.78
41.90
49.83
57.83
57.99
29.56
46.98
70.97
52.40
59.02
67.21
45.37
68.31
36.83
71.31
77.96
45.48
52.90
Direct VA %
47.57
42.33
36.72
39.07
28.62
37.53
57.97
38.14
22.15
36.15
35.29
29.03
35.68
24.86
44.88
15.29
19.53
45.91
35.23
Indirect VA %
0.51
0.75
0.69
0.46
0.99
0.15
0.56
0.82
0.20
0.65
0.26
0.14
0.95
0.20
1.24
0.19
0.10
0.42
0.58
Re-imported VA %
12.82
13.14
20.70
10.64
12.55
4.34
11.91
14.05
6.68
10.81
5.42
3.61
18.00
6.63
17.05
13.21
2.41
8.18
11.29
Foreign VA %
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
Total
0.88
2.28
7.35
0.15
4.73
2.32
0.64
1.80
0.96
0.55
0.59
7.22
0.31
2.75
2.75
0.29
3.00
1.25
39.83
GVC participation
1.14
2.96
1.22
218.08
2.14
8.47
0.30
0.94
10.87
0.92
7.87
15.11
0.97
4.63
0.58
36.99
15.37
2.46
2.53
GVC position
50.20
25.43
29.66
.
32.33
63.58
38.80
39.11
30.28
32.62
63.67
61.83
36.91
55.41
27.32
.
64.24
41.81
35.25
Direct VA %
41.21
54.88
49.05
.
49.21
33.20
51.38
49.11
57.88
52.79
30.76
34.31
48.96
37.71
58.10
.
31.05
49.45
49.59
Indirect VA %
0.10
0.23
0.22
.
0.69
0.04
0.11
0.31
0.11
0.27
0.09
0.05
0.27
0.07
0.43
.
0.08
0.11
0.37
Re-imported VA %
8.50
19.46
21.06
.
17.78
3.18
9.71
11.48
11.73
14.33
5.49
3.81
13.86
6.80
14.15
.
4.63
8.63
14.79
Foreign VA %
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
.
100
100
100
Total
0.13
5.76
6.86
0.29
9.06
1.62
0.23
2.44
0.05
1.39
0.56
3.19
0.39
4.69
5.06
0.96
4.33
0.75
47.75
GVC participation
12.14
2.14
1.47
.
1.10
87.97
2.93
0.97
11.71
2.06
89.94
63.42
2.50
7.29
0.65
.
17.64
13.25
2.23
GVC position
International integration
50.74
29.70
30.92
45.81
37.62
50.14
27.51
37.27
51.53
43.59
70.17
63.95
35.81
45.03
21.90
46.63
59.85
37.77
36.59
Direct VA %
Germany
32.59
32.39
31.71
39.52
36.20
41.90
48.77
36.92
33.43
33.64
23.25
27.31
34.86
34.43
41.86
38.32
29.65
40.55
35.73
Indirect VA %
0.44
1.39
0.96
0.54
0.98
0.15
0.60
1.10
0.48
0.83
0.16
0.19
0.94
0.44
1.97
0.34
0.25
0.71
1.03
Re-imported VA %
16.23
36.52
36.42
14.13
25.20
7.81
23.13
24.71
14.56
21.94
6.42
8.55
28.40
20.09
34.28
14.71
10.25
20.97
26.64
Foreign VA %
Gross export decomposition in value added
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
Total
industries providing domestic inputs to the exporting industry.
Note: Indirect VA % is obtained from data on the TiVA variable EXGR IDCSH (i.e., Indirect Domestic Value Added%). This reflects the value-added created in upstream
Agriculture
Basic Metals
Chemicals & Mine
Construction
Electrical Equip
Finance & Insura
Food Products
Machinery
Mining
Other Manufacture
Other Services
Business Service
Textiles & Appar
Transport & Tele
Transport Equipm
Utilities
Wholesale & Reta
Wood & Paper
Total
Industries
Japan
Gross export decomposition in value added
International integration
USA
Gross export decomposition in value added
Gross export decomposition in value added and International integration in Selected Developed Countries (2009)
Table 3:
0.26
5.92
8.85
0.18
3.91
0.98
0.94
5.62
0.20
0.60
0.91
5.72
0.39
2.61
0.75
7.32
2.85
1.45
49.48
1.54
0.74
0.44
12.57
0.59
5.05
0.13
0.46
5.12
0.49
14.48
10.59
0.43
1.79
8.36
0.21
3.58
0.99
0.86
GVC position
International integration
GVC participation
36
58.23
21.53
20.52
23.36
18.05
68.33
23.51
22.35
39.28
30.27
46.76
38.83
20.10
42.80
21.40
30.06
57.37
22.08
23.80
Direct VA %
36.89
43.15
38.04
50.59
37.07
25.25
51.11
39.98
32.36
44.97
36.22
49.66
58.77
40.41
44.45
44.11
33.08
42.59
42.47
Indirect VA %
0.06
0.42
0.50
0.40
2.30
0.09
0.26
0.87
0.39
0.64
0.37
0.29
0.42
0.23
0.67
0.47
0.21
0.49
1.10
Re-imported VA %
4.82
34.89
40.94
25.65
42.58
6.32
25.12
36.79
27.97
24.12
16.65
11.22
20.71
16.55
33.48
25.35
9.34
34.83
32.63
Foreign VA %
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
Total
0.70
4.13
5.73
0.06
17.03
0.61
0.73
3.36
0.82
1.65
0.33
0.85
3.82
0.80
1.64
0.33
2.30
1.15
46.06
GVC participation
21.87
0.54
0.38
0.36
0.19
164.21
0.35
0.17
3.87
0.21
2.42
3.72
0.17
2.29
0.23
18.56
3.33
0.48
0.41
GVC position
80.28
24.33
24.87
.
