Content uploaded by Slavo Radosevic
Author content
All content in this area was uploaded by Slavo Radosevic on Nov 23, 2015
Content may be subject to copyright.
CENTRE FOR COMPARATIVE ECONOMICS
E&BWP
UCL SSEES
Centre for Comparative Economics
Dynamics of Technology Upgrading of the Central and East
European Countries in a Comparative Perspective: Analysis
Based on Patent Data
Björn Jindraa, Iciar Dominguez Lacasab and Slavo Radosevicc
a Copenhagen Business School, u University of Bremen, c SSEES, UCL
Economics and Business Working Paper No.135
February 2015
Centre for Comparative Economics
UCLSchool of Slavonic and East European Studies
Gower Street, London, WC1E 6BT
Tel: +44 (0)20 7679 8519
Fax: +44 (0)20 7679 8777
2
Abstract
This working paper explores patterns of technology upgrading as a three-dimensional process
which consists of (i) intensity of technology upgrading, (ii) structural change, and (iii) interaction
with the global economy. The specificity of our report is that we depict patterns of technology
upgrading by relying entirely on patent data. We derive patent indicators to capture the three
dimensions.
Patent indicators for intensity of technology upgrading trace technological capabilities at the
technology frontier (transnational patents) and behind the technology frontier
(domestic/resident direct applications to national offices). Structural change in technological
knowledge is depicted by the share of transnational patent applications in high technology
fields and knowledge-intensive activities and by calculating a technological diversification index.
To capture interaction with global economy in the upgrading process indicators measure
technological knowledge sourcing across countries and interactions between foreign and
indigenous actors. Based on 7 patent indicators covering the three upgrading dimensions the
comparative analysis focuses on EU27 and its subregions and on the BRICS countries.
According to the results, in 2011 CEECs were quite homogenous in their upgrading paths. A
typical CEE economy in 2011 is well behind EU12 in terms of frontier technology intensity,
domestic technology intensity, share of high tech patents and technology sourcing abroad.
Moreover, its organizational capabilities are often less advanced. The CEE profile is much less
coherent in terms of technology diversification/specialization and share of joint inventions.
However, differences among CEECs are not significant. Still there are some notable national
features. Poland, Romania and Slovenia have above average domestic technological intensity
which reflects partly their sizes (Romania and Poland) and specific model of innovation system
reliant on domestic R&D intensive firms (Slovenia). Latvia and Lithuania are specific in terms of
high share of HTKI patents.
CEE technology upgrading as depicted by patents is within the BRIC pattern (with exception of
China which in terms of technology upgrading has de facto delinked from BRICS). In the BRIC
context, the CEE characterize very open innovation system with a high share of coinventions and
foreign actors exploiting local inventions. This reveals weak organizational capabilities to
commercialize its own inventions.
According to the results CEE grew during 1990s/2008 based on production, not technological
capability. Their future growth will increasingly depend on building technological capabilities at
world frontier level. Our analysis shows that the basis for such growth exists only to a limited
extent and that speed of upgrading towards world frontier activities is well beyond required for
catching up. Equally, our analysis shows that solutions for improved technology upgrading will
need to be found with their existing innovation model of small open economies integrated into
the EU.
3
1. INTRODUCTION
In this working paper we explore issues related to technology upgrading of the European Union
(EU) peripheral economies, especially of the new EU member states1. Technology upgrading is
the key to further long-term growth as suggested by the growth literature. This has been
already recognised by the EU policy agenda which has promoted Smart Specialization Strategies
(SSS) as the ex-ante conditionality for use of the EU Structural Funds to so call less favoured EU
regions and countries. In addition, EU has been using the European Innovation Scoreboard as it
was called in the past and now Innovation Union Scoreboard (IUS) as the major metrics in
assessing progress of all EU countries in terms of their innovation capacity. This metrics has
become so dominant that some of its either individual or aggregate indicators have been used
as policy objectives and benchmarks in measuring how countries perform in achieving the aims
of SSS and other national policy targets.
Central and Eastern European countries (CEECs) are largely middle-income economies but it is
not certain whether they have achieved a threshold of technological capability required for
catching up to high-income economies status2. The shift from middle income to high-income is
not guaranteed or is not automatic as growth process is usually non-linear and evolves across
several threshold levels with their specific threshold requirements. In order to understand this
process we need to be open to a variety of historical experiences as well as go beyond simple
explanations of growth be they adequate institutions (REF3), human capital (Glaeser et, 2004 4)
or Research and Development (R&D) (OECD, 20045).
In order to advance research in this area we approach to the issue of growth and measurement
of growth through the perspective of technology upgrading. This is a multidimensional
conceptual framework which is open to sensitivities of different levels of development and
which is also empirically informed but also has some theoretical relevance. We consider it as
appreciative theorizing framework which aim to overcome a frequent weakness of composite
indicators which is that they represent “measurement without theory” (Koopmans, 19476). A
conceptual approach is based on the literature review and is developed as part of this task in a
paper by Radosevic and Yoruk (2014) Why do we need theory and metrics of technology
upgrading? as part of this deliverable. Here we rely broadly on this approach but we also
develop it further by applying it based on patent data.
The paper is organised as follows: We first explain approach to technology upgrading by
discussing its elements (section 2). In section 3 we use this approach to analyse individual
1 By European periphery we mean neighboring countries, which are not members of the EU. These are West
Balkan countries, Turkey, and European CIS countries
2 Based on World Bank criteria only Bulgaria and Romania are middle income economies while others are in a high
income group. However, from our perspective this classification is not suitable for categorising CEECs and for
understanding middle income trap.
3 Acemoglu Daron and James Robinson (2008) The Role of Institutions in Growth and Development, Commission
on Growth and Development, Working Paper No. 10, The World Bank
4 Glaeser, Edward L; R. LaPorta F. Lopes-de-Silanes and A. Shleifer (2004). "Do Institutions Cause Growth?". Journal
of Economic Growth.
5 OECD (2004) Understanding Growth, Palgrave Macmillan, London
6 T.C. Koopmans, Measurement without theory, Rev. Econ. Stat. 29 (3) (1947) 161–172.
4
indicators of technology upgrading. Section 4 explores position of EU in technology upgrading
in a comparative perspective of the EU28 and BRICS economies. In section 5 we explore three
dimensions of technology upgrading. Section 6 concludes.
2. DETECTING TECHNOLOGY UPGRADING THROUGH PATENT DATA: A
CONCEPTUAL APPROACH
Our departing proposition is that technology upgrading is multidimensional process. By this we
mean that: it is based on broader understanding of innovation, which goes well beyond R&D. It
is multi-level process which means that it is micro, mezzo and macro grounded but which also
means that at its core is structural change in various dimensions: technological, industrial,
organisational. It is also an outcome of global forces (embodied in international trade and
investment flows) and local strategies (pursued by host country firms and governments)(for
extensive review of literature on this issue see Radosevic and Yoruk, 20147; for perspective
along these lines see Ernst, 20088;Lall, 19929).
In nutshell, based on literature review and at general level we approach to technology
upgrading as three-dimensional process. It consists of dimension 1: which is about intensity of
technology upgrading as depicted by different types and levels of innovation activities, of
dimension 2: which is about spread or width of technology like diversity of technological
knowledge, and of dimension 3: which depicts knowledge inflows into economy through a
variety of forms like trade, FDI and global value chains. All three dimensions have strong
grounding in the respective literatures on firm level technology upgrading, on structural change
and growth, and on integration in global economy. Figure 1 summarizes three dimensions
and paths of technology upgrading.
77 Slavo Radosevic and Esin Yoruk (2014) Are there global shifts in world science base? Analysis of catching up and
falling behind of world regions, Scientometrics, June, DOI 10.1007/s11192-014-1344-1
8 Ernst, Dieter (2008) Asia’s “upgrading through innovation” strategies and global innovation networks: an
extension of Sanjaya Lall’s research agenda, Transnational Corporations, Volume 17, Number 3, December 2008
9 Lall, S. (1992), Technological capabilities and industrialization, Research Policy, Vol. 20, No. 2, 165-186
5
Figure 1: Dimensions and paths of technology upgrading
Source: authors
In the technology upgrading process dimension 1 evolves from domestic behind technology
frontier efforts towards world frontier technology efforts. Dimension 2 goes in direction of
increasing diversification in terms of categories of technological knowledge and increasing
share of knowledge in high growth or dynamic areas. Dimension 3 evolves from invention
process being driven by foreign actors towards joint knowledge generation and then towards
sourcing of technology from abroad. We aim at capturing these dimensions and their evolution
using patent indicators.
The body of research on measuring countries’ performance in growth, competitiveness and
innovation offers a variety of composite indicators. Examples are: the Global Competitiveness
Index (WEF, 201210), the Knowledge Economy Index (Chen and Dahlman, 200411) of the World
Bank, the World Competitiveness Report Index of IMD (http://www.imd.org/), Technological
capability of countries and the ArCo, (Archibugi and Coco, 200512, 200413; Archibugi et al.,
200914), the UNIDO Industrial Performance Scoreboard (UNIPS), the Summary Innovation Index
10 World Economic Forum (2012) Global Competitiveness Report 2011-12, WEF, Geneva
11 Chen, Derek H. C.; Dahlman, Carl J.. 2004. Knowledge and Development: A Cross-Section Approach. World Bank,
Washington, D.C
12 Daniele Archibugi and Alberto Coco (2005) Measuring technological capabilities at the country level: A survey
and a menu for choice, Research Policy 34 (2005) 175–194
13 Daniele Archibugi and Alberto Coco (2004)A New Indicator of Technological Capabilities for Developed and
Developing Countries (ArCo), World Development Vol. 32, No. 4, pp. 629–654, 2004
14 Daniele Archibugi, Denni, M. and Filippetti, A. (2009). The technological capabilities of nations: The state of the
art of synthetic indicators. Technological Forecasting and Social Change Vol. 76: 917-931
DIMENSION1
Intensity of technology upgrading:
from behind technology frontier
to technology frontier
DIMENSION 3
Interaction with global economy
(from inventions driven by foreign actors, to co-inventions
and to technology sourcing from abroad)
DIMENSION 2
Breadth of technology upgrading
(structural changes towards diversification of
technological knowledge and increased share of high
technology and knowledge intensive activities)
6
and the Global Innovation Index, both of the European Commission; the Technological Activity
Index of the UNIDO; the Technological Advance Index of the UNCTAD; the Technology
Achievement Index, developed by UNDP and reported in the Human Development Report
2001, and the S&T Capacity Index (STCI) proposed by the RAND Corporation, the High-Tech
Indicators (HTI) developed at the Georgia Tech Technology Policy and Assessment Center and
reported by the National Science Foundation's Science & Engineering Indicators.
