Full Terms & Conditions of access and use can be found at
https://www.tandfonline.com/action/journalInformation?journalCode=cira20
International Review of Applied Economics
ISSN: 0269-2171 (Print) 1465-3486 (Online) Journal homepage: https://www.tandfonline.com/loi/cira20
Specialization and KIBS in the Euro area: a
vertically integrated sector perspective
Davide Antonioli, Claudio Di Berardino & Gianni Onesti
To cite this article: Davide Antonioli, Claudio Di Berardino & Gianni Onesti (2020): Specialization and KIBS in the Euro area: a vertically integrated sector perspective, International Review of Applied Economics, DOI: 10.1080/02692171.2019.1708278
To link to this article: https://doi.org/10.1080/02692171.2019.1708278
Published online: 07 Jan 2020.
Submit your article to this journal
View related articles
Specialization and KIBS in the Euro area: a vertically
integrated sector perspective
Davide Antonioli a, Claudio Di Berardino band Gianni Onesti b
aDepartment of Economics and Management, University of Ferrara, Ferrara, Italy;bDepartment of Neuroscience and Imaging, University of G. d’Annunzio, Chieti-Pescara, Italy
ABSTRACT
The imbalances among countries belonging to the European Monetary Union (EMU) have been analysed under several angles in recent years, but often neglecting the evolution of economic and productive struc-tures. In this work, we aim tofill this gap analyzing the countries specialization through the differences in the inter-industrial linkages that affect economic systems competitiveness and production pro-cesses. We use the input-output subsystem approach exploiting the latest WIOD release (2018) to investigate the role of business services, with a special focus on KIBS, in shaping the EMU countries productive structures through their integration in the manufacturing sectors. The results show that disparities are growing in the composition of pro-ductive structure and they are even more pronounced when we con-sider intersectoral dynamics; in particular, when KIBS are addressed to satisfy the manufacturing final demand and when we control for manufacturing subsystem technological intensity.
ARTICLE HISTORY Received 31 May 2019 Accepted 17 December 2019 KEYWORDS
KIBS; manufacturing; input-output; EMU; specialization JEL CLASSIFICATION F45; O14; P5
1. Introduction
In recent years, the attention of scholars and policy makers mainly focused on‘monetary
integration’ and its consequences for the European Monetary Union (EMU) countries,
while less attention has been paid to‘real disintegration’ (Bagnai and Mongeau Ospina
2017).
The studies on imbalances among EMU countries analyse such phenomena under several angles (i.e. productivity growth, institutional characteristics, labour market poli-cies, external trade balance, etc.), but often neglecting the evolution of economic
struc-tures. Among the few contributions on this topic Palan and Schmiedeberg (2010), for
example, provide evidence for European countries about structural divergence/conver-gence looking at intersectoral (agriculture, manufacturing, services) and inter-industry (for branches of agriculture, manufacturing and services) dynamics. In addition, two
works of the European Central Bank (ECB) (MPC 2004; Mongelli, Reinhold, and
Papadopoulos2016) show the changes in the economic structure in European countries,
pointing out not only the relevance for the transmission of monetary andfiscal policies
but also their crucial role as factors determining the degree of resilience in front of negative phases of the business cycles and the potential productivity growth.
CONTACTClaudio Di Berardino c.diberardino@unich.it
Furthermore, the recent economic crisis has led to strong changes that have affected not only the path of income growth but also the structural and technological features of
european countries (Bolea, Duarte, and Sánchez Chóliz2018).
As the European Commission put forward, in order to strengthen the EU competi-tiveness, coming out from a long and deep recession, a renaissance of the European
industry is needed (EC,2014a). The central role of manufacturing has been re-recognized
by several authors that highlight its importance in sustaining productivity growth,
employment and innovation (e.g. Stöllinger et al. 2013; Pianta 2014; Mazzucato et al.
2015). Indeed, the industrial policy has recently gained renewed attention at the EU level.
The role of manufacturing in sustaining EU competitiveness should be considered within
a framework of increasing integration between manufacturing and services (EC2013):
Knowledge Business Services (KIBS) (Desmarchelier, Djellal, and Gallouj 2013; EC,
2014b) in particular. Despite this crucial role of KIBS in potentially triggering the EU industrial renaissance little is known about the dynamic of their integration in manufacturing.
The present work has a twofold value added. On the one hand, we aim atfilling the gap
in the literature about the changes in the productive systems of the EMU member countries in the 2000s, with a focus on KIBS integration in manufacturing. On the other hand, we address this topic by adopting a methodological approach that is different from the traditional analysis. Specifically, following Pasinetti’s input-output subsystem approach, the paper analyses countries’ productive structures through the differences in the inter-industrial linkages that affect economic systems competitiveness and produc-tion processes.
In order to empirically address our objective, we use the WIOD database (latest release) that covers a time span from 2000 to 2014. The period considered also allows us to verify the consequences of the Great Recession (e.g. Archibugi, Filippetti, and Frenz
2013; Eurofound 2014; Garbellini, Marelli, and Wirkierman 2014; Bolea, Duarte, and
Sánchez Chóliz2018) on the process of modification of productive structures, given the
impact that the two crisis had on several dimensions of the EU economies.
The remainder of the paper is organized as follows. The next section has the objective of providing a review of the background literatures that constitute the foundation of the present work and illustrates the works on the integration among manufacturing and services, in order to analyze the vertical integration processes in the different countries
examined. Section 3 describes the methodological approach and Section 4 shows the
results and provide comments and interpretations. The last section is left to conclusions and policy implications.
2. Background literature
2.1. Intermediate linkages and the vertical perspective of the economy
The analysis on structural change has suggested the need for a more accurate interpreta-tion of the importance of manufacturing in the economic system as a whole. In fact, catching up across countries seems to be mostly affected by the intensity and growth rate
The traditional analytical view of productive structures‘disparities’ mainly focused on
a ‘horizontal’ perspective of the economy. However, these analyses are not enough to
understand the complexity of structural convergence, characterised by forces promoting significant changes in productive structures, since the production processes become more vertically disintegrated. This approach considers economic sectors as divided from one another and no interdependence is assumed. In fact, they focus on the distribution of employment across sectors or industries only and do not account for
intermediate linkages (Palan and Schmiedeberg2010). In particular, this approach does
not account for the shifting boundaries between markets, and in-housefirms’ activities
(Franke and Kalmbach2005) can overestimate or underestimate the role of
manufactur-ing and services in the economic system as a whole (Di Berardino and Onesti2018).
Empirical evidence confirms an increased blurriness in the distinction between manu-facturing and services and indicates that the economic activities could be complementary
rather than substitutable (Pilat et al.2006). In a strongly interconnected context, one of
the most important sources of economic growth and of the competitiveness of the economic system increasingly depends on the spillover effects among the different
activities composing the productive structure (Barriero De Souza et al.2016).
In order to take into account of the vertical integration among sectors, a dimension through which it is possible to analyse the dynamic of the economic structure, we use the subsystem concept. The subsystem (or vertically integrated sectors) is an autonomous, vertically integrated production system that includes all the factors that directly and
indirectly contribute to satisfying thefinal demand in manufacturing. It can be shown,
adopting this perspective, that direct effects in manufacturing have a smaller role than the indirect effects that derive from the linkages within the production system. Indeed, some authors have found that organizational restructuring in several European countries has
been achieved through a substantial vertical externalization (Dietrich 1999).
Manufacturingfirms, attempting to improve their organizational efficiency and reduce
costs through the outsourcing of service activities, have played a fundamental role in this
shift (McCarthy and Anagnostou 2004). Consequently, outsourcing to services would
actually reflects an overestimation of the reduction in hours worked in manufacturing
(Momigliano and Siniscalco1982). In this interpretation, outsourcing is a particular kind
of structural change that is merely an extension of manufacturing into nonmanufactur-ing sectors.
