ArticlePDF Available

Abstract

This paper examines the role that trade plays in economic development through the channel of technology transfer, approximated by total factor productivity. Three strains of factors influence the process of technology transfer; direct effort that is taken to transfer technologies, the capacity to adopt technologies, and differences in the underlying conditions between donor- and receiving countries. In this context, trade in (capital) goods allows technology import and improved input decisions. Second, trade opens export markets, allowing learning-by-doing. Third and most importantly, trade increases the set of accessible technologies, increasing the scope for imitation. The theoretical insights are compared to the empirical literature that deals with trade and technology transfer. Not surprisingly, it turns out that openness and human capital have a positive influence on the transfer of technology. Yet methodological problems with the data weaken the practical significance of the results, especially as the precise and fundamental mechanism of spillovers and the factors that condition the degree of technology transfer are not profoundly illuminated. These underlying processes have to be better understood in order to be able to give valuable policy recommendations that will go beyond the general advice of increasing openness and human capital formation.
This paper can be downloaded without charge at:
The Fondazione Eni Enrico Mattei Note di Lavoro Series Index:
http://www.feem.it/Feem/Pub/Publications/WPapers/default.htm
Social Science Research Network Electronic Paper Collection:
http://ssrn.com/abstract=657324
The opinions expressed in this paper do not necessarily reflect the position of
Fondazione Eni Enrico Mattei
Corso Magenta, 63, 20123 Milano (I), web site: www.feem.it, e-mail: working.papers@feem.it
Technology Transfer
Through Trade
Mombert Hoppe
NOTA DI LAVORO 19.2005
JANUARY 2005
KTHC - Knowledge, Technology, Human Capital
Mombert Hoppe, DG Development, European Commission
Technology Transfer Through Trade
Technology Transfer as an Additional Benefit from Trade – A Theoretical and
Empirical Assessment
Summary
This paper examines the role that trade plays in economic development through the
channel of technology transfer, approximated by total factor productivity. Three strains
of factors influence the process of technology transfer; direct effort that is taken to
transfer technologies, the capacity to adopt technologies, and differences in the
underlying conditions between donor- and receiving countries. In this context, trade in
(capital) goods allows technology import and improved input decisions. Second, trade
opens export markets, allowing learning-by-doing. Third and most importantly, trade
increases the set of accessible technologies, increasing the scope for imitation. The
theoretical insights are compared to the empirical literature that deals with trade and
technology transfer. Not surprisingly, it turns out that openness and human capital have
a positive influence on the transfer of technology. Yet methodological problems with
the data weaken the practical significance of the results, especially as the precise and
fundamental mechanism of spillovers and the factors that condition the degree of
technology transfer are not profoundly illuminated. These underlying processes have to
be better understood in order to be able to give valuable policy recommendations that
will go beyond the general advice of increasing openness and human capital formation.
Keywords: Technology transfer, Trade, Economic growth, Total factor productivity
JEL Classification: F10, F43, O40
The text represents the author’s personal views and is not to be seen as an official
position of the Institution.
Address for correspondence
Mombert Hoppe
DG Development
European Commission
Brussels
Belgium
Phone: +32 25380671
E-mail: Mombert.Hoppe@web.de
Technology Transfer Through Trade
1 Introduction
This paper defines technology transfer as technology diffusion between economies. While
Vernon (1966) and Krugman (1979) simply assumed technology transfer in their articles,
this paper focuses on the factors that influence it. Taking the importance of technology for
economic growth into consideration,1 technology transfer is of utmost importance for
output growth and the catching-up process of developing countries, especially as nearly all
R&D activity takes place in industrialised countries.
Technology transfer refers to the arrival or the transfer of a certain technology to a country,
where it has not been used before.2 In most cases, this process will be intertwined with a
process of adaptation due to different demands on the produced good and/or the production
environment, such as input prices and existing ways of problem-solving. Also the
utilisation of a certain technology in a similar context for the production of another good is
regarded as a transfer of technology with adoptive action. Together with subsequent
national diffusion and wider utilisation of this technology, technology transfer works in
increasing a country’s Total Factor Productivity (TFP). It will be difficult to distinguish the
two mechanisms within the data, as their effects are identical. Technology transfer in this
context describes both transfer to countries that did not use that technology before and
subsequent diffusion within that country, therefore, the terms technology transfer and
technology diffusion will be used as interchangeable terms in the following. This paper
will commence with defining some important terms before turning to the factors that
determine technology transfer. Next, the theoretical mechanisms of technology transfer
will be described. Subsequently, the results of empirical studies on technology transfer will
be summarised while the concluding section will compare the theoretical to the empirical
results and evaluate them.
2 Some Central Terms and Concepts
Technology and techniques
Technology is created by controlled inventive action following profit incentives and is
used as an input into production. It is in general a public good but the degree of complexity
and practicalities of its use as well as patent rights increase its degree of excludability
largely. First, technology is never fully expressed by the descriptions about the needed
-1-
Technology Transfer Through Trade
material inputs, as Pietrobelli (2000) argues. Technology contains a tacit element,
knowledge that is not readily describable or codifiable such as the way how to use a certain
equipment efficiently. These tacit elements can not easily be transferred but are of crucial
importance to the proper functioning of the technology and the choice of the used
techniques. With regard to patent rights, Eaton and Kortum (1999), argue that protection
takes mainly place in the inventor’s home country and to a much smaller degree in foreign
countries. Under these circumstances, technology transfer of good designs can take place
legally as long as the transferred technology is only used in the imitating country. As
markets in developing countries are generally larger for low-tech goods, this might lead to
more imitation of less sophisticated technologies.