Nasierowski and Arcelus (2000)15show that similarity in ranking across different indexes are
significant. They all point to importance of innovation to economic development but
differences in their conceptual perspectives do not change significantly ranking among
countries. Archibugi et al (2009)16 show similar results but also show that differences in ranking
cannot be substituted by single indicator like R&D.
It is important to bear in mind that different indexes treat ‘technology’ in different ways. Some
of them cannot be taken as a direct measure of innovative performance. Indicators like Global
Competiveness Index depict the quality of the current endowment of a country (including
institutions) and among them also the technology activities as one of determinants of growth.
Our aim is to confine ourselves to technology upgrading and we do not aim to unravel a
complex picture of the entirety of factors that determine growth and competiveness of
economies. Also, unlike the majority of rankings, our aim is not really to focus on ranking but
on different paths of technology upgrading. The learning effect should be in showing diversity
of paths and compare countries in terms of their own upgrading paths.
The specificity of our paper is that we depict patterns of technology upgrading by relying
entirely on patent data. On the one hand, the exclusive reliance on patents has costs in terms
of capturing only a part of technology effort. Their intangible character is more advantageous
as countries move up towards technology frontier and less relevant for countries behind
technology frontier where IPRs are not the major form of protection of technological knowhow.
This is especially important as innovation activities in latecomer economies like CEE are largely
about adoption and improvements on imported machinery. Although, technology as stock of
knowledge should be kept separate from production, technological capacities and production
capacity are in reality strictly interconnected (Bell and Pavitt, 199717). However, use of only
patents means that similar to Archibugi and Coco (200518) we need to abstract production
from technology capability. On the other hand, an important advantage of using patents is the
length and consistency of time series derived as well as the possibility to identify technological
fields or specializations using the patent classification. Unlike macroeconomic variables
15 W. Nasierowski and F.J. Arcelus (2013) On Perceptions of Technical Efficiency of the Basis of the Innovation
Union Scoreboard. Available online at:
https://www.guidedbees.com/~ruben/EDSI/papers/nasierowski_onperceptions.pdf (last accessed January 2015)
16 Archibugi Daniele, Mario Denni, Andrea Filippetti (2009) The technological capabilities of nations: The state of
the art of synthetic indicators Technological Forecasting & Social Change 76: 917–931
17 Bell, M. and Pavitt, K. (1997) Technological accumulation and industrial growth: contrasts between developed
and developing countries, in Archibugi, D. and Michie, J. (eds) 1997a
18 Daniele Archibugi and Alberto Coco (2005) Measuring technological capabilities at the country level: A survey
and a menu for choice, Research Policy 34 (2005) 175–194
7
technological capabilities are changing very slowly even during periods of deep economic crises
or high growth periods (Archibugi, 200919). By using patents we can detect easily stock and
flows and thus depict much better compared to other indicators changes in technology
intensity as well as structural change in technological knowledge. These two dimensions –
technology upgrading and structural change – should be considered jointly with the way
economy integrates itself in global knowledge flows.
In overall, we think that benefits surpass costs in this case provided that we are aware of the
changing nature of patenting as countries move from the position of technology followers to
leaders and as they shift from domestic and behind frontier technology effort to world frontier
technology effort. Figure 2 shows patent indicators used which depict individual dimensions of
technology upgrading.
Figure 2: Dimensions and components of technology upgrading as depicted by patent
indicators
Source: authors
2.1. INTENSITY OF TECHNOLOGY UPGRADING (SCALE)
This dimension of upgrading is about acquiring different types of technology capabilities, which
are also a reflection of different technological levels of economies. Economies that operate
behind technology frontier are more likely to grow based on production capability while high-
19 Archibugi, Daniele, Mario Denni, Andrea Filippetti (2009) The technological capabilities of nations: The state of
the art of synthetic indicators Technological Forecasting & Social Change 76 (2009) 917–931
•Transnational patenting (TN)
•Resident direct patenting (WIPO)
Intensity of
technology
upgrading
• Foreign applications of national inventions (FANI)
•International Co-inventions (COINV)
•National applications of foreign inventions NAFI)
Interaction with
global economy
•Patent applications in high tech and knowledge intensive
services (HTKI)
• Technological knowledge diversification of domestic and
transnational patent categories (Herfindhal index)
Breadth of
technology
upgrading
8
income economies are more likely to grow based on technology frontier (R&D based) activities.
Three types of capabilities (production capability, technology capability, R&D) are present in all
economies to different degrees. Their importance as drivers of growth varies in dependence of
achieved income and technology levels as well as of the structural features of economies.
We use patent indicators to measure domestic technological capability. Nonetheless, for the
analysis it is necessary to differ between domestic technological capability pushing the
technology frontier and domestic technological capability for technological development
behind technology frontier. To capture domestic technological activities pushing the technology
frontier we rely on transnational patent applications of domestic applicants (TN). Transnational
patent applications include applications to the European Patent Office and PCT applications.
These patent filings reflect technological activities relevant for competitiveness in international
markets. This international relevance of patent protection suggests that the technology
protected pushes the technology frontier at a global level. To capture technological capability
for technological development behind the technology frontier we use direct patent applications
by residents to their respective national patent offices. In general terms (even though the
patent strategies may differ from this rule) residents will directly apply for patents in their
home countries disregarding applications abroad if their technological activities do not have a
global industrial relevance.20 To us resident direct patent applications to national patent offices
dominantly proxy technology effort behind the technology frontier. Countries that are behind
technology frontier should have much higher share of resident patents and their share of
transnational patents is marginal. However, as they move towards technology frontier their
transnational patenting increases. This pattern may be somewhat different in very large
catching up economies where domestic patenting may continue to play important role.
However, their transnational patenting as proxy of world frontier technology effort should
continue to increase.
Figure 3 shows on the left the relationship between transnational patents applications per
capita (TN) and GDP per capita for the EU12 (developed or core EU), South EU (Greece, Portugal
and Spain) and the EU CEECs over 1990-2012 period. On the right Figure 1 shows same
relationship but for WIPO patents per capita i.e. for domestic technology effort. The
relationship is much better for transnational patents which indicate close relationship with
levels of GDPpc.
20 We are aware that this strategy is much more relevant for smaller than for larger and more developed
economies where due to their economic size we may expect that more patents will be registered as priority
patents i.e both at home and abroad than in small economies. However, this factor in analysis is controlled by
patents by GDP proxy.
9
Figure 3: Technology intensity at the frontier and behind the frontier vs GDP pc.
Source: RegPat, World Bank and authors’ calculations
GDP pc vs TN Patents pc GDP pc vs. Resident Patents pc
EU12 (1990-2011) EU12 (1990-2011)
GDP pc vs TN Patents pc GDP pc vs. Resident Patents pc
SouthEU (1990-2011) SouthEU (1990-2011)
GDP pc vs TN Patents pc GDP pc vs. Resident Patents pc
CEE (1990-2011) CEE (1990-2011)
20000 40000 60000 80000 100000
GDP per capita
0 200 400 600 800 1000
Transnational patents per 1 million inhabitants
20000 40000 60000 80000 100000
GDP per capita
0 200 400 600
Resident patents per 1 million inhabitants
10000 15000 20000 25000 30000 35000
GDP per capita
020 40 60 80 100
Transnational patents per 1 million inhabitants
10000 15000 20000 25000 30000 35000
GDP per capita
050 100 150
Resident patents per 1 million inhabitants
0
10000 20000 30000
GDP per capita
020 40 60 80
Transnational patents per 1 million inhabitants
0
5000 10000 15000 20000
GDP per capita
050 100 150 200 250
Resident patents per 1 million inhabitants
10
2.2. TECHNOLOGY UPGRADING AND STRUCTURAL CHANGE
There is not general theory of structural change but a variety of theoretical approaches of
different methodological nature that aim to explain structural shifts between three broad
sectors and among industries within these sectors (Krueger, 200821). There is a common
understanding that technological changes affect structural change in the way that industries
with relatively lower rates of productivity growth tend to shrink in terms of shares while those
with higher rates of productivity growth expand. In this way structural change promotes
aggregate productivity growth even if we assume that within industries productivity growth
remains stagnant. However, the empirical evidence on the role of structural change in
aggregate productivity growth escapes broad generalisations. It generates positive as well as
negative contributions to aggregate productivity growth. Since many of these effects net out,
structural change on average appears to have only a weak impact (Peneder, 200322).
So, instead of being focused on structural changes at the level of industries it seems more
appropriate to track changes in the structure of technological knowledge.
We depict structural change in technological knowledge by using two indicators. First,
transnational patent applications in high technology fields and knowledge-intensive services.
Second, we use technological diversification index based on Herfindhal index of transnational
patents across 35 technological fields. This index is based on Lee (201323) who shows that
catching up from middle income to high income status is accompanied by diversification of
technological knowledge.
We should expect that latecomer economies have initially highly concentrated structure of
patents which are diversifying as they are upgrading technologically i.e the number of patents
categories with patents is increasing. This process should be present in the case of both
resident and transnational patents. However, we would expect that dispersion of technology
effort should be more pronounced in the case of transnational than resident patents. Also, we
may expect that as countries are catching up that they are increasingly involved in high growth
patenting areas which are in high tech categories and in knowledge intensive services areas.
21 Krueger, Jens J. (2008) Productivity and structural change: a review of the literature, Journal of Economic Surveys
(2008) Vol. 22, No. 2, pp. 330–363
22 Michael Peneder (2003), Industrial structure and aggregate growth, Structural Change and Economic Dynamics,
14 (2003): 427-448
23 Lee, K. (2013) Schumpeterian Analysis of Economic Catch-up Knowledge, Path-Creation, and the Middle-Income
Trap. Cambridge University Press.
11
2.3. TECHNOLOGY UPGRADING AND INTERACTION WITH GLOBAL ECONOMY
A successful technology upgrading is never entirely autonomous process but is always linked to
inflow of foreign knowledge skills, which are coupled with intensive domestic technology effort
(Radosevic, 199924). The key to catch-up and post-catch-up is leverage of domestic innovation
efforts with global industrial or knowledge networks25. Hence, magnitude of knowledge inflows
and their coupling to domestic innovations efforts are important dimensions of technology
upgrading. A globalisation of technology exploitation and collaboration but also technology
generation through globalization of R&D process has further increased the importance of
international linkages for industrial upgrading (UNCTAD,200526). Drawing on the Cross-border
Ownership approach by Guellec/Pottelsbergue (2001, 2010) we use patent indicators to gauge
technology sourcing from foreigners as well as interaction or cooperation in technological
activity with foreign actors. Guellec/Pottelsbergue (200127, 201028) developed the concept of
the Cross-border Ownership to explore globalization of RD process. We use the indicators from
the perspective of technology upgrading which leads to slightly different interpretations.
Technology sourcing from a global perspective and the nature of interaction with foreign actors
change from the catch up to the post catch up stage, which is reflected in patent indicators. We
use three indicators to explore these processes.