30.87
77.22
16.59
29.00
77.39
26.27
84.96
69.12
27.66
46.88
27.14
35.92
42.20
37.45
43.57
Direct VA %
16.70
53.40
47.70
.
46.87
16.21
69.89
47.97
16.40
24.39
5.64
16.88
54.32
34.07
49.11
46.75
45.92
47.73
34.41
Indirect VA %
0.01
0.07
0.10
.
0.10
0.02
0.05
0.09
0.03
0.25
0.04
0.05
0.09
0.07
0.10
0.06
0.04
0.07
0.10
Re-imported VA %
3.01
22.20
27.33
.
22.15
6.54
13.46
22.94
6.19
49.09
9.36
13.95
17.93
18.98
23.64
17.28
11.83
14.75
21.92
Foreign VA %
100
100
100
.
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
Total
1.38
2.10
4.51
0.23
2.99
1.31
0.43
0.86
2.13
8.38
0.92
7.51
1.89
2.71
1.07
0.33
3.21
0.29
42.27
GVC participation
14.54
0.90
0.76
.
0.42
9.85
0.26
0.58
8.36
0.10
2.20
1.66
0.43
1.85
0.22
1044.81
2.89
1.28
0.93
GVC position
International integration
55.16
19.26
16.21
40.23
21.39
63.22
18.39
25.31
66.38
37.64
57.13
65.15
30.15
50.04
24.69
36.55
55.84
39.00
27.62
Direct VA %
South Korea
28.57
37.01
22.79
37.33
30.90
29.07
52.86
42.68
19.99
36.64
29.84
23.65
37.72
15.19
38.51
19.74
30.46
37.77
31.26
Indirect VA %
0.07
0.21
0.19
0.14
1.14
0.03
0.11
0.25
0.06
0.17
0.08
0.12
0.27
0.14
0.32
0.09
0.06
0.13
0.48
Re-imported VA %
16.20
43.52
60.81
22.31
46.57
7.67
28.64
31.76
13.58
25.55
12.94
11.08
31.87
34.62
36.49
43.63
13.64
23.10
40.64
Foreign VA %
Gross export decomposition in value added
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
Total
industries providing domestic inputs to the exporting industry.
Note: Indirect VA % is obtained from data on the TiVA variable EXGR IDCSH (i.e., Indirect Domestic Value Added%). This reflects the value-added created in upstream
Agriculture
Basic Metals
Chemicals & Mine
Construction
Electrical Equip
Finance & Insura
Food Products
Machinery
Mining
Other Manufacture
Other Services
Business Service
Textiles & Appar
Transport & Tele
Transport Equipm
Utilities
Wholesale & Reta
Wood & Paper
Total
Industries
India
Gross export decomposition in value added
International integration
China
Gross export decomposition in value added
Gross export decomposition in value added and International integration in Selected Emerging Economies (2009)
Table 4:
0.07
5.32
12.24
0.08
20.42
0.86
0.35
2.76
0.12
0.40
0.29
3.53
1.60
5.10
7.90
0.59
2.80
0.61
65.03
2.21
0.59
0.32
1.81
0.50
9.30
0.08
0.34
57.13
1.57
7.70
11.26
0.67
0.74
0.19
100.45
3.20
2.65
0.60
GVC position
International integration
GVC participation
37
47.49
33.18
35.05
41.06
35.65
54.90
30.26
39.20
62.00
35.44
57.79
60.46
44.04
48.89
28.09
46.19
61.89
37.44
38.22
Direct VA %
35.80
33.91
32.32
37.28
30.12
33.79
50.46
31.23
22.06
38.31
28.96
27.43
26.02
32.22
32.66
34.43
26.52
39.33
33.73
Indirect VA %
0.08
0.24
0.12
0.11
0.17
0.04
0.09
0.19
0.08
0.16
0.06
0.05
0.12
0.09
0.33
0.06
0.05
0.12
0.16
Re-imported VA %
16.64
32.67
32.50
21.54
34.07
11.27
19.20
29.38
15.85
26.09
13.18
12.06
29.81
18.80
38.93
19.32
11.54
23.12
27.89
Foreign VA %
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
Total
0.76
5.24
7.01
1.22
3.96
0.34
1.83
2.75
1.63
1.68
0.65
2.65
1.59
3.39
7.45
0.75
3.59
1.86
48.34
GVC participation
1.15
0.87
0.51
0.97
0.27
5.79
0.16
0.43
3.80
0.24
5.01
4.70
0.18
1.53
0.17
4.13
11.69
0.76
0.73
GVC position
68.42
30.55
32.73
48.77
31.88
65.04
28.27
42.07
65.61
29.21
61.93
67.64
32.09
59.83
35.49
40.49
66.74
34.64
41.66
Direct VA %
22.89
37.14
34.18
34.18
38.21
30.39
57.08