Every technology can be operated with a wide range of techniques. While technology
refers to the general idea, machinery or blueprint, techniques refer to the way this
technology is actually used. The choice of the appropriate techniques can depend on factor
prices and differing needs of output. Even though a technology might be existing in a
country, it is not necessarily the case that users of this technology are aware of the
technique that is appropriate in their special position due to the tacit knowledge that forms
part of the technology. As a result, technology might not be used effectively, as Clark and
Feenstra (2001) point out. The process of acquiring the appropriate techniques forms the
second important step in technology transfer after the basic technology has been
transferred. Technology and human capital both increase the effective use of capital and
labour inputs. To the degree that increases in human capital are not observable or
measurable, such as learning-by-doing and experience, its effects will most likely be
picked up by TFP. As a result, this element of human capital is analysed as part of
technology transfer, an idea that is also implicitly contained in Bernanke and Gürkaynak
(2001) and that is in line with the argument of Lloyd-Ellis and Roberts (2002) who
emphasise the interactive working of technology and skills.
Inventing activity does not only take place in research labs or as a controlled and planned
action (primary inventions described by Young (1991)). A second set of inventions exists,
being equally important in the determination of TFP. These so-called secondary inventions
can be described in various ways and they include the concept of learning-by doing
(Arrow, 1962), dynamic learning (Radosevic, 1999) or learning-to-learn (Connolly, 1997).
Secondary inventions relate to primary inventions as do techniques to technologies. While
-2-
Technology Transfer Through Trade
the primary invention describes the conceptual technology, secondary inventions refine
this technology and improve its efficiency. Secondary inventions are based on experience
and are a by-product of final goods production. Their effects on TFP are nevertheless
substantial and we will assess their importance in technology transfer. There exist three
kinds of inventive activity. One is of quality-improving nature, the second deals with the
invention of new goods and the third form is of cost-reducing nature.
Human capital
Human capital is used as an input into production but also fulfils a crucial task in the
creation and adoption of technologies. The stock of human capital is influenced by
primary, secondary and tertiary education, vocational education, on-the-job-training and by
work-experience but also social capabilities and health aspects of the workers can
influence the efficiency of the workers. In a recent attempt to better and more accurately
quantify human capital stocks in different countries, Barro and Lee (2000) take the
completion of “education-levels” as well as average schooling years as the central
indicators of human capital. They also correct for quality while leaving out other factors
such as on-the-job-training and work-experience, which increase human capital, as for
example Lucas (1988) argues.
Measurement of absorption and quality of education (see Lloyd-Ellis and Roberts, 2002) is
difficult and human capital estimation using wage as done by Gollop and Jorgenson (1980)
and Mulligan and Sala-I-Martin (1995) (cited in: Barro and Lee (2000:16-17)) is difficult
and can conceptually run into problems (see e.g. Muysken, Rieder and Hoppe (2002)).
Conceptually, it can be said that human capital includes all knowledge that is embodied in
human beings and that is relevant for production and technology creation. In this light, the
paper by Barro and Lee (2000) leaves the reader with a greatly improved measure of
human capital. Work-experience and on-the-job-training, however, are hard to include and
it will therefore be likely that these improvements of human capacity will be rather accrued
to TFP when output growth is analysed. This has to be taken into account, however, when
analysing the effect of technology transfer on TFP growth.
Models of knowledge creation acknowledge the importance of human capital, either
directly (Romer (1990) or through intermediate inputs (Rivera-Batiz and Romer (1991)).
Aghion and Howitt (1998) point towards the fact that more sophisticated intermediate
inputs into knowledge production increases research efficiency. Eaton and Kortum (2001a)
indicate that it is the amount of researchers and the research efficiency of a country that is
-3-
Technology Transfer Through Trade
of crucial importance for knowledge production. The role of human capital in technology
adoption will be dealt with in the section dealing with “technological capabilities”.
Total Factor Productivity
The effects of technology are usually captured in the applied literature by total factor
productivity. This measure captures the effectiveness of all used inputs and comprises
technology and other unobserved factors such as institutional structures. Technology
increases the output that can be generated with a given set of inputs. The empirical
literature usually analyses TFP as measure for the level of technology within a country, its
distribution varies widely between industrialised and developing countries.
It is difficult to make general statements about the development of TFP-levels in
developing countries. Calculating them as a residual as in the growth accounting approach,
Hall and Jones (1999) report productivity levels for 127 countries in relation to the
productivity of the United States using data from 1988. They find that the average
productivity of these countries equals 51.6 percent of the US-value with a reported
standard deviation of 32.5. The values range from 120.7 percent for Italy and 112.6 percent
for France to values as low as 10.6 percent for China and 16 percent for Zaire. These
numbers show the huge disparities of TFP between countries.