Foreign Applications of Native Inventions (FANI) measure the extent to which technological
development in a country or region is driven by foreign actors. This is primarily important in the
catch-up phase of host countries. If we assume that inventors have the technological
capabilities and applicants exploit these capabilities commercially, this indicator is a proxy for
the involvement of foreign actors in the exploitation of native technological capabilities.
International Co-invention in technological activities (COINV) measure international
collaboration using patent applications with inventors residing in different countries. The share
of patents involving inventors from different countries shows the degree to which knowledge
generation is internationalized.
Native Applications of Foreign Inventions (NAFI) is a proxy for the exploitation of technological
capabilities abroad as it measures the extent to which technological development in a country
is making use of knowledge or technology sourcing from abroad. Arguably, this element
becomes increasingly important in the later stages of the catch-up phase of host countries and
might characterize high-income host countries. In that respect, it may be expected that
24 Radosevic, S. (1999) International technology transfer and catch-up in economic development, Edward Elgar,
Cheltenham.
25 Ernst, Dieter (2008) Asia’s “upgrading through innovation” 31 strategies and global innovation networks: an
extension of Sanjaya Lall’s research agenda, Transnational Corporations, Volume 17, Number 3, December 2008
26 UNCTAD (2005) World Investment Report, UN, Geneva
27 Guellec, D. and van Pottelsberghe de la Potterie, B. 2001. The internationalisation of technology analysed with
patent data. Research Policy, 30, 1253-1266.
28 Guellec, D. and van Pottelsberghe de la Potterie, B. 2010, "Measuring the internationalisation of the generation
of knowledge. An approach based on patent data.," In Handbook of quantitative science and technology research.
The use of publication and patent statistics in studies on S&T systems, H. F. Moed, W. Glänzel, & U. Schmoch, eds.,
Dordrecht: Kluwer Academic Publishers, pp. 645-662
12
countries behind technology frontier have high share of FANI, are increasingly involved in
COINV and have smaller share of NAFI. As they are technology upgrading it may be expected
that share of FANI declines, while shares of COINV and NAFI are increasing.
3. TECHNOLOGY UPGRADING THROUGH PATENT DATA: COMPARATIVE
ANALYSIS
Based on this conceptual framework in this section we analyze patterns of technology
upgrading of the CEECs but in a comparative context of the EU28 and BRICS economies. We
consider CEE countries individually as well as groups of EU countries to analyze the
convergence process in Europe between 1990 and 2011. The groups of countries considered
are EU12, EU South and CEE. Indicators for these group of countries are built using the average
across countries within the respective group (CEE, EU12, South EU). Moreover, CEE are
compared to BRICS. We follow dimensions of technology upgrading as explained in section 2.
3.1. Intensity of technology upgrading
As mentioned above, intensity of technology upgrading is reflected on the technology capability
of the country. To capture domestic technological capability pushing the technology frontier we
rely on transnational patent applications of domestic applicants compiled from the OECD
RegPat Dabase (Version January 2014). To capture technological capability for technological
development behind the technology frontier we use direct patent applications by residents to
their respective national patent offices. The World International Patent Office (WIPO) provides
with data on direct applications by resident applicants to their national offices.
3.1.1. Technological capability pushing the technology frontier
Drawing on Frietsch and Jung (200929) the counts of transnational patents (TN) include all PCT
applications whether transferred to the EPO or not and all direct EPO applications without
precursor PCT application.30 We consider two indicators: Transnational patent applications per
GDP (TNpGDP) and Transnational Patent Applications per capita (TNpc). TNpGDP captures the
technology intensity of the economy at the technological frontier. TNpc capture the technology
intensity of the country. Figure 4 includes the indicators for different CEE countries and group
of countries.
29 Frietsch, R. and Jung, T. 2009, Transnational Patents - Structures, Trends and Recent Developments,
Expertenkommission Forschung und Innovation, Berlin, Studien zum Deutschen Innovationssystem. 7 - 2009.
Available at http://www.e-fi.de/fileadmin/Studien/StuDIS2009/7_2009_Patentreport_ISI.pdf
30 The origin of the invention is defined by the country of residence of the applicants. The indicators use the
applicant country for the geographic designation of the invention in order to be consistent with the data available
from WIPO. The application year (rather than the priority year) is considered for the same reason. If an invention
involves applicants from different countries each country will be assigned with one application (and not a fraction
of it).
13
Figure 4: Indicators to capture technological capability pushing the technology frontier
Source: OCED RegPat, World Bank and authors’ calculations.
14
In what concerns the European Union, per GDP and per capita indicators of technological
intensity display strong growth from early 1990s and deceleration of this process after 2008.
Within EU the data suggest a divergence on core and periphery countries. This is especially
present in terms of growth of patents until 2008 when patenting in the developed EU12 slows
down.
The comparison of CEE with BRICS suggests that in pc terms CEE has higher ‘technology
intensity of country’ (not economy) than China. China’s catch up started in 2000s not in 1990s
as CEE. So, this is quite recent phenomenon which is telling about technology upgrading of
China. A strong catch up CEE in per capita terms is lost in GDP terms while Chinese is not. In
other words, CEEC as countries have become more patent (technology) intensive but not as
economies. The increasing/decreasing gap between TN patenting in pc and GDP terms is an
indicative proxy for increasing or decreasing alignment or misalignment of their National
Innovation Systems (Tunzelmann et al. 2012)31.
In the group of CEE countries Slovenia is the clear leader in terms of transnational patents per
capita. This can be reflection of its very high relative GERD, its high income but also its profile of
R&D system which may be geared more towards patentable sectors especially pharma and
chemicals (OECD, 2012: 10832). Estonia is second leader largely. Slovenia is outlier in per capita
terms but joined with Estonia in GDP terms. Both countries are still above China but given
differences in size this is remarkable for China and puts all CEE successes in perspective. Among
CEECs, it is interesting to see that continuous growth of Poland is reflected in transnational
patents per GDP. Given still very small numbers we consider this to be the reflection rather
than driver of growth.
3.1.2. Technological capability behind the technology frontier
Analog to the use of transnational patents we build per GDP and per capita indicators for the
period 1990-2012. Figure 5 presents the patent indicators per application year.
31 Tunzelmann, N. von; Günther, Jutta; Wilde, Katja; Jindra, Björn: Interactive Dynamic Capabilities and
Regenerating the East German Innovation System, in: Contributions to Political Economy, Vol. 29 (1), 2010, pp. 87-
110
32 OECD (2012) OECD Review of Innovation Policy: Slovenia, OECD. Paris
15
Figure 5: Indicators to capture technological capability behind the technology frontier
Source: OCED RegPat, World Bank and authors’ calculations.
Resident direct patent applications
per capita.
Sources: WIPO, World Bank
0
50
100
150
200
250
300
350
EU12
CEE
SouthEU
0
50
100
150
200
250
300
350
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
CEE
Brazil
Russia
India
CN
South Africa
0
50
100
150
200
250
300
BG
HR
CZ
EE
HU
LV
LT
PL
RO
SK
SI
Resident direct patent applications
per GDP (1990-2011). GDP in $ cp 2005.
Sources: WIPO, World Bank
0
2
4
6
8
10
12
14
EU12
CEE
SouthEU
0
20
40
60
80
100
120
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
CEE
Brazil
Russia
India
CN
South Africa
0
5
10
15
20
25
30
35
BG
HR
CZ
EE
HU
LV
LT
PL
RO
SK
SI
16
In overall, there seems to be much less increase in technology intensity of country in terms of
direct applications to national offices (behind the frontier effort) than in terms of TN patents (at
world frontier). This is expected given decrease in demand for domestic behind frontier effort
when compared to imported technology. A stagnant trend in EU28 and in its subregions shows
the declining importance of technology efforts oriented towards local/national markets (see
figures 1 and 2 above). This may be expected given continuous economic and institutional
changes towards European research area and effects of industrial networks in the EU,
especially between Germany, Austria and Central Europe. Some increase in CE and South EU
after 2008 is difficult to interpret except as the effect of Structural Funds (at least in CEE and
increase in GERD/GDP ratios).
A higher number of direct resident applications per GDP in CEE when compared to the EU12
shows that in terms of behind the frontier technology effort CEE were high in early 1990s,
especially given significant decreases in their GDP. On the other hand, a decline of resident
patents per GDP in CEE shows increasing internationalization of their economies where behind
the technology frontier effort is being increasingly squeezed by opening of their innovation
systems. Hence, we observe a strong convergence. However, it seems that the level has now
stabilized and even slightly increased as the effect of 2008. This is also the case in the EU South.
In what concerns the comparison between CEE with BRICS, China shows a strong increase of
both behind and on the frontier technology effort. So, in the case of China we do not observe
hyper integration features of India or closed economy of Russia but there are elements of
coupling between domestic and technology frontier. Russia is unique in its persistent and high
levels of behind the frontier technology effort. This is quite expected given the nature of its
system of innovation
Within CEE Slovenia is again leader in terms of ‘technology intensity of country’ (not economy):
A strong increase after 2008 in Slovenia is probably due to effects of Structural Funds in support
of domestic RTD system, especially centres of excellence and competence centres.
17
3.2. Breadth of technology upgrading
To analyze breadth of technology upgrading we focus on features of structural change. This is
about widening ‘surface’ of technology efforts or increasing number of technology areas in
which countries get involved or patent as they progress in technology upgrading. We define
two structural change indicators to measure this process: (i) the relevance of high technology
and knowledge intensive services patents in the technological activities and (ii) the
diversification of the technological activities across 35 technological fields.
3.2.1. High Tech Knowledge Intensive Patents
Using transnational patent applications we consider the share of patents in the high technology
fields and knowledge intensive services (HKTI). To define high technology we use the EUROSTAT
definition.33 The indicator used is the share of HTKI patent applications to the total patent
output in the country per application year. We frame HTKI patents as patents that reflect high
growth technology areas or ‘dynamic technology frontier patenting activities’.
Figure 6: Share of HTKI Patents in total patent output per application year (3 Years MA)
Source: OCED RegPat and authors’ calculations.
Figure 6 shows the indicator for different countries and groups of countries. The share of HTKI
patents at technology frontier is on average 6% in CEE, 11.4% in EU15 and 6.1% in South EU. In
the EU periphery technology activities in currently growing and dynamic areas related to ICT
presumably are underrepresented. This seems to correspond to an analysis on based priority
patents (Dominguez Lacasa and Giebler 2014)34. However, there is a positive structural change
of shifting towards HTKI areas which is strikingly similar in both EU South and CEE. A decline in
33 http://ec.europa.eu/eurostat/cache/metadata/Annexes/htec_esms_an6.pdf (last accessed 13.01.2015)
34 Dominguez Lacasa, Iciar; Giebler, Alexander (2014) Technological Activities in CEE Countries: A Patent Analysis
for the Period 1980-2009, IWH Diskussionspapiere 2/2014, S. 1-36. Available at http://www.iwh-
halle.de/d/publik/disc/2-14.pdf (last accessed January 2015)
0
5
10
15
20
25
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
EU12
CEE
SouthEU
0
5
10
15
20
25
30
35
40
45
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
CEE
Brazil
Russia
India
CN
South Africa
0
5
10
15
20
25
30
BG
HR
CZ
EE
HU
LV
LT
PL
RO
SK
SI
18
share of HTKI areas at the EU15 level shows that technology path of EU is quite different from
US or East Asia.