However, empirical results to quantify these claims are far from conclusive. Rather,
Montresor and Vittucci Marzetti (2011), show evidence of deindustrialization in a virtual
world composed of the sum of the I-O tables for the main OECD countries. Sarra, Di
Berardino, and Quaglione (2019) indicate that the deindustrialization process affecting
the EU27 is not homogeneous. Great differences exist between the EU15 and the other EU countries and among the technological classes of manufacturing. The deindustriali-zation mainly appears to have been confined within precise technological boundaries throughout the EU, where the low-tech subsystem suffered the largest contraction both in the EU15 and in the EU12, being greatly affected by strong deindustrialization.
Looking at the manufacturing sector we then pose the following question:
Q1. Are the disparities in the composition of manufacturing productive structures among 19 EMU countries growing?
Although these claims about outsourcing have become more important, there are no studies addressed to quantify economic structures convergence by using subsystem approach. We revise the usual approach of analysis based on economic sectors by applying the Krugman specialization index to the I-O subsystem approach. In doing so, we are able to capture the extent of the change in vertical integration between the manufacturing sector and the KIBS. Hence, we can answer the following questions: Q2. Are the changes in productive structures more evident when a subsystem approach is applied than when the traditional sector-based approach is used?
2.2. The relationship between manufacturing subsystem and knowledge-intensive business services
Many authors show that inter-sectoral interaction tends to be more intense between
manufacturing and knowledge-intensive sectors (e.g. Kox2004; Francois, Manchin, and
Tomberger2015; Ciriaci and Palma2016). Furthermore, it has been argued that many
service activities are required by the manufacturing sector to improve efficiency and
innovation through the optimization or the substitution of some employment functions.
For instance, Ciriaci and Palma (2016), using OECD data for some European countries
over the period 1995–2005, demonstrate the existence of a strategic relationship between KIBS and manufacturing subsystems, dominated by technologically advanced subsys-tems. Therefore, the studies on technological change have identified these services as the
key to capture structural transformations (Montresor and Vittucci Marzetti 2011;
Desmarchelier, Djellal, and Gallouj2013; Di Berardino and Onesti2018).
The KIBS are important because they assume a central role in terms of production,
information processing, knowledge diffusion, and innovation (Miozzo and Miles2002).
Indeed, as Mas-Verdú et al. (2011) highlight, the activities come from their own diverse
perspectives. Thus, manufacturing demonstrates a strong tendency to absorb innovative capacity, whereas KIBS are more likely to spread innovation. As mentioned, the integra-tion of these services is a way for manufacturing to acquire knowledge and skills, as well as advantages in exports, value added, product innovation, employment, and
productiv-ity growth (Baker2007; Castaldi2009; Castellacci2010; Dachs et al.2014).‘We remark
that the industry still appears as a significant factor for explaining the economic growth,
even if it is via the demand of industrial firms for KIBS’. (Desmarchelier, Djellal, and
Gallouj2013, 204).
However, the empirical literature on the effects of the transition to a service-based economy has not always relied on an accurate measure of the phenomenon. For instance, the traditional approach does not allow researchers to distinguish KIBS from other services. It is based on an arbitrary classification of service categories that does not accurately reflect the impact of KIBS on the economy, which hinders reliable analysis
(Cheng and Daniels2014). The subsystem approach applied to the I–O analysis is helpful
in answering these questions because it allows the researchers to measure the amount of
increased employment in KIBS that feeds eitherfinal demand or intermediate demand,
avoiding the inaccuracies related to imputation based on their prevailing destination adopted by the traditional approach.
In the following, we explore the contribution of KIBS to the manufacturing and industrial composition by technological intensity. Looking at the manufacturing sub-system we then pose the following questions:
Q3a. What does it happen to the productive structures movements when we consider the
role of‘integrated’ KIBS?
Q3b. Do some peculiarities emerge if we control for manufacturing subsystem technological intensity?
3. Data and methodological framework
In Pasinetti’s approach, an important distinction is made between ‘sector or industry’ and
‘subsystem’. Pasinetti’s subsystem features a generalization of the concept of subsystem
elaborated by Sraffa. As Sraffa stressed (1960, 89): «the commodities forming the gross
product can be unambiguously distinguished as those which go to replace the means of production and those which together form the net product of the system». The economy can be decomposed into as many parts as there are goods that contribute to the net product, «in such a way that each part is in a self-replacing state with a net product of one
commodity only» (Harcourt and Massaro1964, 717). These parts can be called
«sub-systems». The traditional analysis of sectoral interdependence is, by definition, a static
approach, but Pasinetti offers a way of looking at the same economy from a dynamic
point of view. Indeed, «the differences consist only in how economic relations are
classified, but it is possible to switch from one to the other by means of linear algebraic
transformations: the production coefficients of a vertically integrated sectorial model are
a linear combination of the production coefficients of the corresponding input-output
model» (Pomini2012, 55).
As illustrated by Scazzieri (1990), this method provides a powerful simplifying tool,
which is suitable if one wants to identify «one-way» causal linkages within the complex network of economic relationships. In this manner, the analysis of structural change is based on the sequential character of productive processes, identifying a causal relation-ship that is different from those associated with the theory of the stationary state or of
static equilibrium (Scazzieri 1990). Hence, this empirical framework emphasizes
demand-driven growth, in which inter-industrial relationships are identified according
to the linkages between final goods and all the inputs to production. This approach
indicates that each sector is considered on the basis of its contribution to the production
offinal goods.
The I–O tables that present the direct needs of domestic production are a suitable methodology in analysing the intensity of this phenomenon. Following a method
illu-strated by Pasinetti (1973) these tables can be reworked to define an ‘operator’ which is
independent from the relative prices, and able to decompose a vector that expresses an entity classified for sector (based on a classification compatible with that of the I–O
matrix) in a square matrix in which the same entity is remapped from the‘sector’ or
‘branch’ to a ‘sub-system’ (or ‘vertically integrated sector’ or ‘block’). The subsystem is an aggregation that analytically represents all the activities used (directly or indirectly) to
classifying each sector according to the final product, the subsystem identifies the contribution of every single sector or industry within each process of production.
The operator for the conversion from branch (sector or industry) to subsystems is obtained as follows:
LD ¼ ^vð Þ I Að Þ1^e (1)
In the above equation, A is the coefficient matrix derived from the domestic I–O table;
(I-A)−1represents the Leontief’s inverse, the generic element of which, wij, measures the
value of the output of branch i that is directly or indirectly needed to obtain a unitary
value of the output of branch j available forfinal uses; v is a vector, the generic element of
which, vi, represents the labour coefficients, obtained from the ratio between hours
worked and total output at the current prices of branch i; e is a vector, the generic
element of which, ei, represents the output at the current prices of branch i destined for
final demand. The latter includes the domestic consumptions and the total exports. The
symbol‘^’ indicates that the underlying vector is transformed in a diagonal matrix which
has the components of the vector on the principal diagonal and 0 elsewhere. The columns
of LDoperator will represent the number of hours worked in each branch contributing
directly or indirectly to the production of the subsystems.
The data used in the paper were obtained from the Word Input–Output Database. The
WIOD is a time-series of national symmetric input–output tables (industry x industry)
that cover 43 countries and a time span from 2000 to 2014 according to the International
Standard Industrial Classification revision 4 (ISIC Rev. 4). (Dietzenbacher et al.2013; Los,
Timmer, and de Vries2015; Timmer et al.2015). This long period of time allows us to
consider the effects of the economic world crisis. In fact, our elaborations are
distin-guished for the two sub-periods (2000–2008; 2008–2014). The time partition permits a
more complete and accurate view of the general trend. The availability of specific
socio-economic accounts (SEAs) in the WIOD database, which are entirely complementary to
the sectoral classification of national input–output tables, makes the WIOD dataset
suitable for the analysis of the production structure changes, because they contain data on employment, hours worked and value added for each sector of the economies (Sarra,
Di Berardino, and Quaglione2019).