Coe, Helpman and Hoffmaister (1997) report the development of TFP levels for 77,
showing that large differences in development exist. While ten countries experienced an
increase of more than 50 percent of their TFP from 1971-1990, another set of twelve
countries saw their TFP fall by 20 percent and more. In total, TFP decreased for 37 out of
the 77 countries during this time period. Regional differences in TFP development are
clearly visible. While the Middle Eastern and European countries experienced an average
productivity growth of 46 percent during these 19 years, the countries in the Western
Hemisphere actually lost 5 percent of their productivity on average. Without surprise, also
the dynamic East Asian economies experienced strong TFP growth of 58 percent over 19
years, representing an average productivity growth rate of 2.4 percent, a figure roughly in
line with the analysis of Young (1995). Klenow and Rodríguez-Clare (1997) find similar
results.
Measurement
Technology is difficult to measure as the process of technology production is not linear.
Inputs do not directly translate into technology and not all outputs are observable
-4-
Technology Transfer Through Trade
(secondary inventions, production processes, non patenting), constraining both methods of
assessing the stock and creation of technology. Moreover, the rate of knowledge decay is
difficult to assess and unobserved factors might blur the picture. Usually, a growth-
accounting approach is used when assessing national technology stocks. In this case, the
productivity effect of changes in the input-composition as well as the effects of all non-
described inputs or effects of external changes will accrue to TFP. While producing some
comparative explanatory potential, not only changes in the underlying productivity are
reflected. There exists no generally satisfactory measure.3 Changes in the exchange rate,
oil-prices, positive or negative output-shocks, the rates of capital- or labour utilisation all
can have an effect on the TFP-measure without reflecting any underlying change in real
productivity in case these changes are not taken care of. Especially for developing
countries the stock of human capital is likely to be overestimated, when enrolment rates are
used as Islam (1995:1153) points out. Lastly, the labour market structure of many
developing countries is characterised by a large informal sector (Allen and Thomas, 2000),
leaving the TFP-measure represent the productivity of the formal sectors. Under these
circumstances, the precise measurement of output and inputs becomes more difficult and
measurement errors in TFP-values are to be expected such that they do not necessarily
reflect the precise underlying theoretical value.
3 Theoretical aspects of technology transfer
The incentives for technology transfer lie, in parallel to technology creation, in competitive
rewards while lacking the uncertainty aspect of technology creation. An adopter knows
before imitating a technology that it is viable, while this is never clear a priori in the
process of technology creation. Moreover, the costs are generally much lower (see e.g.
Grossman and Helpman (1991a)). As long as the marginal costs of adapting another
technological innovation are smaller than the expected marginal (monopoly) profits,
inward technology transfer of this particular innovation will be worthwhile and firms will
try to gain access and implement foreign technologies. The degree of product market
integration and the size of transport costs influence the amount of technology transfer.4
Active efforts and inputs are needed for technology to be transferred.5 Being closely
related to technology creation (e.g. Teece (1976), cited in Wang (1989)), the same factors
work also in technology transfer. R&D spending can be seen as an investment in the
-5-
Technology Transfer Through Trade
special form of human capacity to adopt technological knowledge and the amount of
researchers increases the set of possible ideas, a fact that Eaton and Kortum (2001a) point
towards in the context of knowledge production. As in the model of Wang (1989), R&D
can interact with the size of the technology gap in increasing the amount of transferred
technology. The effectiveness is also influenced by other important factors.
Technology’s property of being largely non-excludable makes it prone to diffusion. When
a design is not very different from an existing one, copying might not need many resources
for being successful. This sort of knowledge diffusion stands central to the paper. Evenson
and Westphal (1995) present a general form of a national diffusion model, based on
Mansfield (1961). Diffusion takes place without underlying explanation but with a certain
pace and it is precisely this lack of underlying reasoning that can be found back in most of
the diffusion literature. While the gros of the literature takes spillovers to be automatic,
Grossman and Helpman (1990) point out that the mechanisms by which these spillovers
take place have been generally ignored. A precise model of the phenomenon lacks, even
though some factors such as labour movements and vertical relationships might give some
explanation (see Teece (1976), or Saggi (2000)).
The capacity to adopt
The capacity to adopt depends on a set of factors and is often referred to by termes such as
“technological capabilities” (Wang, 1989; Lall, 1992) or “national absorptive capacity”
(Movery and Oxley, 1995). It is normally assumed to be influenced by certain form of
human capital, and experience with imitation. Human capital and particularly tertiary
education is central in technology transfer. Trained workers largely influence the adoptive
capacity of firms, reducing the costs of technological imitation or adoption. Gruber and
Marquis (1969:268) also point out the importance of the kind of highly trained human
capital and in particular the workers’ experience and motivation.
In his influential paper, Sanjaya Lall (1992) describes the technological capabilities at the
firm level as composed of three factors. Investment capabilities, production capabilities
and linkage capabilities. He describes investment capabilities as the skills that determine
the capital costs and allow the appropriate selection of projects, a sort of experience in
investing. Production capabilities then describe the skills such as quality control or
knowledge on how to adapt machinery that function in an efficient utilisation of the
technology. The last set of capabilities as described by Lall includes the capabilities to
-6-
Technology Transfer Through Trade
transfer skills and technology from other actors in the economy. While these factors
explain the degree of technology transfer by capacity, determination of these capacities is
not clarified. Lall’s argumentation should therefore be rather seen as supporting the
importance of adoptive capacity in the process of technology transfer. All of these factors,
however, are rather difficult to measure and to integrate into one coherent measure of
human capital or adoptive capacity, as was already indicated. Human capital, in particular
secondary and tertiary education, directly influences the adoptive capacity of workers,
reducing the costs of adoption. This “adoptive” element of human capital is also influenced
by the amount of experience with technology adoption.