BRICs shows a gradual but upward increase in the share of HTKI patents with China having the
highest share but also decline in the share after 2007/08. This maybe reflects a changing
orientation of Chinese growth towards more domestic technology based growth (after 2008)
and technology diversification in transition from middle to high income as argued by Lee
(201335). The indicator for China nicely shows that its boom does not have anything to do with
dot.com wave in 2001. The same holds for the other BRIC countries. This suggests that the
nature of globalization is largely about absorptive capacities of catching up countries not
catching up at frontier (Radosevic and Yoruk, 201436, OECD 201037). It is quite surprising to
observe a low share of HTKI of India given possible hypothesis on ‘hyperintegrationist’ mode of
development as opposed to China. Interestingly, they are on two different sides within the
BRICS spectrum. Within the BRICS the biggest surprise is India which shows that its
technological strengths in services are not yet in services that can be captured by patent
indicators. Its export of software is not of patentable type.
In general terms CEE falls clearly within BRIC spectrum even in terms of secular increase in
share of HTKI areas.
3.2.2 Technology diversification
Drawing on Lee’s (201338 ) idea that catching up process translate into an increasing
diversification of technological activities we aim at analysing trends in diversification of
technological capabilities. To measure technological diversification we use the Herfindhal index
of transnational patent applications and resident direct patent applications to the national
offices across 35 technological fields (Schmoch 2008)39. The assignment of an invention to a
technological sector or specific technology field follows a fractional counting methodology.40
The Herfindhal index is normalized between 0 and 1. Values close to 1 mean concentration.
Values close to 0 mean diversification.
35 Lee Keun (2013) Schumpeterian Analysis of Economic Catch-up: Knowledge, Path Creation, and the Middle-
income Trap, Cambridge University Press.
36 Slavo Radosevic and Esin Yoruk (2014) Are there global shifts in world science base? Analysis of catching up and
falling behind of world regions, Scientometrics, June, DOI 10.1007/s11192-014-1344-1
37 OECD (2010) Global Perspectives on Development, Shifting Wealth , OECD, Paris.
38 Lee, K. (2013) Schumpeterian Analysis of Economic Catch-up Knowledge, Path-Creation, and the Middle-Income
Trap. Cambridge University Press.
39 http://www.wipo.int/ipstats/en/statistics/technology_concordance.html (last accessed 13.01.2015)
40 If a transnational patent application includes patent classes that belong to different technological areas or
technologies a fraction (and not a whole count) will be considered for each technological area or technology.
19
Figure 7: HH-Index: Transnational Patents (3 Years MA)
Source: OCED RegPat and authors’ calculations.
Figure 8: HH-Index: resident direct patent applications to the national offices (3 Years MA)
Source: OCED RegPat and authors’ calculations.
The indicators presented in Figures 7 and Figure 8 suggest a clear trend towards diversification
which is in line with Lee’s (201341) hypothesis and results except for China for TN patents and
for India for domestic patents. .
For TN patenting, trends in the EU periphery shows strong diversification though at a somewhat
higher levels of concentration in CEECs than in the EU South. A diversification trends is feature
of all CEECs despite their quite different starting levels of concentration/diversification. There is
strong convergence of both the EU South and CEE to the core which presumably should mean
that the overall technological knowledge structure in the EU is becoming strongly determined
by the EU core. However, this trend has slowed down significantly after 2001 despite economic
growth which was high until 2008.
Trends towards diversification of technological knowledge are also feature of the BRICs except
China after 2000. First, we observe very strong diversification of India and CEE which suggest
technology upgrading via diversification. Second, there is a very slow diversification trend in
Russia, Brazil and South Africa which may reflect slow structural change in their technology
systems. Third, China shows opposite trend – towards concentration or decrease in number of
transnational patent categories. How do we interpret this seemingly counterintuitive trend?
41 Lee, K. (2013) Schumpeterian Analysis of Economic Catch-up Knowledge, Path-Creation, and the Middle-Income
Trap. Cambridge University Press.
20
Has China already moved towards technology structure of the upper income economies? Lee
(2014) shows that diversification is trend in transition from middle to upper income stage after
which countries continue to specialize. China does not seem to conform to this trend.
As we would expect diversification is much less pronounced in resident patenting which largely
reflect domestic and behind the frontier technology effort. The slow tendency towards
diversification is present in all countries with exception of India after 1997, South EU after 2001
and China after 2004. Without in depth analyses of each regions technology systems it is quite
difficult to interpret structural changes in generation of technological knowledge behind
technology frontier. Also, we see need for further research in exploring diverging vs. converging
trends between structural change of TN and resident patenting.
3.3. Interaction with the Global Economy
In general, the key idea here is to use patent based indicators to gauge technology and
knowledge flows as well as interaction or cooperation in technological activity with foreign
actors. The flows and the modes of interaction with foreign actors change along the catch up
process, which should be reflected in the indicators. We use three indicators originally
developed by Guellec/Pottelsbergue (200142, 201043) to track technology sourcing from a global
perspective and international knowledge cooperation.
3.3.1. Foreign applications of National inventions (FANI)
FANI shows the share of TN patents that are invented by inventors in country x but applicants
are from country y. Guellec/Pottelsbergue (200144, 201045) interpreted the indicators as the
extent to which technological development in a country or region is driven by foreign actors. A
large FANI Rate suggests the strong importance of foreign actors exploiting the technological
activities of a country or region. A low FANI Rate suggests that native inventions are mainly
applied by native actors. If we assume that inventors have the technological capabilities and
applicants have commercial and organizational capabilities this indicator can tell us something
about the relationship between technical and non-technical capabilities. According to Teece
(1986) for successful innovation and technological development at the firm level it is not
42 Guellec, D. and van Pottelsberghe de la Potterie, B. 2001. The internationalisation of technology analysed with
patent data. Research Policy, 30, 1253-1266.
43 Guellec, D. and van Pottelsberghe de la Potterie, B. 2010, "Measuring the internationalisation of the generation
of knowledge. An approach based on patent data.," In Handbook of quantitative science and technology research.
The use of publication and patent statistics in studies on S&T systems, H. F. Moed, W. Glänzel, & U. Schmoch, eds.,
Dordrecht: Kluwer Academic Publishers, pp. 645-662.
44 Guellec, D. and van Pottelsberghe de la Potterie, B. 2001. The internationalisation of technology analysed with
patent data. Research Policy, 30, 1253-1266.
45 Guellec, D. and van Pottelsberghe de la Potterie, B. 2010, "Measuring the internationalisation of the generation
of knowledge. An approach based on patent data.," In Handbook of quantitative science and technology research.
The use of publication and patent statistics in studies on S&T systems, H. F. Moed, W. Glänzel, & U. Schmoch, eds.,
Dordrecht: Kluwer Academic Publishers, pp. 645-662.
21
enough to have technology capabilities but also complementary assets to put these capabilities
into use. At firm levels this means organisational capabilities in addition to only invention
capacity. His answer to who actually profits from innovation, pointed to owners of
complementary assets, particularly when they are specialized and/or co-specialized. So,
following Teece (1986)46 we interpret large FANI as a proxy for organisational capabilities of
firms or individuals to commercialize inventions on their own. For firms that are applicants of
foreign inventions this indicates presence of organisational capabilities to commercialize
inventions as well as understanding of available technological inventions abroad which are
patentable.
From the perspective of complementary or organizational capabilities, a declining FANI rate all
else equal is a sign of upgrading in complementary or organisational capabilities in the country
or capacity to profit from their technological activities. Figure 9 presents the indicator for
different countries and group of countries.
Figure 9: Rate of Foreign Applications of Native Inventions (FANI Rate) (3 year moving
averages)
Source: OCED RegPat and authors’ calculations.
From the complementary or organisational capabilities view, the sudden increase in FANI Rate
in 1990-1993 in CEE is a reflection of weak organisational capabilities of firms in newly opened
economies to handle invention process on their own and also of better understanding of
foreigners what are available technological inventions which are patentable. However, situation
has stabilised and if we take mid-1990s as the beginning of normal period we do not observe
improvements in organisational or complementary capabilities. In fact, average between 1995-
1998 and 2010-2012 shows a minor decline in all CEECs. We observe similar weakening of
complementary capabilities in South EU as well as in the EU12. This trend can be a reflection of
weakening of these capabilities across Europe (i.e. of declining role of EU large firms as
organisers of innovation processes) but this can also be a reflection of globalisation of
innovation process.
46 Teece, David J. 1986, Profiting from technological innovation: Implications for integration, collaboration,
licensing and public policy. Research Policy 15 (6): 285-305.
22
Given our interpretation of FANI we would expect that successful technology upgrading would
be reflected in decreasing FANI. Data for BRICs and CEE are in line with this hypothesis. For
example, China’s FANI rates have declined dramatically reflecting organisational power of
Chinese MNEs. Indian complementary capabilities as reflected in FANI have improved until
2001/2002 (dot.com period) and have declined afterwards as reflected in increased FANI
indices. Russian and especially Brazilian FANI Indices are gradually and slowly decreasing
reflecting gradually improving complementary capabilities of their firms, especially MNEs.
Within BRIC context CEE FANI rates seems quite stagnant reflecting possibly very weak
endogenous organisational capabilities i.e a low share of domestic large firms in technology
activities.
3.3.2. Indicators for Knowledge Cooperation: Coinventions (COINV)
As countries upgrade technologically their capability for joint international generation of
inventions should increase. An increase in joint patents also reflects changing nature of
invention process which is becoming more globalized as depicted also by FANI and NAFI
indicators. Guellec/Pottelsbergue (200147, 201048) measure international collaboration using
patent applications with inventors residing in different countries: the share of patents resulting
from international research co-operation (inventors from different countries) in the total
number of patents invented by residents of a given country. Here we use identical measure.
Figure 10: Share of International Co-Inventions (COINV Rate) (3 year MA)
Source: OCED RegPat and authors’ calculations.
As given in Figure 10, the indicator shows significant globalization of knowledge generation in
the EU and in its three sub-regions. By 2012 in all three sub-regions of the EU around 40% of all
TN patents applications involve at least one foreign and one domestic inventor (COINV).
47 Guellec, D. and van Pottelsberghe de la Potterie, B. 2001. The internationalisation of technology analysed with
patent data. Research Policy, 30, 1253-1266.
48 Guellec, D. and van Pottelsberghe de la Potterie, B. 2010, "Measuring the internationalisation of the generation
of knowledge. An approach based on patent data.," In Handbook of quantitative science and technology research.