4. KIBS integration in manufacturing and convergence in EMU: where do we stand?
4.1. Production structures of EMU19 countries
We start the analysis with an evaluation of the characteristics of the production structure and of the changes that have taken place since 2000. We break down the production structure of the overall Euro zone into seven industry groups: agriculture, manufacturing,
public utilities, construction, market services (and KIBS1 inside it) and non-market
services. Table 1 presents some interesting facts in this respect and highlights many
differences in the productive composition obtained with a sectoral and a subsystem
approach using data on employees, hours worked or value added (Montresor and
Vittucci Marzetti 2011; Ciriaci and Palma 2016; Di Berardino and Onesti 2018). This
several activities to the economy as a result of intersectoralflows. We can outline two general tendencies with regard to manufacturing and services.
The traditional sectoral approach reveals an‘underestimation’, when compared to the
subsystem approach, in the ability to generate employment, hours worked and value
added in manufacturing (Table 1). Indeed, if we use a subsystem approach, the weight of
manufacturing becomes higher and exceeds the 20% in all aggregates considered: the aim of Horizon 2020 program about the weight of value added in (vertically integrated) manufacturing in the European countries would be already reached (EU Commission
2014a). The large outsourcing of manufacturing and the general reorganization of
production are associated to a ‘minor deindustrialization process’ if we compare it to
the traditional sectoral approach. This suggests that policies should overcome traditional sectoral boundaries that are becoming more blurred, while it should be taken into
account the processes of integration between the different production activities.
When looking at market services we observe a substantial‘overestimation’ using the
sectoral composition because its weights in all the aggregates obtained through a sub-system approach are lower. This trend seems mainly prominent for KIBS, corroborating
the hypothesis that their indirect role to satisfying the final demand of other activities
inside the production system is commonly rather underestimated.
The strong intensity of inter-sectoral linkages demonstrates that the evolution of services, KIBS in particular, and manufacturing is actually symbiotic, and thus, if the growth in services depends on manufacturing, then, at the same time, structural changes
in services affect manufacturing2
.
To conclude, the economic transition to services is accompanying the deindustrializa-tion process, but in this trend, it should be considered the increasing vertical integradeindustrializa-tion between services and manufacturing that can be attributed to the greater complexity of managerial functions, on the one hand, and to strong vertical decentralization by
manufacturingfirms, on the other.
In the following analyses, we provide evidence on the hours worked, instead of value
added or employment. As seen inTable 1the trends are very similar for all the aggregates
(hours worked, employment and value added). Further, changes in manufacturing value added shares are complicated by changes in relative prices. The fall in the relative prices
of industries ‘might make it difficult to pin down the real decline in manufacturing
output, given the limitations of sectoral deflators’ (Tregenna2009, 123). This aspect is a
point in favour of the choice to use hours worked. Moreover, considering the hours
worked has several advantages over employment as stressed by Portella-Carbó (2016), as
they are directly comparable over time and among countries and are not related to institutional arrangements, social conventions, or the length of the working day (i.e. part-time work).
Figure 1 suggests that the amount of hours worked in the EMU19 manufacturing subsystems is decreasing in the period 2000–2014. Therefore, we cannot neglect the
evidence of ‘deindustrialisation’ even when integrated services in manufacturing are
considered. This decline is more pronounced during the crisis. Indeed, we found evidence of structural breaks in 2008 with a collapse for manufacturing of about 14 percentage points. The results are similar for the hours worked in KIBS integrated into manufacturing subsystem, according to a lower manufacturing capability to absorb KIBS during the unfolding of the crisis. Otherwise, the total KIBS subsystem shows a constant
Table 1. Productive structure composition by sectors and sub-systems (EMU19). Sector Employment Hours Worked Value Added 2000 2008 2014 2000 2008 2014 2000 2008 2014 A 4.9 3.7 3.5 2.4 1.9 2.0 2.4 2.1 2.0 M 17.4 15.1 13.9 20.5 17.7 16.3 17.2 16.8 16.4 PU 1.2 1.1 1.1 1.5 1.3 1.5 2.8 2.7 2.7 C 7.4 7.6 6.0 7.8 8.1 5.9 6.6 6.2 4.9 MS 40.3 43.1 44.2 38.5 41.6 42.9 48.7 50.3 51.0 KIBS 11.0 13.4 10.6 9.9 12.3 13.5 11.5 12.4 12.9 NMS 28.9 29.4 31.2 29.3 29.4 31.4 22.3 21.9 23.0 Subsystem A 3.0 2.5 2.2 1.7 1.5 1.5 1.8 1.6 1.6 M 24.6 22.8 20.5 25.7 23.8 21.7 24.4 24.5 22.5 PU 1.2 1.3 1.3 1.4 1.4 1.5 2.0 2.1 2.0 C 9.6 10.0 7.4 10.1 10.6 7.4 9.0 8.9 6.7 MS 30.4 31.7 35.0 29.5 30.9 34.2 37.4 38.0 41.2 KIBS 4.3 5.3 5.8 4.1 5.1 6.8 4.6 5.3 7.2 NMS 31.2 31.8 33.6 31.7 31.8 33.8 25.3 24.9 26.0 Source: Own elaboration. WIOD database.
growth in the economic system, confirming that it is one of the main features of the consolidation of the knowledge economy over recent decades.
4.2. Specialization in the EMU19 countries
The present section examines the cross-country differences in manufacturing
specializa-tion among the EMU19 countries through the implementaspecializa-tion of specializaspecializa-tion index to the subsystem approach.
In so doing we overcome a shortcoming of the current empirical literature (see among
others Palan and Schmiedeberg 2010), which usually neglect inter-sectoral linkages.
More specifically, we apply the Krugman Specialization Index (KSI) to the subsystem approach in order to understand how much of the specialization process is explained by interlinkages among different activities. The KSI is one of the most diffused measures of specialization. It can be seen as a relative specialization that compares one country to a reference group, EMU19 in our analysis. We can define the Subsystem Krugman Specialization Index (SKSI) as follows:
SKSI¼X
i
Vikð Þ Vt ið Þt
(2)
where Vik(t) is the share of subsystem i in country k at time t based on hours worked
and Vi(t) is the share of subsystem i in the European union less country i3. The higher the
index, the more the economic structure of one country deviates from the EMU19 area and the more a country is considered to be specialized.
60 80 100 120 140 160 180
MANUFACTURING KIBS KIBS INTEGRATED
Figure 1.Hours worked in manufacturing subsystem, total KIBS subsystem and KIBS integrated into
manufacturing subsystem. 2000 = 100. EMU19.
Specialization is a growing force in the manufacturing among EMU19 countries, as
shown inTable 2. Indeed, the results by applying the KSI show an increasing trend in
specialization over time, which potentially implies a diverging trend among the EMU19 economies. The specialization process over time seems to be even more evident when we
abandon the sector approach in favour of the subsystem one, as shown inTable 2: the
SKSI increases more, in the period considered, than the specialization index based on
sectors4. When the trend of SKSI is considered over two sub-periods, 2000–2008 and
2008–2014, in order to single out the potential ‘impact’ of the crisis5, it is possible to note
that the specialization process persists over the two time spans; only the pace of specialization changes. This evidence highlights the fact that the subsystem approach seems to be a good alternative for the investigation of specialization because it allows to capture more information about production structure than the traditional sector approach.
The answers to ourfirst research questions (Q1 and Q2) are positive. Over the time
span considered we assisted to an increase of disparities in the trend of KIBS integration and the subsystem approach seems to capture some information that the sector approach is not able to capture.
4.3. KIBS integration in the manufacturing subsystem
The leading role of KIBS as important leverage in manufacturing development is recognized by the increasing use of them as inputs into the production organization in
the EMU countries (see Table A1 in Appendix for the countries code).Figure 2illustrates
the changes in the degree of KIBS vertical integration in the manufacturing subsystem before (left panel) and during the crisis (right panel). This distinction allows us to observe some interesting differences between the two periods. The subsequent analyses allow us
to answer the manifold ‘question marks’ implied by the third research question (Q3a
and Q3b).
In the period 2000–2008 there is a general evidence of growth in almost all countries
(position above the bisector). Despite this common trend, we can observe strong
differences among the groups of countries, which tend to cluster. Germany,
Netherland and France (leading group) are characterized by a strong increase in the integration process of KIBS in manufacturing so as to be above the EMU19 average.