In their analysis of the functioning of human capital in the growth process, Benhabib and
Spiegel (1994) point towards the indirect effects of human capital on growth. In their eyes,
the major contribution of human capital lies precisely in its strength to make workers better
at “creating, implementing and adopting new technologies”. They draw on the theoretical
model by Nelson and Phelps (1966) who formalised the effect of human capital on
technology transfer. Human capital, however, forms only one factor in this model,
interacting with the technology gap.
While R&D spending directly affects the amount of technology transferred, it also works
in increasing the absorptive capacity of a firm. Wang (1989:78) points out that firms use
R&D not only to create new inventions themselves but also to increase their capacity to
recognise, to understand and to use technologies developed abroad. This is also contained
in Wang’s model of technological capability where the change in technological capability,
is determined by the resources that are devoted to transfer activity. While Wang sees
investment as directly influencing the technological capability of a firm, Lall (1992) argues
that national adoptive capacity is directly influenced by both the amount of research
experience and the amount of technological success. Griffith et al. (2000) find empirical
support for these factors, using industry-level data of 12 OECD countries.
As Pietrobelli (1998) and Lall (1992) argue, the national sum of technological capabilities
exceeds the sum of firm-level capabilities due to interlinkages and externalities that arise
from the accumulation of these technological capabilities such as scale and synergy effects.
As firms most likely do not take this positive externality into account, a state can be active
in investing in human capital formation, increasing productivity through the direct effect of
human capital on output and indirectly via lower technology transfer costs and the
resulting increased technology transfer.
-7-
Technology Transfer Through Trade
Differences in the underlying conditions of countries
Countries that are “leaders” and “followers” differ in their level of technological
development, their engineering history and the kinds of technology that are appropriate to
the existing factor endowments. Of importance is the technology gap between the
technological leader and the country that is trying to adopt technologies. On the one hand,
a larger gap between the leader and the follower gives rise to more technologies that can be
transferred (e.g. Wang (1989), Findley (1978), or Griffith et al. (2000)). However, these
technologies must be accessible by the country that wants to imitate them. Trade acts in
making formerly unknown designs available to actors in other countries. Due to the
embodied nature of technology, this leads to the accessibility of more technologies to
potential imitating firms.
On the other hand, the larger the gap, the more difficult the leading technologies are to be
understood as for example Blomström and Sjöholm (1998) argue. If in this case adoption
of older technologies takes place to a large degree as these technologies are closer to the
level of technological development of the adopting nation, it depends on the speed of
invention and adaptation whether the technology gap will actually in- or decrease. The two
opposing views have some truth to it and it is difficult to decide which factor dominates. It
can be said, though, that at least some technology transfer will take place. When the
technologies transferred become older and older, the imitating country will never catch up
in terms of technological development and due to the crucial role of technology for income
per capita neither in terms of income. Taking the large correlation between income per
capita and the technology level into account, Navaretti and Soloaga (2002) show that this
scenario actually represents what can be observed in reality, using a sample of developing
Central-Eastern European and Southern Mediterranean countries. The unit value of
machinery (and therefore the complexity) imported by the developing countries constantly
lags behind the unit value of machinery that the United States are importing, leading to a
consistent and increasing technology gap.
As e.g. Eaton and Kortum, (1995) argue, distance has a large influence on technology
transfer. Taking account of trade intensity should remove this factor as the increase of the
knowledge pool through trade should not depend on geographical distance. Still, the
compatibility of culture or machinery of the receiving and the transferring country can be
important. Technologies must be compatible and different approaches to the same problem
might exist in different cultural backgrounds. Therefore, imports from culturally distinct
-8-
Technology Transfer Through Trade
origins might increase the knowledge pool with technologies which are more difficult to
adopt within a certain technological background as compared to ideas that were developed
in countries belonging to the same cultural background. Long-during trade relations lead to
an equalisation of existing sets of technology. While geographic distance can explain some
of these intense trade relations also historical relations with certain nations can be seen as
determining factor. This idea gains some support when analysing the rates of return of
industrial countries’ R&D spending to developing nations (see Coe, Helpman and
Hoffmaister, 1995). While the United States as technological leader have a very important
influence on nearly all developing countries, the relative importance of other countries for
certain nations differs. Comparing the effect of Japanese R&D on different countries, it can
be seen that its effect is larger for China, Korea, Singapore, Hong Kong, and Kenya when
compared to other countries. While the first four countries lie in the proximity of Japan,
Kenya is a very important trade partner of Japan and receives large assistance from Japan.6
This indicates that Japanese R&D has a relatively larger effect on culturally and
historically related countries.
A similar pattern can be found, when analysing the former British Colonies. While British
R&D spending in general has a lower rate of return to African countries then does German
or French R&D spending, Zimbabwe, Uganda, and Kenya benefit clearly more from
British than from the former’ spending. India also benefits strongly from British R&D
spending. The same sort of colonial ties can still be observed in the return of French R&D
spending on Cameroon’s output. While cultural closeness does not fully explain the
observed pattern, it nevertheless helps in intuitively explaining large parts of it.