The use of publication and patent statistics in studies on S&T systems, H. F. Moed, W. Glänzel, & U. Schmoch, eds.,
Dordrecht: Kluwer Academic Publishers, pp. 645-662.
23
However, there are significant differences in trends between three sub regions. At EU
periphery there seems to be stagnation in COINV rates after 2001 (South EU) but especially
after 2008 (South and CEE). This may possibly reflect the effect of worsening of macroeconomic
conditions after 2008 on R&D based investment and thus on technology knowledge co-
generation.
Levels of technology co-generation are lower in BRICS than in the CEE and the rest of the EU.
Among BRICS China is distinctive as its share of co-inventions declines continually reflecting
much stronger patenting by Chinese companies themselves. Hence, this relative decline should
not be confused with absolute very strong growth of Chinese TN patents. Russia and Brazil
again have similar trend of stagnant COINV rate but given size of these economies the share of
co-inventing is actually quite high. India’s patenting was during the 1990s more than half based
on co-inventions but COINV was also rapidly declining reflecting increasing indigenous
technological capabilities. After 2001 India has been increasingly involved in technology
cooperation at very high level for such large economy. Again, compared to China it is on the
other side of the BRIC spectrum. Its share of technology co-inventions is similar now to the CEE
which is a much smaller region.
3.3.3. National Application of foreign Inventions (NAFI)
Drawing again on Guellec/Pottelsbergue (2001 49 , 2010 50 ) we compute the share of
transnational patent applications with applicants located in a country that involve at least one
inventor located abroad. This indicator is a proxy for the exploitation of technological
capabilities abroad (Native Applications of Foreign Inventions - NAFI). These patent-based
indicators aim at measuring the extent to which technological development in a country is
making use of knowledge or technology sourcing from abroad. Arguably, this element become
increasingly important in the later stages of the catch-up phase of host countries and might
characterize high-income host countries. The operationalization follows the logic outline for
FANI above. Counting transnational applications per application year, the number of
transnational patents applied by natives and invented by foreigners and (NAFI) is divided by the
total number of transnational patents with at least one national applicant (NAFI-Rate). From
the perspective of technology upgrading, we interpret the capacity of countries to source
technology from abroad as measured by NAFI as the sign of high or increase organizational
capabilities all else equal. A high or increased NAFI would indicate improvement in these
capabilities and vice versa.
49 Guellec, D. and van Pottelsberghe de la Potterie, B. 2001. The internationalisation of technology analysed with
patent data. Research Policy, 30, 1253-1266.
50 Guellec, D. and van Pottelsberghe de la Potterie, B. 2010, "Measuring the internationalisation of the generation
of knowledge. An approach based on patent data.," In Handbook of quantitative science and technology research.
The use of publication and patent statistics in studies on S&T systems, H. F. Moed, W. Glänzel, & U. Schmoch, eds.,
Dordrecht: Kluwer Academic Publishers, pp. 645-662.
24
Figure 11: Rate of Native Applications of Foreign Inventions (FANI Rate) (3 year MA)
Source: OCED RegPat and authors’ calculations.
Figure 11 includes NAFI rates for different countries and groups of countries. NAFI indices for
EU regions shows that technology sourcing abroad has initially declined in CEE and has
remained stagnant and at comparatively very low level since mid-1990s while it has increased
significantly at EU12 and South EU. Surprisingly levels of NAFI for EU12 and South EU are
relatively similar which should reflect similar capacities for technology sourcing abroad. Among
CEECs, there were initial differences in NAFI but these have been gradually converging as times
goes by. NAFI, which in our context denote capacities for technology sourcing abroad, have
been stagnant in BRICs which may seems surprising given the newly emerging literature and
evidence on emerging markets MNEs, some of which have relied on technology sourcing as one
of their strategies orientations. In particular, declining NAFI of China seems to suggest that
despite individual high profile cases of BRICS MNEs sourcing technology abroad these cases do
not yet represent trend or technology sourcing is not their key strategic orientation. However,
we should bear in mind that NAFI or share of transnational patent applications with applicants
located in a country that involve at least one inventor located abroad is quotient and we should
bear in mind that it is dependent on total number of TN patents. A catching up country that has
high and growing number of TN patents but still low number of its patents invented abroad is
actually doing still better than country that has high NAFI but low number of its TN patents. This
is exactly the case between the CEE and China where former has higher NAFI but much lower
number of TN patents.
4. TECHNOLOGICAL UPGRADING IN EUROPE IN A COMPARATIVE PERSPECTIVE
In this section we merge three dimensions and all indicators into one graphic form – network
diagram - to explore levels and patterns of changes of technology upgrading. Each graphic
includes 7 indicators. The technological intensity of a country is represented by the number of
patent applications by residents at the national filing office per GDP (domestic technological
intensity) and the by the number of transnational patent applications by national applicants per
GDP (frontier technological intensity). The breadth of technological upgrading is represented
by share of high tech knowledge intensive transnational patent applications in total
transnational patent applications (High tech patents) and the degree of concentration of patent
25
applications by residents at the national filing of the country across 35 technological areas
(specialisation). The technological interaction with the global economy is represented by three
indicators: The share of applications with at least one national applicant and at least one
foreign inventor in total transnational inventions filed by at least one national applicant (NAFI
Rate); the share of foreign applications with at least one foreign applicant and at least one
national inventor in total transnational applications with at least one national inventor (FANI
Rate); and the share of transnational patent applications involving at least one foreign as well
as one domestic inventor in the total number of transnational patent applications invented by
at least one native (COINV Rate).
First, we analyze each of the CEE countries in comparison to other EU countries at a particular
point in time (2011). In a second section we explore the position of the CEE in relation to BRICS
using identical approach.
4.1. Technological upgrading in the EU
We consider the seven indicators for CEE countries for the year 2011. In addition, we indicate
the relative change in percent for each indicator for the respective CEE country in comparison
to the year 1995 (or the latest available). In the diagrams we compare each of the eleven CEE
countries to the other ten CEE countries, South European countries as well as EU12 countries.
The values for each indicator used for graphical representation are scaled between 0 and 1
using all country values for 26 EU countries . Then we generated simple unweighted average for
the other ten CEE countries, the group of South European economies as well as the group of
the EU12 countries. Thus, the graphical space represented by the seven dimensions in each of
the diagrams corresponds to the possible maximum values by 26 EU countries at the point of
observation (2011).
Figure 12 and Figure 13 bellow show the profile of technology upgrading of the individual CEECs
in relation to the EU12, South EU and other CEECs. We do not go into detailed description of
profiles of each of 11 CEECs but draw only two general conclusions. First, technology upgrading
profiles of the CEECs are pretty homogenous which reflects their technological levels and
relative distance to the EU-12. A typical CEE economy is well behind EU12 in terms of frontier
technology intensity, domestic technological intensity, share of high tech patents and
technology sourcing abroad (NAFI). Its organizational capabilities are often less advanced as
reflected in high share of FANI. The CEE profile is much less coherent in terms of technology
diversification/specialization and share of joint inventions. Second, differences among CEECs
are not significant in the sense that we can talk of distinct national technology profiles. Poland,
Romania and Slovenia have above average domestic technological intensity which reflects
partly their sizes (Romania and Poland) and specific model of innovation system reliant on
domestic R&D intensive firms (Slovenia). Latvia and Lithuania are specific in terms of high share
of HTKI patents.
26
Figure 12: 2011 Indicators for Bulgaria, Czech Republic, Estonia, Croatia, Hungary and
Lithuania. Comparison with EU 12, South EU and other CEE
Source: OCED RegPat, WIPO, World Bank and authors’ calculations.
27
Figure 13: 2011 Indicators for Latvia, Poland, Romania, Slovenia, and Slovakia. Comparison
with EU 12, South EU and other CEE
Source: OCED RegPat, WIPO, World Bank and authors’ calculations.
28
4.2. Technological upgrading of Emerging Economies
Again, using network diagrams we aim at a graphical presentation of changes in the selected
indicators for each of the three dimensions of technological upgrading for the CEE region in
comparison to the BRICS countries between 1995 and 2011.
First we create two summary network diagrams that integrate all countries under observation
in 1995 and 2011 to show the change in structural indicators in these selected emerging
economies and the CEE region (Figure 14). Next, we offer a diagram for the CEE region and
each of the BRICS countries (Figure 15) based on seven indicators in 1995 and 2011. The values
for BRICS are country specific. For the CEE region we create a simple unweighted average for
across the eleven CEE countries. Before drawing the graphs, we scale all indicators for the
BRICS countries and the CEE region between 0 and 1. Thus, the graphical space represented by
the seven dimensions in each of the diagrams corresponds to the possible maximum values by
the BRICS countries and the CEE region in 1995 and 2011.
Figure 14: Indicators for BRICS and CEE (average) in 1995 and 2011
Source: OCED RegPat, WIPO, World Bank and authors’ calculations.
A comparison of the CEECs and BRICS profiles in 1995 and 2011 offers few very interesting
insights. First, 1995 profiles are more diverse than 2011 reflecting divergences and
convergences among these catching-up economies. In 1995, Russia had distinctive prolife
characterised by comparatively the highest both domestic and frontier technological intensity
and together with China the highest share of high tech patents. CEE had the least diversified
technological knowledge portfolio with comparatively high frontier technological intensity.
China had the highest FANI rate which by 2011 became the lowest next to Brazil reflecting
increase organisational capabilities of their MNEs to commercialize their own inventions. India
had very low ranking on all dimensions of technological upgrading except in terms of NAFI or
sourcing technology abroad. This quite diverse set of profiles changed significantly by 2011.
29
China has delinked from BRICS by highly increased domestic and frontier technological intensity
as well as by very high share of high-tech patents. CEE has lost its initial high ranking in terms
of frontier technological intensity, has significantly diversified its technological knowledge,
increased co invention rate but also became the region with the highest FANI rate which
reflects weak organisational capabilities to commercialize its own inventions. India has
continued to be comparatively the strongest in sourcing technology abroad but it also reduced
diversification of its technology portfolio of inventions. Other BRICS – Russia, Brazil and South
Africa – have features which fall within these three specific cases of China, CEE and India. Russia
has lost its advantages in terms of the highest frontier and domestic technological intensity. In
overall, we have seen a shift from much more diverse technology upgrading profiles in 1995
towards four profiles: China, CEE, India and rest of BRICS (Russia, Brazil and South Africa).
Next, we explore in greater detail changes in between 1995 and 2011 by each of BRICs and CEE
(Figure 15). CEE technology upgrading profile has substantially changed in between 1995 and
2011. Its technology intensity, both at frontier and domestic, has been declined and its
openness has significantly changed as shown by increased co-invention, NAFI and FANI rates.