32.47
22.98
42.75
30.12
23.68
49.13
30.83
34.87
25.20
26.59
42.98
36.50
Indirect VA %
0.02
0.08
0.06
0.04
0.07
0.01
0.03
0.07
0.02
0.08
0.02
0.02
0.06
0.02
0.09
0.04
0.01
0.04
0.06
Re-imported VA %
8.68
32.23
33.04
17.01
29.84
4.56
14.62
25.40
11.38
27.96
7.93
8.65
18.72
9.32
29.54
34.26
6.66
22.34
21.79
Foreign VA %
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
Total
0.77
6.81
5.16
0.74
1.53
0.87
0.66
1.86
0.94
0.69
0.44
1.21
4.17
4.12
3.87
0.68
2.73
0.49
37.73
GVC participation
1.49
0.40
0.41
0.43
0.18
4.37
0.11
0.24
5.67
0.23
1.35
11.42
0.54
3.88
0.08
2.86
6.53
0.94
0.73
GVC position
International integration
51.38
30.00
31.89
35.79
35.11
61.27
31.53
.
61.85
46.59
55.24
52.79
33.24
50.71
25.01
54.69
53.82
32.12
47.23
Direct VA %
South Africa
34.97
43.22
46.13
47.15
38.13
34.65
53.76
.
27.87
34.63
33.22
37.63
48.69
36.17
36.52
37.01
38.08
49.82
36.26
Indirect VA %
0.02
0.03
0.02
0.02
0.04
0.00
0.01
.
0.01
0.02
0.01
0.01
0.02
0.01
0.05
0.01
0.01
0.02
0.02
Re-imported VA %
13.63
26.75
21.96
17.04
26.73
4.08
14.69
.
10.27
18.77
11.52
9.57
18.05
13.11
38.42
8.29
8.09
18.04
16.49
Foreign VA %
Gross export decomposition in value added
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
Total
industries providing domestic inputs to the exporting industry.
Note: Indirect VA % is obtained from data on the TiVA variable EXGR IDCSH (i.e., Indirect Domestic Value Added%). This reflects the value-added created in upstream
Agriculture
Basic Metals
Chemicals & Mine
Construction
Electrical Equip
Finance & Insura
Food Products
Machinery
Mining
Other Manufacture
Other Services
Business Service
Textiles & Appar
Transport & Tele
Transport Equipm
Utilities
Wholesale & Reta
Wood & Paper
Total
Industries
Turkey
Gross export decomposition in value added
International integration
Poland
Gross export decomposition in value added
Gross export decomposition in value added and International integration in Selected Developing and Transition
Countries (2009)
Table 5:
0.74
6.42
2.48
0.12
0.45
0.84
0.61
0.00
11.66
0.51
0.55
0.92
0.21
2.68
2.64
0.35
1.75
0.88
33.82
0.12
0.67
0.21
9.14
0.56
15.79
0.09
.
2.03
0.17
6.52
18.55
0.40
2.78
0.04
7.91
1.72
0.63
1.05
GVC position
International integration
GVC participation
Table 6: Variables used in the analysis
Variable name
Dependent variable
Labour productivity
Covariates
Exporter
Inward fdi
K intensity
Human K
Employment
Firm size
Tech
Export value
Excluded instruments
Definition
Sales per worker (US$ 2010)
Firm with at least 10 percent of its annual sales derived from direct exports
Firm with at least 10 percent of ownership held by private foreign investors
Capital stock per worker
Number of full-time unskilled workers at end of the surveyed fiscal year
Number of permanent and temporary full-time workers
Micro (less than 10 employees), small (between 10 and 50), medium (between 50 and 250)
and large enterprises (over 250 employees)
(Technology innovation). tech=1 if firms use technology licensed
from a foreign owned company (excluded office software),
own internationally recognized quality certification (e.g., iso),
use own website and/or emails to communicate with clients and
suppliers, and tech= 0 otherwise
Sales exported directly (% of sales)
Average time to clear imports from customs (days)
Days to obtain import license
38

Download Report