Complexity of technology and skill-levels
This section deals with the complexity of technology and the (in)compatibility of different
skill-levels with a certain technology that Acemoglu and Zilibotti point out in their (1998)
seminar paper. Here, they construct a model of two regions, one being rich in skills (the
North) with the other (the South) lagging behind and being poor in skills. Only the North is
investing in R&D and they can develop machineries to their needs. The authors assume
that the utilisation of skill-intensive machinery with unskilled labour leads to lower
productivity (see also Lall, 1992:168). This means that machinery might be so complex
that imitating firms cannot reap profits from using them as the labour input they have at
their disposition is lacking the appropriate skills. Therefore, less technology will be
transferred. Moreover, the technology that is transferred and operated with sub-optimal
-9-
Technology Transfer Through Trade
labour input will lead to a lower TFP increase than an identical transference with qualified
labour input would generate. Basu and Weil (1998) point to the importance of the
appropriateness (in terms of capital-labour ratios) of the existing technology. In line with
Acemoglu and Zilbotti (1998), this leaves most high-quality technology inappropriate for
utilisation in developing countries. Their idea is supported by the arguments of Evenson
and Westphal (1995), arguing that not only differences in factor endowment but also
differences in physical, economic and social conditions affect the value of technologies
and therefore its appropriateness. While this effect is not empirically quantified, the idea of
appropriateness should be kept in mind, when analysing technology transfer.
As Lall (1992) points out, firms have knowledge of the technologies they are actually
using. He argues that this knowledge decreases for similar technologies that other firms use
and decreases further for dissimilar alternatives. As a result, technological progress is
expected to appear in the surrounding of already existing technologies. When technological
improvements are seen as consisting of a spectrum of small improvements, a larger
complexity difference or an upgrading of larger scale become increasingly different.
Following Lall’s argumentation, these technologies will be less known to firms and will
therefore be more difficult to implement, when large changes are supposed to take place at
the same time. In investment theory, a very similar concept exists. Installation costs are
supposed to increase with the size of an investment but are expected to be transitory. Based
on the increased complexity of technology and the decreasing knowledge of technologies
that are farther away from the ones that are used, the same can be argued for technology
transfer. It should become clear that the technological distance between the actually used
and the potentially transferred technology has a positive impact on transfer costs and
technology transfer will therefore be likely to take place in small steps. It will equally be
more costly between than within sectors.
Competition is of large importance for multinational corporations as Blomström et al.
(1992), and Blomström and Wang (1992) argue. The underlying idea can also be applied to
the case of spillovers. The technological advantage of an affiliate translates into better
quality products and resulting higher mark-ups and profits. Similar to the argumentation in
the case of technology creation, the aggregate effect of increased competition on
technology transfer is difficult to assess. The role of ownership has not been answered
sufficiently clear to my knowledge (see also Blomström and Sjöholm (1998)).
-10-
Technology Transfer Through Trade
Summarising, the marginal revenue of technology adoption depends ambiguously on the
level of competition. The costs of adoption are negatively influenced by the existing and
accessible technology pool, the quality of the researchers, best approximated by the tertiary
education rate, the amount of cumulative past R&D spending, the appropriateness of the
accessible technologies, and a measure of cultural or engineering closeness, as described
above. The working of the different factors that influence marginal revenue and marginal
costs of technology transfer have been analysed and their effect on the speed of technology
transfer has been pointed out. Of these factors, the accessible set of technology will be
central in the subsequent analysis as this is the factor that is most strongly influenced
through trade.
National diffusion succeeds technology transfer
Diffusion within a recipient country follows its transfer International technology diffusion
[technology transfer] appears to be slower than national technology diffusion. This is due
to the differences in the underlying conceptions, the lower degree of communication and
the differences in the functioning of firms. The more widely used a technology becomes,
the more will this technology be adopted by different firms. The awareness of an existing
technology is greatly improved when it is geographically close not only with respect to its
productive output but also with respect to its productive processes. Personal contacts, news
and especially labour movements play a crucial role in the diffusion of technology once it
has reached a certain geographical location (Saggi, 2000). Once a technology is employed
in one firm of a less developed country, labour mobility between firms and also between
sectors can be central in further diffusion. Transfer might take place in an unchanged way
from one firm to another within the same economic sector, or might be adjusted to
different sectors, where adoption of the technology might take place. Larger (or more
advanced) firms that start acquiring foreign technologies will be able to pay skilled labour
a large wage premium in developing countries. It remains highly doubtful, whether small
or young firms will have the financial means to compete the qualified (and technologically
advanced) labour away from the large or adoptive enterprises. This would lead to a slower
triple-down effect within the country than it could be hoped. On the other hand, this
worker-rigidity will lead to higher profits and the forces pulling technology towards the
developing countries will be larger. An overall evaluation of the two opposing forces is not
possible with the data at hand.
-11-
Technology Transfer Through Trade
4 Trade and Technology Transfer
While many approaches point to static gains from trade, this paper points to other, dynamic
mechanism that works through trade. Through trade, product designs and product
characteristics spread to developing countries in the sense that they are widely available
and accessible by firms in these countries. The more contact points exist (the larger the
trade volume is), the larger is the assumed effect. This section will present the three core
mechanisms/variables that help transfer technology and increase TFP in turn. First, imports
of capital goods and the effects of an increased set of intermediate goods will be described,
as these are the most direct effects. Next, trade can lead to a dynamic effect in production
and to the learning of techniques, increasing TFP. Last and most important, trade increases
the amount of accessible technology and the knowledge stock. As a result, knowledge
(re)production in developing countries is likely to increase.