On positive side, its technology profile has diversified as should be expected when countries are
transiting from middle towards high income status. Also, its capacity for sourcing technology
abroad has also somewhat improved. However, invention process in CEE has become much less
intensive but it is now taking place in cooperation with foreign partners (COINV) who have
organisational capabilities to commercialize local inventions (NAFI). The CEE case contains
interesting lessons regarding costs and benefits in terms of openness and autonomy of
technology systems.
Changes in profile of Russian technology upgrading have been similar but also much more
dramatic when compared to the CEE. First, its decline of frontier and domestic technological
intensity has been much sharper than in the CEE. Also its share of high tech patents has
significantly declined. This loss of technology intensity of CEE has been compensated by stringer
interaction with global economy through high coinvention rate which was not the case in
Russia. Also, its FANI and NAFI rates have remained relatively unchanged. As in CEE, there has
been positive tendency of increased technological diversification.
China’s profile of technology upgrading shows very strong increase in both domestic as well as
in frontier technological intensity at the same share of high tech patents. On the other hand,
technological upgrading was not followed by its increased technological openness. Its
coinvention rate has dropped significantly and its capacity for sourcing technology abroad has
declined somewhat. FANI rate for China has declined dramatically which actually shows
increased capability of its MNEs to commercialize their own inventions. Given huge increases in
China’s technological intensity this dimension of interaction with global economy should be
seen in relative terms as relatively less intensive given much higher increase in technological
intensity. In this respect, a Chinese model of technology upgrading is quite different from the
CEE which had to compensate its decreasing technological intensity by more technological
openness.
30
India has very low technological intensity which despite its high economic growth in this period
has further shrank questioning whether its further growth can rely on technology or on other
production factors. Similar to CEE India has to compensate much less dramatic loss of
technology intensity by increase knowledge cogeneration (COINV). Its capacity to source
technology from abroad has remained constant but its technology portfolio has further
concentrated which is not the best basis for technology upgrading of such a large economy.
Changes in Brazilian technology upgrading profile have been much less intensive when
compared to China, Russia, and CEE. Relatively small decreases in technology intensity and in
share of high tech patents have also resulted like in CEE and India to increases in knowledge
cogeneration at relatively similar NAFI and FANI rates. South Africa followed similar pattern as
its domestic and especially frontier technology intensity have declined as well as share of high
tech patents. As in Brazil, CEE and India this has led to increases in co-invention rates and with
slight changes in NAFI and FANI rates at unchanged degree of technology diversification.
Figure 15: Indicators for BRICS and CEE (average) in 1995 and 2011
Source: OCED RegPat, WIPO, World Bank and authors’ calculations.
31
5. EXPLORING DIMENSIONS OF TECHNOLOGICAL UPGRADING
So far, we focused on identifying differences and changes to the technological profile of CEE
countries in comparison to other European economies and the BRICS countries by using
selected patent based indicators. In the final part of the analysis, we aim to explore all three
dimension of technological upgrading: technology intensity, structural changes and knowledge
interaction with global economy. We have departed from the proposition that technology
upgrading is multidimensional process and these three dimensions are different facets of this
complex process.
Statistically, it is possible to create a simple composite indicator of technological upgrading
based on selected indicators across countries within the observation period. In order to this, we
need to make some assumptions about the relation between our indicators and technological
upgrading. This seems straight forward in case of technological capability or intensity. Here we
assume that a higher value for technological intensity with regard to domestic or frontier
technology corresponds to a higher stage in technological development of the country. Second
we assume that the breadth of technological upgrading is higher, if the share of high tech
knowledge intensive patents in transnational patent applications is higher as well as the degree
of diversification of domestic technological activity across technological areas is higher. Finally,
we assume that higher NAFI rates (i.e. transnational patent applications with national
applicants and foreign inventors) correspond to stages of higher technological development as
capacity of countries to source technology globally increases. In turn, we assume that lower
FANI rates (i.e. transnational patent applications with foreign applicants and national inventors
in total transnational inventions) corresponds to stages of higher technological development as
countries organisation capabilities to commercialize inventions generated in their own country
increases. Finally, we assume that COINV Rates (i.e. transnational patent applications involving
at least one foreign as well as one domestic inventor in the total number of transnational
patent applications with national applicants) should decline as countries develop technology
capability to invent but also to commercialize their own inventions.
However, we think that constructing a composite indicator of technology upgrading would defy
our main analytical aim in this paper which is to understand the interactions between different
dimensions of technology upgrading and their changes. ‘Burying’ different dimensions and their
interactions into one composite indicator is in contradiction to our departing proposition to
build metrics which takes into account different drivers of technology upgrading. Synthesizing
three relatively independent but related processes – technology intensification, structural
changes and knowledge exchange - into one indicator leads to decontextualized metrics. Given
generally poor understanding of the processes of technology upgrading each of the above
stated assumptions can more or less stand scrutiny but only as a stylized fact on its own.
However, we are much less certain about their mutual interaction and whether the overall
construct or composite indicator of technology upgrading is really theoretically and statistically
grounded.
32
In view of these limitations as well as of the greater learning potential in exploring different
dimensions of technology upgrading, we present summary sub-indexes for each of three
dimensions of technology upgrading. Our analysis uses information for 1995 and 2011.
Following the above outlined assumptions we inverse the original values for the indicators
Herfindhal, FANI Rate and COINV Rates for each country. Using values for the year 1995 (or the
earliest year available) we rank each of the seven indicators separately, where the highest value
corresponds to the highest rank. Then we add the ranks across the relevant indicators for each
dimension for each country. The ranking of technological intensity is based on measures for
domestic and frontier patenting intensity. The measure of diversification of domestic inventions
and the relative importance of inventions in high tech and knowledge intensive activities are
grouped into the indicator for structural changes. Finally FANI, NAFI and COINV rates are group
into the ranking of interaction with the global economy. We give each indicator equal weigh
into one composite indicator for each dimension. The country with the lowest sum has the
highest overall rank per dimension. The procedure is repeated for the 2011 values. Finally we
can identify relative changes in the ranking for each of the country between 1995 and 2011 in
each dimension of technology upgrading. It is important to realize that this is not composite
indicator of the overall technology upgrading but of upgrading as reflected in patent data. In
that respect, this indicator shares all virtues and drawbacks of patents as indicators.
We need to acknowledge that the five indicators used to measure breadth and global
interaction of technological upgrading are measures independent from the underlying ‘size’ or
intensity of patenting activity. For example, similarly low FANI rates (i.e. high rankings) are
obtained in case of Malta and Finland in 2011. However, Finland has the second highest
technological intensity and Malta is ranked 22. The FANI rate is calculated with a base of 39
transnational patent applications in case of Malta and with 2.324 in case of Finland. The
distortion is amplified in case of the NAFI rates. As a result Malta comes in first on the ranking
for global interaction. Similar cases apply basically to all CEE countries in the year 1995 and to
the majority of smaller CEE countries (Baltic economies) still in 2011. Given the upward bias in
the rankings of structural change and global interaction for countries with low or very low
frontier or domestic technological intensity, we need to interpret the ranking dynamics of the
corresponding countries with appropriate caution.
Having these limitations in mind the ranking dynamics suggest the following:
1. Technological intensity: China has increased by far the most its patenting intensity due to
remarkable increase of both TN and resident patents. Still, Germany and Finland are two of the
most technology (patent) intensive economies. Given their income levels China and Slovenia
are surprisingly highly located. This indicates their high potential for technology upgrading but
also it shows that their current growth is not yet based on R&D. Russia’s relatively high position
is largely due to domestic technology effort. CEE (with exception of Slovenia) are firmly in the
second half of table together with South EU which is expected given that drivers of their growth
are not related to technology but to production capability.
33
Table 1: Dimension of technology upgrading: patent-based rankings 1995-2011
Technological intensity
Structural Change
Global Interaction
Country
1995
2011
Change
1995
2011
Change
1995
2011
Change
Germany
2
1
1
6
26
-20
13
11
2
Finland
1
2
-1
5
2
3
10
2
8
China
18
3
15
15
1
14
19
23
-4
Slovenia
7
4
3
16
30
-14
5
16
-11
France
8
5
3
3
5
-2
14
9
5
Denmark
12
6
6
12
15
-3
12
8
4
Austria
10
7
3
21
20
1
11
13
-2
Sweden
3
8
-5
2
6
-4
6
3
3
Latvia
14
9
5
30
31
-1
28
22
6
Netherlands
13
10
3
9
14
-5
3
5
-2
Russia
4
11
-7
7
29
-22
25
30
-5
Romania
16
12
4
31
32
-1
31
25
6
United Kingdom
6
13
-7
1
4
-3
18
18
0
Luxembourg
19
14
5
29
23
6
23
6
17
Hungary
9
15
-6
24
22
2
26
32
-6
Italy
15
16
-1
8
25
-17
4
26
-22
Estonia
33
17
11
32
8
24
30
12
18
Poland
20
18
2
19
13
6
27
29
-2
India
30
19
11
27
24
3
17
33
-16
Czech Republic
22
20
2
13
18
-5
9
27
-18
Bulgaria
5
21
-16
11
11
0
24
21
3
Malta
28
22
6
33
33
0
1
1
0
Belgium
21
23
-2
4
9
-5
16
14
2
Ireland
11
24
-13
14
10
4
15
4
11
Croatia
17
25
-8
18
21
-3
21
28
-7
Spain
25
26
-1
10
3
7
8
19
-11
Lithuania
26
27
-1
23
27
-4
33
10
23
Cyprus
27
28
-1
25
17
8
7
7
0
Brazil
29
29
0
22
12
10
22
31
-9
South Africa
23
30
-7
20
19
1
32
24
8
Greece
31
31
0
26
16
10
29
20
9
Portugal
32
32
0
28
7
21
2
17
-15
Slovakia
24
33
-9
17
28
-11
20
15
5
Sources: OCED RegPat, WIPO, World Bank and authors’ calculations. Authors calculation.
The underlying indicators for technological intensity are strongly shaped by industry structure
and favor those economies where ‘patenting industries’ like chemicals and pharma are
important. This partly explains the relatively high position of Slovenia. As technology intensity
measure does not differentiate between frontier and behind the frontier patenting some
economies will be higher than expected (Russia, Romania) or lower than expected (United
Kingdom, Ireland). Based on patenting intensity BRICs are not homogenous entity but widely
differing group with thus very different opportunities for growth based on technology.
34
Beside China the biggest improver in terms of technology intensity (in relative ranking) are
Estonia and India. Bulgaria and Ireland have fallen substantially behind similar to fall behind of
Russia, Romania, Croatia and South Africa.
2. Structural changes as depicted through indicators of patent diversification and shift towards
high tech patenting is favoring countries behind the frontier as they have much more scope for
convergence or reaching structure of the frontier economies. China and Estonia are again the
biggest improvers, which is quite important additional evidence of their technology upgrading
given that China is third ranked and Estonia 17th in terms of technological intensity. The biggest
improver in terms of structural change is actually Portugal but it has also a fairly low
technological intensity.