Another group of countries, that we can call ‘following group’ (Italy, Spain, Austria,
Portugal, Ireland, Slovenia, Slovakia) shows a similar increase, but these countries have a lower level of integrated KIBS than the leading group. This following group is formed by
Table 2.KSI (Krugman Specialization Index), SKSI (Subsystem Krugman Specialization Index) and CV
(coefficient of variation) in EMU 19 countries by sector and sub-system manufacturing (2000–2008– 2014). Hours worked.
Deviation 2000–2014 Deviation 2000–2008 Deviation 2008–2014 Manufacturing 2000 2008 2014 2000 = 100 2000 = 100 2008 = 100
KSI 0.73 0.766 0.8 109.6 104.9 104.4
SKSI 0.884 1.113 1.122 126.9 125.9 100.8
Sector CV 0.232 0.272 0.292 125.9 117.2 107.4
Subsystem CV 0.229 0.296 0.335 146.3 129.3 113.2
the second (Italy) and fourth (Spain) most important manufacturing countries in Europe, but also by developing countries such as Slovakia and Slovenia. A further category of
countries can be labelled‘lagging group’ (Greece, Lithuania, Latvia, Estonia and Cyprus)
because they are quite below the average in terms of integrated KIBS in both at the
beginning (2000) and at the end (2008) of thefirst period considered.
In the period 2008–2014 we can observe a general slowdown in the growth of KIBS
integrated in manufacturing subsystem. KIBS integrated decrease also in the ‘leading
group’ countries, especially for Netherland while Germany knows a very low drop. In the ‘following group’, the share of KIBS remains quite steady, with the exception of Spain and Ireland. It should be noted the jump to the top of Spain and the collapse of Ireland. The ‘lagging group’ is still very far behind without evident signs of convergence in the integration of KIBS.
The core/periphery model can be applied to interpret the evidence: the EMU core countries, Germany, France and Netherland (and also Belgium, to some extent), are the ‘leading countries’, while peripheral countries, either southern or eastern ones, are clustered in the other two groups. Despite a slow process of convergence in the KIBS integration, the gaps among groups of countries are still large. The issue of polarization
(Cirillo and Guarascio2015) emerges for EMU countries when the dynamics of KIBS
integration is considered and the‘two peripheries model’ (Celi et al.2018) emerges as
well: southern periphery and eastern periphery. However, the eastern ‘Germanized’
countries (e.g. Slovakia and Slovenia) are becoming more similar in terms of KIBS integration to the southern ones (following group), while it is a third group of non-‘Germanized’ eastern countries (e.g. Latvia and Estonia) that constitute – with Greece,
which is a special case– the ‘lagging group’. It seems we are seeing the emergence of three
peripheries, already caught in the dynamics of manufacturing and integrated KIBS
shares: Southern, Germanized-Eastern, Non-Germanized Eastern countries6.
The followingfigures (Figures 3–5) report the degree of KIBS vertical integration in
the manufacturing subsystem in the 19 EMU countries in 2000 and 2014, considering,
respectively, the manufacturing subsystem as a whole (Figure 3), the low-tech
manufac-turing subsystem (Figure 4) and the high-tech manufacturing subsystem (Figure 5). We
consider the growth rates of hours worked in the two dimensions (manufacturing
Figure 2.Share of hours worked in KIBS in the total manufacturing subsystem 2000–2008 (left), 2008–
2014 (right).
subsystem and integrated KIBS), distinguishing between two sub-periods: the pre-crisis (2000–2008) and the crisis period (2008–2014). Integration between manufacturing and KIBS can be seen as a process that overcomes the traditional boundaries among the factors of production and adjusts the contribution of activities in the economy of each country.
Figure 3 (bottom graphs) illustrates the degree of KIBS vertical integration in the manufacturing subsystem as a whole: the degree of integration of KIBS into manufactur-ing can be assessed through the share of direct and indirect inputs of KIBS in the
production offinal goods. The extent of this integration and the weight of manufacturing
subsystem differ significantly among countries. In the left panel (2000) the countries are closer in terms of manufacturing subsystem weights (X-axis), especially the largest countries in the EMU area and, consequently, their distribution was more vertically
oriented, showing that the main differences in the productive systems were to search in
the integration of KIBS into the manufacturing subsystems (Y-axis). In 2014 (right
Figure 3.Top graphs: rates growth of hours worked in the manufacturing subsystem and in the
integrated KIBS. 2000–2008 (left), 2008–2014 (right). Bottom graphs: KIBS integration in the manu-facturing sub-system (% of hours worked) in 2000 (left) and 2014 (right).
Source: Own elaboration. WIOD database. Notes Top graphs: x = Rates growth of hours worked in the manufacturing subsystem; y = Rates growth of hours worked in the integrated KIBS in the manufacturing subsystem; r is the correlation between rates growth of hours worked in the manufacturing subsystem and in the integrated KIBS; the internal axes represent the average values of the EMU19. Bottom graphs: x = Share of Manufacturing subsystem; y = Integration of KIBS in Manufacturing subsystem; size of circles: represents the size of each country’s manufacturing subsystem in the total manufacturing subsystem in EMU19 in 2014, calculated in terms of total hours worked; r is the correlation between KIBS integration and manufacturing systems weights; the internal axes represent the average values of the EMU19.
panel), on the contrary, the dispersion of the countries over the horizontal line indicates that while the gaps in terms of KIBS integration are closing over time, there is a divergence in the weights of the manufacturing subsystems: with a block of manufactur-ing countries, led by Germany and Italy, followed by Slovenia, Slovakia, Estonia and Lithuania, with evident differences in the level of KIBS integration, and a second block of
more‘de-industrialized’ countries, which is led by France and Spain, with a higher level
of KIBS integration.
We can observe different trends for the four largest countries. France and Spain are
characterized by a strong tertiarization process accompanied by a large decrease in the weight of manufacturing subsystem, while despite a reduction of manufacturing produc-tion in Italy, it remains, with Germany, a leading manufacturing country in EMU. However, only Germany conjugates high manufacturing weight and high level of KIBS integration, reaching a sustainable equilibrium between a high level of growth in KIBS
Figure 4.Top graphs: rates growth of hours worked in the manufacturing subsystem and in the
integrated KIBS. 2000–2008 (left), 2008–2014 (right) in Low-Tech and Medium Low-Tech industries. Bottom graphs: KIBS integration in the manufacturing sub-system for LT and MLT industries (% of hours worked) in 2000 (left) and 2014 (right).
Source: Own elaboration. WIOD database. Notes Top graphs: x = Rates growth of hours worked in the LT-MLT manufacturing subsystem; y = Rates growth of hours worked in the integrated KIBS in the LT-MLT manufacturing subsystem; r is the correlation between rates growth of hours worked in the manufacturing subsystem and in the integrated KIBS in Low-Tech and Medium Low-Tech industries; the internal axes represent the average values of the EMU19. Bottom graphs: x = Share of LT-MLT manufacturing subsystem on total manufacturing subsystem; y = Integration of KIBS in LT-MLT Manufacturing subsystem; size of circles: represents the size of each country’s LT-MLT manufacturing subsystem in the LT-MLT manufacturing subsystem in EMU19 in 2014, calculated in terms of total hours worked; r is the correlation between KIBS integration and manufacturing systems weights; the internal axes represent the average values of the EMU19.
integrated to manufacturing, with the weight of manufacturing in the economy remained unchanged if compared to the pre-crisis period. Finally, the correlation between the KIBS integration and the weight of manufacturing systems reinforces over time confirming that the business services play a leading role in the development of the manufacturing
sector, as shown in different studies (Ciriaci and Palma2016), and they could contribute
to improve the competitiveness of the manufacturing sector and of the overall produc-tion system.
If we examine the sub-periods (Figure 3 – top graphs), we discover an interesting
evidence. In the pre-crisis period (2000–2008) manufacturing growth and KIBS integra-tion went hand-in-hand for the majority of the countries, while in the period of crisis (2008–2014) this ‘marriage’ partially broke, as the drop in the correlation index suggests (see Tables A2 and A3 in Appendix).