Direct effects
Import of capital goods
The most direct relation between trade and technology transfer remains in the direct
imports of machinery goods (a very obvious but often neglected form of technology
transfer via trade, see Eaton and Kortum (2001b) and the criticism by Navaretti and
Soloaga (2001)). Machines are imported and they have an immediate impact on
productivity through the technology that is embodied in them. The pure import of capital
goods does not necessarily lead to an appropriate use of the machinery, indicating that also
disembodied or tacit knowledge must be transferred.7
The effects of machinery import on TFP are difficult to separate from the effects of other
channels. Moreover, the “pure” productivity of imported machinery is hard to evaluate as
the technical and knowledge aspects only together result in the observed productivity
gains. The amount and quality of imported capital goods and the underlying efficiency of
the machinery is difficult to assess even when capital imports are disaggregated in import
data. Navaretti and Soloaga (2001) argue that more complex production goods will
increase productivity by a larger degree than older and less complex technologies. They
find for the European countries that there does exist a positive correlation between income
per-capita and the complexity of imported capital goods: more backward countries import
less advanced technologies.
-12-
Technology Transfer Through Trade
Quantity aspects of intermediates
International trade consists to the largest extent of trade in producer rather than consumer
goods (Ethier, 1982; Coe and Helpman, 1995). Consequently, the range of inputs that can
be used by producers in every country is increased through trade, a point that is used by
Keller (1999) in his model. The theoretical foundation for this lies in the models of Ethier
(1982) and Grossman and Helpman (Grossman and Helpman, 1991a). Here, the production
of the final good depends on the amount of intermediate inputs (goods and services) that
are used in the production process. The production function that Grossman and Helpman
(1991:47) present exhibits constant returns to scale in the production of the final good and
is based on a model by Dixit and Stiglitz (1977).
(4.1)
α
α
1
0
)(
=
n
djjxD , for , 10<
where is the amount of intermediate product j used and α represents a parameter that
allows to calculate the elasticity of substitution between the different intermediate inputs
which equals
)(jx
α
=1
1
ε. Differentiated inputs are the result of inventive activity and it is
assumed that all intermediate inputs are produced with the same constant-returns-to-scale
production function. As Ethier (1982) argues, all will be equal in equilibrium and the
resources that are used therefore equal . As represents total inputs, TFP can be
written as
)( jx
nxnx
(4.2) α
α
==
1
n
nx
D
TFP , with 0>
n
TFP for . 10<
In this model for a closed and small economy output therefore increases with the amount
of available and used intermediate goods. With the non-existence of obsolescence in this
model, and with profit-maximising agents, all available goods will also be used. This
property is explained by Ethier (1982) with an ever increasing division of labour into
separate production processes. When moving from this closed economy to an open
economy model, trade will increase the amount of available intermediates and will
therefore increase TFP.
Quality aspect of intermediates
Next to the issue of a wider variety of intermediate inputs into production, also the effects
of quality improvements through technology have been discussed (e.g. Grossman and
Helpman (1991a; 1991b)). Quality improvements are of large importance as they include
-13-
Technology Transfer Through Trade
the concept of obsolescence, meaning that new generations of products (and intermediate
inputs) replace older versions. This is also valid in a developing country context. Quality
ladder models (e.g. see Connolly 1997, p.6) state that inputs of higher quality that are
invented are more productive and therefore raise total output, while keeping input prices
constant (equally, production costs of intermediate goods might fall, lowering input prices
while keeping output constant, again raising TFP). A certain number of intermediate
products (J) exists whose quality is improved by invention or imitation (denoted by k).
Quality improvement are assumed to be of a constant factor q>1 making it q times more
effective than the older version. If the knowledge how to produce a certain intermediate
good is available, the good can be produced at marginal costs which are independent of the
quality level of this good due to perfect competition in the final goods market as Connolly
(1997) points out. Production of final goods can be described as follows (Connolly 1997,
citing Barro and Sala-i-Martin (1995))
(4.3)
=
=J
j
ik
k
iii j
ij xqLAY
1
1
)( αα
Here, the parameter describes the effectiveness of institutions in country . This
variable will also include the amount of human capital and probably the capital stock
which is not taken care of in this model. The factor q describes a quality adjusted
measure of inputs into production. Connolly argues that firms will use limit pricing to
capture the entire market and push older products out of the market and concludes that
assuming and as given, output solely depends on the quality measure of
intermediate products that are used in a country. For a developing country this means that
under an open trade-regime, it can use intermediate inputs of supposedly higher quality
that are imported from developed countries in their own production process. Hereby, the
quality and value of the final goods increases directly through the intermediate input of
higher quality.
A
i
i
j
ij
ik
kx
i
A L
A wider set of inputs to choose from
When producers have to chose from a discrete rather than a continuous set of inputs, their
input decisions most likely are less than optimal as some needed intermediate products or
intermediate inputs of a certain desired quality or quantity are not available. With trade,
both the quantity and the set of quality of intermediate inputs that producers can choose
from rises and access to them becomes easier. Trade therefore allows producers to improve
-14-
Technology Transfer Through Trade
their input decisions, thereby reducing the amount of inputs needed for the desired level of
output.
Dynamic gains from trade – learning-by-doing
While technology is transferred via direct imports and intermediate goods, techniques are
improved through better input decisions as indicated above and due to learning-by-doing.