The smallest structural changes can be observed for Russia and Germany but for quite different
reasons. Germany is at the technology frontier and it may be expected that it will further
specialize. In fact, several technology intensive and high income economies are located very
low in terms of technology diversification (Germany, Austria, Denmark, Italy). Finland is quite
specific in the sense that it is technology intensive economy but also with high degree of
diversification of patent portfolio. As we would expect it has reached limits of diversification
and thus has not further improved in that respect. Russia in contrast lost considerable ground
in terms of frontier and domestic technological intensity, which seems to have been paralleled
by a narrowing diversification of domestic technological activity as well as a massive drop in the
share of high tech patenting.
As outlined above our underlying approach is aimed to measure technology upgrading of
middle income economies towards high income. This is clearly visible from changes in relative
position in terms of structural change where five economies from the bottom group in terms of
technology intensity are major diversifiers in expected direction while from the top group only
China belongs to the biggest diversifiers. Germany as economy at technology frontier has
reached saturation in that respect and has been actually specializing. So, within our framework
indicators of structural change do not have a priory positive or negative interpretation. This
depends on where countries stand in relation to the technology frontier.
3. Global interaction in patenting inventions is composed of three indicators (FANI, COINV and
NAFI rates) that indicate different stages and modes of interaction with global economy as
countries are technologically upgrading. So, identical change in degree of openness should be
interpreted in the context of technological level of economy and the actual mechanism of
interaction. The biggest changes in terms of increased openness in patenting activities took
place in Lithuania, Estonia, Luxemburg and Ireland while the biggest relative ‘withdrawals’ took
place in Italy, Portugal, Slovenia, Czech Republic and Spain. It is interesting to see that
‘globalizaton of technology’ is not universal process but evolves very unevenly reflecting very
much country specific interactions between national technology systems and external
environment.
35
The Chinese system is quite autonomous given its high technology intensity and direction of
structural change in patenting portfolio. Latvian technology system has generated in narrow
technology area high technology intensity but unlike neighboring Lithuania it is actually very
little open in terms of knowledge exchange. On the other hand, Finnish system is technology
intensive, quite diversified and also very open by being ranked second in terms of interaction
intensity. Also, Swedish system is quite open and relatively highly ranked in terms of both
diversity of patent portfolio and technology intensity. Slovenia as very technology intensive
economy has not opened in terms of knowledge exchange but it has actually closed further in
relative terms. Italy as large EU economy has further ‘delinked’ while its technology intensity
remains medium.
With the exception of South Africa the BRICS have in relative terms not further opened up but
actually have reduced their ranking positions in terms of global interactions at very different
levels of technology intensity. For example, China and India reduced the relative ranking
positions in terms of global interaction with increasing technological intensity, whereas Russia
lost relative ground in terms technological intensity as well as global interaction. This is
contrasted by the development in South Africa, which also observed drop of its relative position
in terms of technological intensity but at the same time score relatively higher in terms of
global interaction. This raises interesting issues about the role of autonomy and openness in
technology system in the catching-up process. However, our data only allows the interpretation
of changing positions in global interaction in relative terms looking at EU and BRICS economies.
Each of them may have become more or less open in their own terms as we observed above
(see for example Figure 11).
6. CONCLUSIONS
This report measures patterns of technology upgrading as three-dimensional process which
consists of (i) intensity of technology upgrading, (ii) structural change, and (iii) interaction with
the global economy. All three dimensions have strong grounding in the respective literatures on
firm level technology upgrading, on structural change and growth, and on integration of
technological activities in the global economy. We compare countries in terms of technological
levels and changes along their own upgrading paths as reflected in these three dimensions.
The specificity of our report is that, considering the 3 dimensions, we depict patterns of
technology upgrading by relying entirely on patent data. This has its major advantages in terms
of length and consistency of time series derived as well as in the possibility to identify
technological fields or specializations based on patent classifications.
The indicators for intensity of technology upgrading trace technological capabilities at the
technology frontier and behind the technology frontier. Transnational patent applications (TN)
capture inventions pushing the technology frontier while resident direct patent applications to
national patent offices dominantly proxy technology effort behind the technology frontier. It
may be expected that as countries technologically upgrade their patent intensity increases and
shifts form resident toward TN patents.
36
Structural change in technological knowledge is depicted by using transnational patent
applications in high technology fields and knowledge-intensive services and by a technological
diversification index based on Herfindhal index of transnational patents across 35 technological
fields. Drawing on Lee (201351) we assume that technology upgrading of middle income
economies is depicted by increasing diversification of their technology profiles in terms of
patents while this is not necessarily the case with high income economies.
To capture interaction with global economy in the upgrading process we focus on technological
knowledge sourcing across countries and interactions between foreign and indigenous actors.
We draw on indicators developed by Guellec/ Pottelsbergue (2001, 2010). We apply them for
exploring technology upgrading which leads to new perspectives in their interpretation.
Technology sourcing and the nature of interactions with foreign actors change from the catch
up to the post catch up stage, which is reflected in patent indicators. We use three indicators.
Foreign Applications of Native Inventions (FANI) measure the extent to which technological
development in a country or region is driven by foreign actors. International Co-invention in
technological activities (COINV) measure international collaboration using patent applications
with inventors residing in different countries. Native Applications of Foreign Inventions (NAFI)
measure the extent to which a country is able to exploit technological knowledge from abroad.
It may be expected that countries behind technology frontier do not have the organizational
capabilities to exploit their own technological knowledge which is then exploited by foreign
actors (high share of FANI), they increasingly interact with foreign partners for technology
development (increasing COINV) but do not have the capabilities for exploiting foreign
knowledge by themselves (smaller share of NAFI). As they are technology upgrading it may be
expected that share of FANI declines, while shares of COINV and NAFI are increasing.
Based on these indicators and modes of interpretation our comparative analysis focuses on
EU27 and its subregions (EU-12, CEE and South EU) and on the BRICS countries. We identify the
following developments.
In terms of intensity of technology upgrading we observe different trends in the accumulation
of technological capabilities at the technology frontier and behind the technology frontier,
especially in what concerns CEE.
On the one hand, all parts of the EU28 have increased their technological capabilities pushing
the technology frontier. TN patenting in the EU display strong growth from early 1990s and
deceleration of this process after 2008. Within EU the data suggest a divergence on core (EU12)
and periphery (CEE and South EU) countries which has been especially present until 2008 when
patenting in the developed EU12 slows down.
The comparison of CEE with BRICS suggests that in pc terms CEE has the highest TN patents in
pc terms. However, CEE are well behind China in terms of TN per GDP or in technology intensity
51 Lee, K. (2013) Schumpeterian Analysis of Economic Catch-up Knowledge, Path-Creation, and the Middle-Income
Trap. Cambridge University Press.
37
of economy as measured by TN patents. Nonetheless, CEE is ahead of other BRICs. In terms of
technology intensity at the world frontier CEE has advanced but it is beset by structural issues
as reflected by big difference between lower technology intensity of its economy vs. higher
intensity of country.
On the other hand, when it comes to technology effort behind technology frontier as measured
by resident patents we observe a stagnant trend in EU28 and in its subregions. This may be
expected given continuous economic and institutional changes towards European research area
and effects of industrial networks in the EU. A strong decline of resident patents per GDP in CEE
is the effect of their increasing internationalization and substitution of domestic technology
effort by opening of their innovation systems.
In terms of structural change, there is a shift towards HTKI areas in both EU South and CEE
towards EU12 shares. This is reflection of the strong convergence of both the EU South and
CEE to the core which presumably means that the overall technological knowledge structure in
the EU is becoming strongly determined by the EU core. However, a decline in share of HTKI
areas at the EU12 level shows that technology path of EU is quite different from the US or East
Asia. CEE falls clearly within BRIC spectrum in terms of share of HTKI patents.
What concerns the diversification of patent portfolios, there is a clear trend towards
diversification in BRICS and CEE except for China in terms TN patents and for India in terms
resident patents. A diversification trends is feature of all CEECs despite their quite different
starting levels of concentration/diversification. As we would expect, diversification is much less
pronounced in resident patenting which largely reflect domestic and behind the frontier
technology effort.
With regard to technology upgrading and the interaction with global economy in terms of
technology sourcing and interaction with foreign actors, the results for CEE match our
expectations to some extent. Given our interpretation of FANI we would expect that successful
technology upgrading would be reflected in decreasing FANI. Data for CEE and for BRICS are in
line with this hypothesis. Within BRICS context CEE FANI rates seem quite stagnant reflecting
possibly very weak endogenous organisational capabilities i.e a low share of domestic large
firms in technology activities. Another strong feature of the CEE is a high share of coinventions.
In all three sub-regions of the EU around 40% of all TN patents applications involve at least one
foreign and one domestic inventor (COINV). Levels of technology co-generation are lower in
BRICS than in the CEE and the rest of the EU which may be expected.
Interestingly, NAFI Rates for EU regions show that technology sourcing abroad has initially
declined in CEE and has remained stagnant and at comparatively very low level since mid-1990s
while it has increased significantly at EU12 and South EU. NAFI, which in our context denote
capacities for technology sourcing abroad, have been stagnant in BRICs which may seems
surprising given the newly emerging literature and evidence on emerging markets MNEs. It
seems that despite individual high profile cases of BRICS MNEs sourcing technology abroad,
38
these cases do not yet represent trend or technology sourcing is not their key strategic
orientation.
To identify specific technology upgrading paths for the different regions and countries we
develop technological upgrading profiles involving al indicators. These upgrading profiles have
been used for the comparative analysis in 2011 and 1995.
In 2011 CEECs were quite homogenous in their upgrading profiles which reflects their
technological levels and relative distance to the EU-12. A typical CEE economy in 2011 is well
behind EU12 in terms of frontier technology intensity, behind frontier technology intensity,
share of high tech patents and technology sourcing abroad (NAFI). Moreover, its organizational
capabilities are often less advanced as reflected in high share of FANI. The CEE profile is much
less coherent in terms of technology diversification/specialization and share of joint inventions.
However, differences among CEECs are not significant in the sense that we can talk of distinct
national technology profiles. Still there are some notable national features. Poland, Romania
and Slovenia have above average domestic technological intensity which reflects partly their
sizes (Romania and Poland) and specific model of innovation system reliant on domestic R&D
intensive firms (Slovenia). Latvia and Lithuania are specific in terms of high share of HTKI
patents.
Technology upgrading profiles of BRICs and CEECs for 1995 are more diverse than for 2011
reflecting divergences and convergences among these catching-up economies. CEE had the
least diversified technological knowledge portfolio with comparatively high frontier
technological intensity. By 2011 CEE has lost its initial high ranking in terms of frontier
technological intensity, has significantly diversified its technological knowledge, increased co
invention rate but also became the region with the highest FANI rate which reflects weak
organisational capabilities to commercialize its own inventions. On positive side, its technology
profile has diversified as should be expected when countries are transiting from middle towards
high income status. Also, its capacity for sourcing technology abroad has also somewhat
improved. However, invention process in CEE has become much less intensive but it is now
taking place in cooperation with foreign partners (COINV) who have organisational capabilities
to commercialize local inventions (NAFI).