Figure 5.Top graphs: rates growth of hours worked in the manufacturing subsystem and in the
integrated KIBS. 2000–2008 (left), 2008–2014 (right) in Medium High-Tech and High-Tech industries. Bottom graphs: KIBS integration in the manufacturing sub-system for MHT and HT industries (% of hours worked) in 2000 (left) and 2014 (right).
Source: Own elaboration. WIOD database. Notes Top graphs: x = Rates growth of hours worked in the HT-MHT manufacturing subsystem; y = Rates growth of hours worked in the integrated KIBS in the HT-MHT manufacturing subsystem; r is the correlation between rates growth of hours worked in the manufacturing subsystem and in the integrated KIBS in Medium High-Tech and High-Tech industries; the internal axes represent the average values of the EMU19. Bottom graphs: x = Share of HT-MHT manufacturing subsystem on total manufacturing subsystem; y = Integration of KIBS in HT-MHT Manufacturing subsystem; size of circles: represents the size of each country’s HT-MHT manufacturing subsystem in the HT-MHT manufacturing subsystem in EMU19 in 2014, calculated in terms of total hours worked; r is the correlation between KIBS integration and manufacturing systems weights; the internal axes represent the average values of the EMU19.
In synthesis, we can identify different models of specialization among the production systems. These differences are one of the element at the basis of the real divergence between the economic systems that we are experimenting in the Eurozone, with its corollary of asymmetric reaction of countries against exogenous shocks, supporting the
Krugman hypothesis (Krugman1993).
Since several authors (Ciriaci and Palma2016; Sarra, Di Berardino, and Quaglione
2019) show that inter-sectoral interaction tends to be more intense in the
medium/high-tech and high-medium/high-tech manufacturing sectors, given that the medium/high-technological progress has
expanded the diffusion of inter-industry linkages in the economic system and the
intermediate stages of the value chains have become increasingly important in final
production (Silva and Teixeira 2008), we replicate the analysis illustrated in Figure 3
for sectors disaggregated by technological intensity. We use the well-known classification
adopted by the OECD (OECD2003) to distinguish our manufacturing subsystems in
Low-Tech (LT), MediumLow-Tech (MLT), MediumHigh-Tech (MHT) and High-Tech (HT) industries. For our aim we then aggregate the four groups in two, LT-MLT and HT-MHT, since it is sufficing to point out the main differences in the trends of integration. When we turn our attention to the subsystems disaggregated in LT-MLT and HT-MHT (Figure 4) we notice several differences accompanying the development over time of the
integration of KIBS in the manufacturing systems (Q3b). As we can see from the bottom graphs there is a negative relation between the weight of the LT-MLT subsystems on the total manufacturing subsystem and KIBS integration: the higher the LT-MLT share, the lower the KIBS integration. LT industries do not develop over time a sensibly higher KIBS integration and the picture of the 2000 is quite similar to that of 2014, except for some small
reallocation of countries7. Despite the evidence provided by the shares of manufacturing
and integrated KIBS, when we look at the growth rates (top-graphs) we see a positive relationship between LT-MLT manufacturing subsystems growth and KIBS integration
growth. This positive relation is undermined by the economic crisis (2008–2014), but
remains quite stable, with some countries’ positions that is worth noting: Spain is the only large country that seems to strongly push KIBS integration in LT-MLT manufacturing, while are the small countries that, on average, seem to privilege this kind of integration; Greece, Portugal and many Eastern countries are pushing integration of KIBS in LT-MLT subsystems, although they are still lagging behind other EMU countries.
The evidence is substantially different when we concentrate on HT-MHT subsystems
(Figure 5), which show a positive relation with KIBS integration: the higher the HT-MHT subsystems share in the overall manufacturing, the higher the KIBS integration
(bottom-graphs). From 2000 to 2014 we observe some‘radicalisation’ in the manufacturing
sub-systems development in the EMU area. In 2014 is quite evident the hegemonic role of Germany: it is the leading manufacturing country, as a whole and especially in the HT-MHT industries, and it is the only country located in the top-right panel, given the high level of KIBS integration in HT-MHT industries. This dominant German model contrasts
with other two‘models’ which are led by the following large countries, respectively; Italy, on
the one hand, and France with Spain, on the other. Italy is located in the bottom-left panel, showing a lower than average share of HT-MHT industries coupled with a lower than average KIBS integration. Traditional manufacturing industries, largely LT-MLT, are the core of the Italian manufacturing system, requiring low levels of KIBS integration. France and Spain, on the contrary, are able to locate in the top-left panel, showing a higher than
average KIBS integration, while maintaining a lower than average share of HT-MHT. These two countries are experiencing over time a rather evident de-industrialisation process which is compensated by a strong integration of the KIBS into manufacturing subsystems. The consolidation of the German leadership in manufacturing is evident when looking at the high-tech subsystems and it seems to come at the expenses of other traditional manufacturing countries, such as Italy, while contextually in favour of eastern European countries like Slovakia. The evident Germanisation of the Eurozone manufacturing triggers the specialization of a country like Slovakia, which is a strong German supplier, but depress the manufacturing system of other countries, as Italy (Cirillo and Guarascio
2015; Simonazzi, Ginzburg, and Nocella 2013; Stöllinger et al.2013). This framework
does not seem to be consistent with the path of reorganization of manufacturing included
in the proposal of the European Commission (EU Commission2013,2014a) with the
so-called‘industrial leadership,’ a pillar of the Horizon 2020. Finally, we can also notice a
diverging path between big and small countries, in terms of manufacturing weight: the gap between large and small countries widens from 2000 to 2014.
When looking at the growth rates we can appreciate how the positive relation witnessed by a high level of correlation in the pre-crisis period is mainly driven by small eastern countries, which grew more than other countries both in terms of
manu-facturing systems and in terms of KIBS integration, being‘committed’ in a process of
catching up with respect to western countries. When the crisis hit, the positive relation drops significantly, especially because the rate of growth of KIBS integration largely slowdowns, becoming even negative for France, a large country with the highest share of KIBS integrated in manufacturing.
4.4. A disaggregated subsystem approach: industries specialization
At the industry level (19) for all the EMU countries (19) it is possible to produce an analysis based on the construction of four development patterns according to the two dimensions: rates of growth for manufacturing and KIBS integration. Each industry in each country can be located in one of the four following states: Industrialization/
KIBSIntegration, Deindustrialization/KIBSIntegration,
Industrialization/No-KIBSIntegration, Deindustrialization/No-KIBSIntegration. In addition, we can also dis-criminate the share of LT-MLT and HT-MHT industries that can be located in each of
the states of the world (Tables 3and4).
Table 3.Rates of growth of hours worked in the manufacturing subsystem and in the integrated KIBS
(2000–2008) observations: 19 (industries subsystems) x 19 (countries).
Development patterns 2000–2008 Var. HW Man Sub Var. HW KIBS N° industries Share on the total industries number Share on the total industries in the LT-MLT Share on the total industries in the HT-MHT Industrialization/KIBSIntegration + + 176 48.8 40.8 62.4 Deindustrialization/KIBSIntegration - + 74 20.5 24.1 14.3 Industrializarion/No-KIBSIntegration + - 14 3.9 5.3 1.5 Deindustrialization/No-KIBSIntegration - - 97 26.9 29.8 21.8 Source: Own elaboration. WIOD database.
In the pre-crisis period, almost all of the industries show a pattern characterized by increasing manufacturing share and increasing integration. In a relatively stable and growing phase of the economic cycle, the manufacturing shows, on average, a tendency to grow and to vertically integrate with services. This especially holds for HT-MHT industries, but also the share of LT-MLT that grows and integrates with KIBS is not negligible. Moreover, are, indeed, the LT-MLT industries, that during the crisis period (2008–2014) are more capable of maintaining a positive growth and a positive integration rate of growth. This means that the hours worked decrease less in LT-MLT than in HT-MHT industries and that the former may still constitute an important manufacturing part, which the policies for re-industrialization in EU should not neglect.