This process of market integration also allows the production of goods that are not
demanded in the developing country’s market (e.g. because they are based on a General
Purpose Technology that is not widely used yet in that country).
The concept of learning-by-doing as formulated by Arrow (1962), aims at the efficiency
gains of repeated production, increasing TFP. Aghion and Howitt (1998) present a model
with internalised learning-by-doing. In their model, workers can use their human capital in
the R&D sector to perform basic research or they can use their human capital in
production. In this case, learning-by-doing will take place and secondary inventions can be
made. While R&D focuses on the creation of new products, secondary inventions only deal
with the improvement of already existing goods and processes. In correspondence to the
existing models of product variety and product quality, R&D relates to the former, while
learning-by-doing relates to the latter. The initial quality of intermediate goods depends on
the moment of time in which they were invented; that is they depend on the amount of
general knowledge that was available when they were invented. In the second stage,
learning-by-doing and the resulting secondary inventions increase the quality of these
goods. Both aspects therefore work together and both raise TFP. While the innovation of
new intermediate goods depends on the Poisson-arrival rate of each researcher, the quality-
evolution depends only on the rate of learning-by-doing within each firm and not on the
general stock of knowledge that increases with primary and secondary innovations. This
rate is determined by the amount of labour used in the production process and by a
parameter determining the productivity of learning by doing. Both parameters are
exogenously determined and as they largely determine the growth rate, the basic reasoning
for an increased growth rate still remains with the factors influencing these exogenous
factors. In both cases they should be related to education with the Poisson-arrival rate of
each researcher related to higher education (as well as experience in imitating), and the
productivity of learning by doing related to primary, secondary, and vocational education,
as well as work-experience. Aghion and Howitt argue that the stock of general knowledge
-15-
Technology Transfer Through Trade
depends on both sorts of innovation, basic and secondary ones. This way of interpreting
learning-by-doing will be used in the remainder of this paper.
Production of final output leads also to an externality through learning-by-doing.
Consequently, more production will lead to a larger increase in the knowledge stock,
increasing TFP.8 Market integration and trade increase the incentives to transfer also more
advanced technologies to developing countries, giving rise to learning possibilities.
Foreign Direct Investment (FDI), sourcing relationships and import/export competing
industries give rise to these learning possibilities.
Foreign Direct Investment (FDI) and Joint Ventures
In the context of learning-by-doing and an increasingly integrated world economy, FDI
and Joint Ventures are important for the transfer of technology to developing countries.
Often, re-imports based on low labour costs are an important reason to invest in a country
with low labour costs. Integrated markets can lead to an increase in production in less
developed countries increasing TFP directly through the used capital goods as well as
through the learning effects of workers. Moreover, the larger the responsibility and control
of the domestic firm, the better also the understanding is expected to be as involvement is
higher. A better understanding gives a higher incentive for workers to defect and to found
independent firms. Therefore, Joint Ventures, which leave more control and more
responsibility to the domestic firm will give more incentives for workers to learn and might
lead to a more rapid technology diffusion within the host country. Still, sourcing and
import- or export-competition give again stronger incentives to domestic firms to innovate
and to learn.
Sourcing and subcontracting
While world markets are ever more integrating, the production of goods becomes more and
more specialised and therefore dispersed. Multinational enterprises outsource the
production of intermediate products or parts of products to other markets, mainly due to
geographical reasons or cost-considerations. In contrast to international production such as
FDI or Joint Ventures, being characterised by a centralised system of ownership and
control, subcontracting or sourcing deals with the relation between independent firms.
International subcontracting can be distinguished from normal import/export relationships
through the existence of a stable contract. This contract establishes either a commitment to
buy a certain amount of specific goods over an agreed time-horizon, allowing the
-16-
Technology Transfer Through Trade
subcontractor to function as a second-source for supplying customers, or leaves the
subcontractor produce final goods which will be labelled as the principal’s goods. Also
other steps of the production process such as assembly, testing, or marketing can be
outsourced. The subcontractor will be given certain requirements, giving scope for
understanding and learning. In a second step, a development from an “economic”
subcontractor (based on cost-reduction considerations) to a “specialised” subcontractor can
take place through dynamic learning (Radosevic, 1999). The subcontractor will gain more
knowledge in the production of “its” good, and will gain a higher position on the “value
ladder” of goods. This, however, is not easily and especially not automatically
accomplished. The critical factors determining the successful dynamic learning still remain
unidentified. It appears, however, that they also depend on the stock of human capital,
absorptive capacity and profit incentives of the sub-contractor. The increase in sourcing
activity can be recognised by the increases in intra-industry trade between developing and
developed nations. Again, the precise transfer process remains unidentified.
In the case of FDI, competition between firms might lead to higher technology transfer
towards subsidiaries in developing countries, but the threat of spillovers might as well lead
to the transfer of older or less technology. When dealing with sourcing, inventive capacity
is needed to defend the technological advantage that might exist from entering the industry
first or to gain a qualitative advantage. Next to competition with foreign based or sourcing
firms, also competition with domestic firms that are extracting monopoly profits can exist.
While these positive dissemination effects can lead to virtuous circles, the opposing
process of trapping sourcing-firms in low value processes and the possibility of driving out
local competition due to high cost, technology and capacity advantages can form a vicious
circle (Radosevic, 1999). Again, this will greatly depend on firm-level capacities of
sourcing-firms and (potential) competitors.