These changes in the CEE technology upgrading profiles contrast well with BRIC countries.
Changes in profile of Russian technology upgrading have been similar but also much more
dramatic when compared to the CEE. Its decline of frontier and domestic technological
intensity has been much sharper than in the CEE. This loss of technology intensity of CEE has
been compensated by stronger interaction with global economy through high coinvention rate
which has not been the case in Russia.
China’s profile of technology upgrading shows very strong increase in both domestic as well as
in frontier technological intensity at the same share of high tech patents. On the other hand,
technological upgrading was not followed by its increased technological openness. In this
39
respect, a Chinese model of technology upgrading is quite different from the CEE which had to
compensate its decreasing technological intensity by more technological openness.
India has very low technological intensity which despite its high economic growth in this period
has further shrank questioning whether its further growth can rely on technology or on other
production factors. Similar to CEE, India has to compensate much less dramatic loss of
technology intensity by increase knowledge cogeneration (COINV).
Changes in Brazilian technology upgrading profile have been much less intensive when
compared to China, Russia, and CEE. Relatively small decreases in technology intensity and in
share of high tech patents have also resulted like in CEE and India to increases in knowledge
cogeneration at relatively similar NAFI and FANI rates. South Africa followed similar pattern as
its domestic and especially frontier technology intensity have declined as well as share of high
tech patents. As in Brazil, CEE and India this has led to increases in co-invention rates and with
slight changes in NAFI and FANI rates at unchanged degree of technology diversification.
Finally, considering all EU28 economies plus BRICS we rank the countries according to each
indicator in the years 1995 and 2011. By adding ranks we calculate one rank for each of the
three dimensions. The goal is to identify relative changes in the rankings for each of the
countries between 1995 and 2011 in each dimension of technology upgrading.
In terms of technology intensity, CEE (with exception of Slovenia) are firmly among the low
performers (holing positions in second half of the ranking) together with South EU. This is
expected given that drivers of their growth are not related to technology but to production
capability. In this dimension BRICs are not a homogenous entity. Their positions in terms of
technology intensity differ widely signaling very different opportunities for growth based on
technology. China has increased by far the most due to remarkable increase of both TN and
resident patents.
In what concerns structural change, economies that are weak in terms of technology (patent)
intensity show large changes in their benchmark position in terms of structural change. These
results are in line with our assumption of structural change underlying technology upgrading of
middle income economies towards high income. China and Estonia are again the biggest
improvers in terms of ranking which is a quite important additional evidence of their
technology upgrading given that China is third ranked and Estonia 17th in terms of technology
(patent) intensity.
The biggest changes in terms of increased openness in patenting activities took place in
Lithuania, Estonia, Luxemburg and Ireland while the biggest relative ‘withdrawals’ took place in
Italy, Portugal, Slovenia, Czech R and Spain. Our data shows that ‘globalizaton of technology’ is
not universal process but evolves very unevenly reflecting very much country specific
interactions between national technology systems and external environment. Interestingly,
four out of five BRICS (with exception of South Africa) have not further opened in terms of
knowledge exchange in terms of patenting. This suggests that despite foreign presence in R&D
40
in these economies, in particular China and India, this by itself has not led to relatively higher
openness of their technology systems.
In overall, our analysis puts the upgrading paths of the EU28 and CEE in the context of BRICS
and shows strong and weak features of the CEE technology upgrading. Our research shows
clearly that paths of technology upgrading are very country specific though differences among
CEECs are relatively much less present than when compared to BRICS.
CEECs are positioned relatively well in terms of technology intensity at the world technology
frontier though their economic growth is not triggered by these type of technology activities.
Our data indicate lower technology intensity of the CEE as economies (TN/GDP) vs. their
relatively higher technology intensity as countries (TNCpc). This is an indication of mismatches
in innovation systems of the CEECs especially regarding the relationships between technology
activities in business and public sectors. These mismatches will need to be addressed so that
the technology activities outside BES could be made more economically relevant.
Nonetheless, CEE technology upgrading as depicted by patents is within the BRIC pattern (with
exception of China which in terms of technology upgrading has de facto delinked from BRICS).
In the BRIC context, the CEE characterize very open innovation system with a high share of
coinventions and high FANI rates but also weak organizational capabilities to commercialize its
own inventions.
In terms of relative changes in technology upgrading CEE are firmly in the lower half of the
EU28 list together with four out of five BRICS (except China and Slovenia). The only really big
relative improver in terms of technology intensity is Estonia while other countries have
recorded much less significant relative changes. Diversification of their technology profiles as
proxy for technology upgrading of middle income economies is also well behind Chinese
changes again with the exception of Estonia. The interaction of CEE with the global economy in
terms of knowledge exchange interaction with global economy in overall is also not very strong
again with exception of Lithuania and Estonia.
In overall, CEE region shows good relative position in relation to BRICs but degree of changes in
technology upgrading between 1995-2011 falls within BRIC (except China) spectrum. The
biggest difference compared to BRICS is much higher openness of CEE in terms of patent
generation and weak control of patenting process. We interpret this as reflection of weak
organizational capabilities of the CEE larger local firms. A specific position of the CEE as part of
the EU has huge implications on how technology upgrading will evolve. Also, given their size,
the policy approaches to technology upgrading in the CEE are and will continue to be quite
different when compared to BRICs. However, the challenge to couple domestic with foreign
technology efforts is much more pronounced in this region than elsewhere.
Finally, our analysis shows that technology upgrading is multidimensional construct and that
aiming for aggregate composite indicator may actually mask the key issues which arise from
different stages of technology upgrading in which countries find themselves and from their
41
specific paths of technology upgrading. CEE grew during 1990s/2008 based on production, not
technological capability. Their future growth will increasingly depend on building technological
capabilities at world frontier level. Our analysis shows that the basis for such growth exists only
to a limited extent and that speed of upgrading towards world frontier activities is well beyond
required for catching up. Equally, our analysis shows that solutions for improved technology
upgrading will need to be found with their existing innovation model of small open economies
integrated into the EU.
Literature
Archibugi, Daniele and Coco, Alberto (2004)A New Indicator of Technological Capabilities for
Developed and Developing Countries (ArCo), World Development Vol. 32, No. 4, pp. 629–654,
2004
Archibugi, Daniele and Coco, Alberto (2005) Measuring technological capabilities at the country
level: A survey and a menu for choice, Research Policy 34 (2005) 175–194
Archibugi Daniele, Mario Denni and Andrea Filippetti (2009) The technological capabilities of
nations: The state of the art of synthetic indicators Technological Forecasting & Social Change
76: 917–931
Daron, Acemoglu and James Robinson (2008) The Role of Institutions in Growth and
Development, Commission on Growth and Development, Working Paper No. 10, The World
Bank
Dominguez Lacasa, Iciar; Giebler, Alexander (2014) Technological Activities in CEE Countries: A
Patent Analysis for the Period 1980-2009, IWH Diskussionspapiere 2/2014, S. 1-36. Available at
http://www.iwh-halle.de/d/publik/disc/2-14.pdf (last accessed January 2015)
Guellec, D. and van Pottelsberghe de la Potterie, B. 2001. The internationalisation of technology
analysed with patent data. Research Policy, 30, 1253-1266.
Guellec, D. and van Pottelsberghe de la Potterie, B. 2010, "Measuring the internationalisation
of the generation of knowledge. An approach based on patent data.," In Handbook of
quantitative science and technology research. The use of publication and patent statistics in
studies on S&T systems, H. F. Moed, W. Glänzel, & U. Schmoch, eds., Dordrecht: Kluwer
Academic Publishers, pp. 645-662.
Ernst, Dieter (2008) Asia’s “upgrading through innovation” 31 strategies and global innovation
networks: an extension of Sanjaya Lall’s research agenda, Transnational Corporations, Volume
17, Number 3, December 2008
de Rassenfosse, G., Dernis, H., Guellec, D., Picci, L., and van Pottelsberghe de la Potterie, B.
2013, The worldwide count of priority patents: A new indicator of inventive activity. Research
Policy, 42, (3) 720-737.
42
Frietsch, R. and Jung, T. 2009, Transnational Patents - Structures, Trends and Recent
Developments, Expertenkommission Forschung und Innovation, Berlin, Studien zum
Deutschen Innovationssystem. 7 - 2009. Available at http://www.e-
fi.de/fileadmin/Studien/StuDIS2009/7_2009_Patentreport_ISI.pdf
Glaeser, Edward L; R. LaPorta F. Lopes-de-Silanes and A. Shleifer (2004). "Do Institutions Cause
Growth?". Journal of Economic Growth, 9, pp. 271-303.
Koopmans, T.C. (1947) Measurement without theory, Rev. Econ. Stat. 29 (3) 161–172.
Krueger, Jens J. (2008) Productivity and structural change: a review of the literature, Journal of
Economic Surveys (2008) Vol. 22, No. 2, pp. 330–363
Lee, K. (2013) Schumpeterian Analysis of Economic Catch-up Knowledge, Path-Creation, and the
Middle-Income Trap. Cambridge University Press.UNCTAD (2005) World Investment Report, UN,
Geneva
W. Nasierowski1 and F.J. Arcelus (2013) On Perceptions of Technical Efficiency of the Basis of
the Innovation Union Scoreboard, mimeo
OECD (2012) OECD Review of Innovation Policy: Slovenia, OECD. Paris
OECD (2004) Understanding Growth, Palgrave Macmillan, London.
Peneder, Michael (2003), Industrial structure and aggregate growth, Structural Change and
Economic Dynamics, 14 (2003): 427-448
Radosevic, S. and Yoruk, E. (2014) Are there global shifts in world science base? Analysis of
catching up and falling behind of world regions, Scientometrics, June, DOI 10.1007/s11192-014-
1344-1
Radosevic, S. (1999) International technology transfer and catch-up in economic development,
Edward Elgar, Cheltenham.
Schmoch, U. (2008), “Concept of a Technology Classification for Country Comparisons, Final
Report to the World Intellectual Property Organisation (WIPO)”, revised in January 2013,
www.wipo.int/export/sites/www/ipstats/en/statistics/patents/pdf/
wipo_ipc_technology.pdf
Teece, David J. 1986, Profiting from technological innovation: Implications for integration,
collaboration, licensing and public policy. Research Policy 15 (6): 285-305.
Tunzelmann, N. von, Günther, Jutta, Wilde, Katja and Jindra, Björn: Interactive Dynamic
Capabilities and Regenerating the East German Innovation System, in: Contributions to Political
Economy, Vol. 29 (1), 2010, pp. 87-110