We here answer to another aspect implied by our third research question (Q3a and Q3b): HT-MHT integrates more with KIBS in the pre-crisis, but in LT-MHT integration remains more stable in the crisis. These results indicate that there are not systematic linkages between KIBS and the technological intensity of manufacturing. This could mean that LT-MLT industries, because of fewer resources available (e.g. innovation and R&D), must strongly rely on KIBS (such as advertising and marketing research or professional and scientific activities) in order to survive or differentiate their products
from competitors (Stöllinger et al.2013).
5. Conclusions
This study is focused on structural heterogeneity in EMU countries through the analysis of the integration between manufacturing and KIBS in terms of hours worked. We address two main gaps in the literature: one concerning the scarce presence of works on the changes in the economic structures of the EMU countries; the other regarding the measuring of KIBS integration in manufacturing. The paper is based on the last release of WIOD database and covers a period, from 2000 to 2014, which we divide in two sub-periods (pre-crisis, 2000–2008, and crisis, 2008–2014). By adopting a vertical perspective (subsystem approach) of the production structure we measure all the activities that need
to be integrated to create thefinal production in a specific branch. As such, the shifting of
hours worked, our main measurement unit, from manufacturing to services (and vice versa) is influenced by the demand of products and services.
The results show that disparities are growing in the composition of national produc-tive structure and they are even more pronounced when we consider intersectoral
Table 4.Rates of growth of hours worked in the manufacturing subsystem and in the integrated KIBS
(2008–2014) observations: 19 (industries subsystems) x 19 (countries). Development patterns 2008–2014 Var. HW Man Sub Var. HW KIBS N° industries Share on the total industries number Share on the total industries in the LT-MLT Share on the total industries in the HT-MHT Industrialization/KIBSIntegration + + 68 18.8 19.7 17.3 Deindustrialization/KIBSIntegration - + 83 23.0 25.9 18.0 Industrialization/No-KIBSIntegration + - 13 3.6 1.8 6.8 Deindustrialization/No-KIBSIntegration - - 197 54.6 52.6 57.9 Source: Own elaboration. WIOD database.
dynamics, that is to say a subsystem approach, confirming the Krugman position about the increasing specialisation among EMU countries. Adopting the sub-systems perspec-tive seems to be crucial in understanding how the producperspec-tive structures move and develop since it is possible to capture inter-industries linkages (one of the pillar of smart-factory) and it seems to be useful in order to inform policy makers about the magnitude intersectoral integration, traditionally neglected if compared to the role of sectors.
Indeed, when we focus on the KIBS integration for the EMU countries we notice the
emergence of diversified models of integration, whose heterogeneity is mainly
appreci-able when looking at the following pair of countries: core vs. periphery. The core countries (central EMU) show a higher level of KIBS integration in manufacturing than peripheral ones (southern and eastern EMU). The disaggregation by technological intensity shows a positive relation between KIBS integration and technological intensity over time; however, the crisis undermines the virtuous linkage between manufacturing and KIBS emerged in the pre-crisis period.
The evidence provided by the analysis also permits to draw some policy implications. First of all, the EMU endogenous forces did not push the member states toward a process of less specialization and structural and real convergence. On the contrary, real diver-gence is also implied by differences in the dynamic of the productive structures, read through the lens of KIBS integration in manufacturing. This process leads toward a low (er) degree of business cycles synchronization among EMU countries, with negative
economic consequences given the EMU institutional arrangement (De Grauwe2016),
which callfiscal union and/or risk sharing policies, besides a revision of the stringent
rules of the Stability and Growth Pact. Second, given the evidence emerged we can draw some specific lines of intervention for a new industrial renaissance in the EMU area: promote best practices to reinforce manufacturing/services relations; strengthen the pre-requisites that allow to exploit the potentiality of manufacturing/KIBS integration for those countries lagging behind the leaders in terms of integration (e.g. policies to foster ICT adoption and implementation); foster manufacturing growth in those industries that better integrate KIBS, which are mainly HT-MHT industries, but without neglecting the role of traditional industries (largely LT-MLT).
Although the analysis here presented provides a comprehensive view of the dynamic of KIBS integration in manufacturing for EMU countries, further lines of analyses can depart from the current results: extending, through econometric modelling, the analysis of the effects on productivity induced by the productive structures dynamics here illustrated; focusing on the bilateral relations among EU countries according to the view of global-value-chains in order to detect losers and winners in the after crisis stabilization of productive structures; analyzing the issues at stake at a regional level.
Notes
1. The definition of KIBS generally includes (ISIC Rev. 4): Computer programming,
consul-tancy and related activities, information service activities; legal and accounting activities, activities of head offices, management consultancy activities; architectural and engineering activities, technical testing and analysis; scientific research and development; advertising and market research; other professional, scientific, administrative and technical activities.
2. The evidence, as shown inTable 1, holds both before and during the economic crisis.
3. The same index can be calculated, more traditionally, by sectors, as also reported inTable 2, in which KSI is the traditional Krugman Specialization Index calculated by sectors.
4. We have a confirmation of this trend, also applying the coefficient of variation – the so-called sigma convergence– (calculated as the ratio between standard deviation and arith-metic mean in the share of manufacturing among the EMU19 countries).
5. In our comments, we refer to the two sub-periods as pre-crisis (2000–2008) and crisis (2008–2014), since after the Great Recession, several EMU countries (mainly southern ones), and among them two of the largest (Spain and Italy) experienced a period of severe economic turmoil due to the Sovereign Debt Crisis. In addition, the effects of the Great Recession lasted some years after its burst in EMU countries.
6. The changes in trade relationships among the EMU countries may also be considered factors at the basis of both manufacturing and KIBS integration dynamics; however, we are not analyzing it in the present paper, leaving the deepening of this issue to future works.
7. The largest countries seem to reduce their disparities both in terms of LT-MLT shares and of KIBS integration; some small countries like Portugal and Greece increase their divide with respect to other countries in terms of LT-MLT shares.
Acknowledgements
We would like to offer thanks to the participants of the following conferences for their valuable comments: AISRe 2018, Bolzano Italy; Euromemo 2018, Helsinki Finland; Astril 2018, Rome Italy; International Conference Developments in Economic Theory and Policy 2019, Bilbao -Spain.
Disclosure statement
No potential conflict of interest was reported by the authors.
ORCID
Davide Antonioli http://orcid.org/0000-0001-5150-3714 Claudio Di Berardino http://orcid.org/0000-0002-3191-3297 Gianni Onesti http://orcid.org/0000-0002-2621-3536
References
Archibugi, D., A. Filippetti, and M. Frenz.2013.“Economic Crisis and Innovation: Is Destruction Prevailing over Accumulation?” Research Policy 42 (2): 303–314. doi:10.1016/j. respol.2012.07.002.
Bagnai, A., and C. A. Mongeau Ospina. 2017. “Monetary Integration Vs. Real Disintegration: Single Currency and Productivity Divergence in the Euro Area.” Journal of Economic Policy Reform 21 (4): 353–367. doi:10.1080/17487870.2017.1403755.
Baker, P.2007.“The Impact of Business-services Use on Client Industries: Evidence from Input-output Data.” In Business Services in European Economic Growth, edited by L. Rubalcaba and H. Kox, 97–115. London: Palgrave Macmillan. doi:10.1057/9780230228795_6.
Barriero de Souza, K., S. Q. de Andrade Bastos, and F. S. Perobelli. 2016.“Multiple Trends of Tertiarization: A Comparative Input–Output Analysis of the Service Sector Expansion between Brazil and United States.” EconomiA 17 (2): 141–158.
Bolea, L., R. Duarte, and J. Sánchez Chóliz.2018.“From Convergence to Divergence? Some New Insights into the Evolution of the European Union.” Structural Change and Economic Dynamics 47: 82–95. online first. doi:10.1016/j.strueco.2018.07.006.
Castaldi, C.2009.“The Relative Weight of Manufacturing and Services in Europe: An Innovation Perspective.” Technological Forecasting and Social Change 76 (6): 709–722. doi:10.1016/j. techfore.2008.08.001.
Castellacci, F.2010.“Structural Change and the Growth of Industrial Sectors: Empirical Test of a GPT Model.” Review of Income and Wealth 56 (3): 449–482. doi:10.1111/j.1475-4991.2010.00380.x. Celi, G., A. Gingzburg, D. Guarascio, and A. Simonazzi.2018. Crisis in the European Monetary
Union. A Core-periphery Perspective. London: Routledge.
Cheng, D., and P. W. Daniels.2014.“What’s so Special about China’s Producer Services? an Input-Output Analysis.” China & World Economy 22 (1): 103–120. doi:10.1111/j.1749-124X.2014.12055.x. Ciriaci, D., and D. Palma.2016.“Structural Change and Blurred Sectoral Boundaries: Assessing the Extent to Which Knowledge-intensive Business Services Satisfy Manufacturing Final Demand in Western Countries.” Economic Systems Research 28 (1): 55–77. doi:10.1080/09535314.2015.1101370. Cirillo, V., and D. Guarascio. 2015. “Jobs and Competitiveness in a Polarised Europe.”
Intereconomics 50 (3): 156–160. doi:10.1007/s10272-015-0536-0.
Dachs, B., S. Biege, M. Borowiecki, G. Lay, A. Jäger, and D. Schartinger.2014.“Servitisation of European Manufacturing: Evidence from a Large Scale Database.” The Service Industries Journal 34 (1): 5–23. doi:10.1080/02642069.2013.776543.
De Grauwe, P.2016. Economics of Monetary Union. Oxford: Oxford University Press.
Desmarchelier, B., F. Djellal, and F. Gallouj.2013.“Knowledge Intensive Business Services and Long Term Growth.” Structural Change and Economic Dynamics 25 (C): 188–205. doi:10.1016/j. strueco.2012.07.003.
Di Berardino, C., and G. Onesti.2018.“The Two-way Integration between Manufacturing and Services.” The Service Industries Journal 1–21. Online first. doi:10.1080/02642069.2018.1438415. Dietrich, M. 1999. “Explaining Economic Restructuring: An Input-output Analysis of Organizational Change in the European Union.” International Review of Applied Economics 13 (2): 219–240. doi:10.1080/026921799101670.
Dietzenbacher, E., B. Los, R. Stehrer, M. Timmer, and G. de Vries.2013.“The Construction of World Input-Output Tables in the WIOD Project.” Economic Systems Research 25 (1): 71–98. doi:10.1080/09535314.2012.761180.
EU Commission.2013.“Towards Knowledge Driven Re-industrialisation.” European competitive-ness report, Brussels.
EU Commission.2014a.“Towards an Industrial Renaissance.” Industrial Policy Communication Update, Brussels.
EU Commission.2014b.“High Level Group on Business Services.” Final Report, Brussels. Eurofound.2014. Impact of the Crisis on Industrial Relations and Working Conditions in Europe.
Dublin
Francois, J., M. Manchin, and P. Tomberger. 2015. “Services Linkages and the Value Added Content of Trade.” The World Economy 38 (11): 1631–1649. doi:10.1111/twec.12307.
Franke, R., and P. Kalmbach.2005.“Structural Change in the Manufacturing Sector and Its Impact on Business-Related Services: An Input–Output Study for Germany.” Structural Change and Economic Dynamics 16 (4): 467–488. doi:10.1016/j.strueco.2004.09.001.
Garbellini, N., E. Marelli, and A. L. Wirkierman.2014.“Domestic Demand and Global Production in the Eurozone: A Multi-regional Input-output Assessment of the Global Crisis.” International Review of Applied Economics 28: 335–363. doi:10.1080/02692171.2013.872086.
Harcourt, G. C., and V. G. Massaro.1964.“Mr. Sfraffa’s Production of Commodities.” Economic Record 40 (91): 442–454. doi:10.1111/j.1475-4932.1964.tb02173.x.
Kox, H. L. M.2004.“The Contribution of Business Services to Aggregate Productivity Growth.” In Fostering Productivity: Patterns, Determinants and Policy Implications, edited by G. Gelauff and L. Klomp, 243–264. Bingley: Emerald Group Publishing Limited.
Krugman, P.1993. “Lessons from Massachussets for EMU.” In Adjustment and Growth in the European Monetary Union, edited by F. E. Torres and F. Giavazzi, 241–266. Oxford, Mew York, Melbourne: Cambridge University Press.
Los, B., M. P. Timmer, and G. J. de Vries.2015.“How Global are Global Value Chains? A New Approach to Measure International Fragmentation.” Journal of Regional Science 55 (1): 66–92. doi:10.1111/jors.12121.
Mas-Verdú, F., A. Wensley, M. Alba, and J. M. G. Álvarez-Coque.2011.“How Much Does KIBS Contribute to the Generation and Diffusion of Innovation?” Service Business 5 (3): 195–212. doi:10.1007/s11628-011-0110-1.
Mazzucato, M., M. Cimoli, G. Dosi, J. E. Stiglitz, M. A. Landesmann, M. Pianta, R. Walz, and T. Page. 2015. “Which Industrial Policy Does Europe Need?” Intereconomics 50 (3): 120–155. doi:10.1007/s10272-015-0535-1.
McCarthy, I., and A. Anagnostou.2004.“The Impact of Outsourcing on Transaction Costs and Boundaries of Manufacturing.” International Journal of Production Economics 88 (1): 61–71. doi:10.1016/S0925-5273(03)00183-X.
Miozzo, M., and I. Miles.2002.“The Relation between the Internationalization of Services and the Process of Innovation: A Research Agenda.” In Internationalization, Technology and Services, edited by M. Miozzo and I. Miles, 15–32. UK: Edward Elgar Publishing Limited. doi:10.4337/ 9781781952924.00009.
Momigliano, F., and D. Siniscalco. 1982. “Note in Tema Di Terziarizzazione E Deindustrializzazione.” Moneta E Credito 138: 143–182.
Mongelli, F. P., E. Reinhold, and G. Papadopoulos.2016.“What’s so Special about Specialization in the Euro Area? Early Evidence of Changing Economic Structures.” ECB Occasional Paper Series n. 168. Montresor, S., and G. Vittucci Marzetti.2011.“The Deindustrialisation/tertiarisation Hypothesis Reconsidered: A Subsystem Application to the OECD7.” Cambridge Journal of Economics 35 (2): 401–421. doi:10.1093/cje/beq009.
MPC. 2004. “Sectoral Specialisation in EU. A Macroeconomic Perspective.” Occasional Papers. Frankfurt: European Central Bank.
Organisation for Economic Cooperation and Development (OECD). 2003. Science, Technology and Industry Scoreboard 2003. Paris, France: Organisation for Economic Co-operation and Development. doi:10.1787/sti_scoreboard-2003-en.
Palan, N., and C. Schmiedeberg. 2010. “Structural Convergence of European Countries.” Structural Change and Economic Dynamics 21 (2): 85–100. doi:10.1016/j.strueco.2010.01.001. Pasinetti, L. L.1973.“The Notion of Vertical Integration in Economic Analysis.” Metroeconomica
25 (1): 1–29. doi:10.1111/j.1467-999X.1973.tb00539.x.
Pianta, M.2014.“An Industrial Policy for Europe.” Seoul Journal of Economics 27 (3): 277–305. https://paper s.ssrn.com/sol3/paper s.cfm?abstr act_id=25303 44.
Pilat, D., A. Cimper, K. B. Olsen, and C. Webbet.2006.“The Changing Nature of Manufacturing in OECD Economies.” OECD Science, Technology and Industry Working Papers, 2006/09. Paris: OECD Publishing. doi:10.1787/308452426871.
Pomini, M.2012. “Structural Dynamics in Historical Perspective.” In Structural Dynamics and Economic Growth, edited by R. Arena and P. L. Porta, 49–68. Cambridge: Cambridge University Press.
Portella-Carbó, F.2016.“Effects of International Trade on Domestic Employment: An Application of a Global Multiregional Input–Output Supermultiplier Model (1995–2011).” Economic Systems Research 28 (1): 95–117. doi:10.1080/09535314.2016.1142429.
Rodrik, D. 2016. “Premature Deindustrialization.” Journal of Economic Growth 21 (1): 1–33. doi:10.1007/s10887-015-9122-3.
Sarra, A., C. Di Berardino, and D. Quaglione.2019.“Deindustrialization and the Technological Intensity of Manufacturing Subsystems in the European Union.” Economia Politica 36 (1): 205– 243. doi:10.1007/s40888-018-0112-8.
Scazzieri, R.1990.“Vertical Integration in Economic Theory.” Journal of Post Keynesian Economics 13 (1): 20–46. doi:10.1080/01603477.1990.11489829.
Silva, E. G., and A. A. C. Teixeira.2008.“Surveying Structural Change: Seminal Contributions and a Bibliometric Account.” Structural Change and Economic Dynamics 19 (4): 273–300. doi:10.1016/j.strueco.2008.02.001.
Simonazzi, A., A. Ginzburg, and G. Nocella.2013.“Economic Relations between Germany and Southern Europe.” Cambridge Journal of Economics 37 (3): 653–675. doi:10.1093/cje/bet010. Sraffa, P. 1960. Production of Commodities by Means of Commodities: Prelude to a Critique of
Economic Theory. Cambridge: Cambridge University Press.
Stöllinger, R., N. Foster-McGregor, M. Holzner, M. Landesmann, J. Pöschl, R. Stehrer, and C. Stocker- Waldhuber.2013. A“Manufacturing Imperative” in the EU-Europe’s Position in Global Manufacturing and the Role of Industrial Policy. Vienna: WIIW.
Timmer, M. P., E. Dietzenbacher, B. Los, R. Stehrer, and G. J. de Vries.2015.“An Illustrated User Guide to the World Input-output Database: The Case of Global Automotive Production.” Review of International Economics 23 (3): 575–605. doi:10.1111/roie.12178.
Tregenna, F.2009.“Characterising Deindustrialisation: An Analysis of Changes in Manufacturing Employment and Output Internationally.” Cambridge Journal of Economics 33 (3): 433–466. doi:10.1093/cje/ben032.
Appendix
Table A1.Countries code.
AUT Austria BEL Belgium CYP Cyprus DEU Germany ESP Spain EST Estonia FIN Finland FRA France GRC Greece IRL Ireland ITA Italy LTU Lithuania LUX Luxembourg LVA Latvia MLT Malta NLD Netherlands PRT Portugal SVK Slovakia SVN Slovenia
Table A2. Rates of growth of hours worked in the manufacturing subsystem and in the integrated KIBS (2000 –2014; 2000 –2008; 2008 –2014). 2000/2014 2000/2008 2008/2014 TOTAL LOW-TECH HIGH-TECH TOTAL LOW-TECH HIGH-TECH TOTAL LOW-TECH HIGH-TECH country Man INT-KIBS Man INT-KIBS Man INT-KIBS Man INT-KIBS Man INT-KIBS Man INT-KIBS Man INT-KIBS Man INT-KIBS Man INT-KIBS AUT -0.003 0.573 -0.065 0.476 0.096 0.698 0.054 0.598 0.023 0.597 0.103 0.599 -0.054 -0.015 -0.086 -0.076 -0.007 0.062 BEL -0.292 0.034 -0.277 -0.037 -0.313 0.127 -0.045 0.200 -0.001 0.297 -0.105 0.074 -0.259 -0.138 -0.276 -0.257 -0.233 0.049 CYP -0.371 -0.235 -0.423 -0.286 0.072 0.203 -0.103 -0.105 -0.152 -0.153 0.314 0.304 -0.299 -0.145 -0.320 -0.157 -0.184 -0.077 EST -0.073 0.188 -0.179 -0.080 0.851 2.423 0.056 -0.077 -0.068 -0.199 1.149 0.941 -0.123 0.287 -0.119 0.149 -0.139 0.764 FIN -0.231 -0.193 -0.185 0.064 -0.286 -0.318 -0.054 0.065 -0.108 0.138 0.010 0.029 -0.188 -0.242 -0.087 -0.066 -0.293 -0.338 FRA -0.299 -0.383 -0.254 -0.206 -0.352 -0.506 -0.088 0.039 -0.068 0.134 -0.113 -0.027 -0.231 -0.406 -0.200 -0.300 -0.269 -0.492 DEU 0.025 0.407 -0.005 0.435 0.050 0.388 0.037 0.440 0.018 0.430 0.052 0.447 -0.011 -0.023 -0.023 0.003 -0.002 -0.041 GRC -0.019 0.826 -0.007 0.921 -0.089 0.490 0.127 0.541 0.070 0.506 0.450 0.666 -0.130 0.185 -0.072 0.276 -0.372 -0.106 IRL -0.450 -0.708 -0.426 -0.417 -0.476 -0.860 -0.157 -0.038 -0.149 0.077 -0.165 -0.097 -0.347 -0.696 -0.325 -0.458 -0.372 -0.845 ITA -0.170 0.235 -0.215 0.187 -0.096 0.296 0.054 0.501 0.023 0.437 0.106 0.582 -0.212 -0.177 -0.232 -0.174 -0.182 -0.181 LVA -0.199 0.960 -0.254 0.859 0.342 1.693 -0.062 0.616 -0.126 0.533 0.566 1.220 -0.146 0.213 -0.146 0.213 -0.143 0.213 LTU -0.163 0.896 -0.201 0.825 0.394 3.134 -0.044 0.558 -0.097 0.498 0.741 2.440 -0.125 0.217 -0.115 0.218 -0.199 0.202 LUX -0.185 -0.641 -0.214 -0.646 -0.059 -0.626 -0.033 -0.244 -0.044 -0.224 0.016 -0.315 -0.157 -0.525 -0.178 -0.543 -0.073 -0.454 MLT -0.504 -0.662 -0.393 -0.497 -0.650 -0.784 -0.308 -0.702 -0.268 -0.623 -0.360 -0.760 -0.284 0.133 -0.170 0.335 -0.454 -0.102 NLD -0.339 -0.447 -0.360 -0.428 -0.303 -0.468 -0.026 0.120 -0.079 0.011 0.061 0.241 -0.321 -0.507 -0.306 -0.434 -0.343 -0.572 PRT -0.181 0.411 -0.177 0.476 -0.197 0.292 -0.133 0.270 -0.130 0.289 -0.143 0.233 -0.056 0.112 -0.054 0.145 -0.063 0.048 SVK -0.060 0.378 -0.281 -0.048 0.440 1.099 0.075 0.511 -0.095 0.018 0.458 1.345 -0.125 -0.088 -0.206 -0.065 -0.012 -0.105 SVN -0.156 0.341 -0.271 0.137 0.062 0.685 0.002 0.299 -0.109 0.135 0.214 0.575 -0.158 0.032 -0.181 0.002 -0.125 0.069 ESP -0.164 1.182 -0.136 1.288 -0.208 1.056 0.058 0.893 0.106 1.056 -0.020 0.700 -0.209 0.152 -0.219 0.113 -0.192 0.209 EMU19 -0.138 0.104 -0.163 0.183 -0.103 0.036 0.007 0.309 -0.009 0.337 0.028 0.286 -0.144 -0.157 -0.156 -0.115 -0.128 -0.194
Table A3.Correlation matrix of the rate of growth of manufacturing systems and KIBS integration by different periods (2000–2014; 2000–2008; 2008–2014) and technological classification.
Tot 00-14 Tot 00-08 Tot 08-14 LT-MLT 00-14 LT-MLT 00-08 LT-MLT 08-14 HT-MHT 00-14 HT-MHT 00-08 HT-MHT 08-14 Correlation: grw Man/grw KIBS 0.688 0.713 0.549 0.661 0.739 0.601 0.842 0.764 0.406