Having become a ‘specialised’ supplier in a sourcing relationship, also the industrial
countries’ markets start to present profit opportunities to developing countries’ firms, as
their labour costs remain lower than in developed countries. Exporting to developed
countries, however, increases the competitive pressure on developing countries’ firms as
they must compete with all firms and not only with selected firms in their market. This
increases the set of inventions that a firm is exposed to and gives rise to further incentives
and possibilities for innovation. This paper will turn back shortly to this problem later
when discussing models of knowledge transfer.
-17-
Technology Transfer Through Trade
Access to technology
The knowledge pool or the set of accessible technologies
Having analysed the direct and dynamic effects of trade on TFP, the effect of an increased
knowledge pool remains to be analysed. This channel is of utmost importance when
analysing the role that trade plays in technology transfer. As was already stated, trade
increases the number of designs and production mechanisms that are accessible within a
certain country. While it could be argued that controlled imports of products and
technology can do the same trick, it has to be taken into account that a wider exposure to
technology gives rise to more possible and potential points of transfer. This point is
explained by Connolly (1997) who argues that importing firms are responsible for
distributing goods and therefore have more knowledge concerning them, reducing the costs
of adoption. The more goods are imported, the larger supposedly the number of firms for
which adoption costs are reduced, as long as imports are not monopolised. The remainder
of this section will present the theoretical approaches that support the assumption that more
trade leads through a wider accessibility of technology to more technology transfer. It will
start with the mechanism of reverse engineering before turning towards models of
technology diffusion. Subsequently, the empirical literature will be compared to this
section.
Reverse engineering and imitation
Reverse engineering aims at strapping products, understanding them and to rebuilt them
afterwards. This means that existing products, designs or methods are copied and adopted
towards local needs. For the imitating country this process functions as a kind of invention,
with the difference that this sort of imitative invention demands less labour than inventing
goods from a set of unknown possibilities, as Grossman and Helpman (1991a) argue. In the
case of imitation, this uncertainty does not exist, as a functioning product with the desired
characteristics is already existing and “only” has to be copied. Assuming that a technology
can be mastered by pure physical reverse engineering and without the participation in the
actual production process or direct communication between scientists of the producing and
the imitating firm, reverse engineering becomes possible through the exposure to goods.
The costs of adoption are influenced by the set of technologies that is accessible in a
country. Two opposing factors exist. On the one hand, the more technologies are available
in an economy but have not yet been adopted by the economy, the larger is the set that
firms can choose from when copying, increasing the possibilities for imitation. On the
-18-
Technology Transfer Through Trade
other hand, the large set of technologies might result from large differences in the
technology level between the importing and the exporting country. More complicated
technologies are more difficult to copy, as Connolly (2001) argues. Still, ceteris paribus,
more trade leads to a larger available variety of designs and goods to copy, increasing the
probability of imitation.
The larger the level of technology that is existent in an economy, the closer is that country
to the Technological Frontier Area (TFA). This implies that, generally, the gap between the
existing knowledge and the knowledge that has to be acquired through reverse engineering
is smaller. As a smaller technological gap between existing knowledge and desired
knowledge should make it easier for an engineer to understand and copy the technology at
hand, the stock of existing knowledge will have a positive effect on knowledge imitation.
In this context, reverse engineering functions as “knowledge production” for the
developing country as the transferred knowledge is new to the developing economy and
raises the parameter of the technology stock, , in the developing country. Yet, a higher
level of technological development reduces the scope of copying, as the imitating country
comes closer and closer to the TFA. The set of technologies to copy is reduced and the
total effect on imitative activity is likely to be negative. Besides, the higher the intensity
with which a country is exposed to a certain technology, the more actors have the
possibility to imitate the technology. Analysing the determinants of innovation and
imitation, Connolly (2001) argues that the probability of successful imitation is positively
influenced by the resources that are used in the imitative process and past imitating
experience, while the complexity of the good that is to be imitated has a negative effect on
the probability of imitation. This section will now turn towards its central point, the effects
of an enlarged and more widely accessible set of technologies on technology transfer. As
technology transfer functions as knowledge production for the imitating country, this paper
will turn to describing models of technology transfer. These models are often based on
models of knowledge production and are related to an increase in the set of accessible
technologies through trade, leading to technology transfer. Increased access to unknown
technologies increases the rate of technology growth in imitating countries as long as the
accessible technologies are not too complex to be understood.
A
Models of knowledge transfer
Grossman and Helpman (1990) argue, that not only domestic R&D spending but also
foreign R&D spending add to a “stock of knowledge capital” that reflects the local
-19-
Technology Transfer Through Trade
understanding of technology, engineering, and industrial know-how. Foreign R&D-
spending, so their argumentation, lets the local knowledge stock grow if and to the degree
that there exist contacts between the two countries. Trade, in this line of argumentation,
increases the number of contacts greatly and therefore increases the local stock of
knowledge capital and therefore the effectiveness of research or imitating activity. They do
not point out, however, in which ways this process works precisely. This section will
present different models of technology diffusion or will relate already presented models to
the effects of an increased knowledge stock through trade.
Nelson and Phelps (1966) present a model of technological implementation. They
formulate the change in the level of applied technology:
(4.4) )(
)()(
)( tA
tAtT
H
A
A
&
.
Nelson and Phelps use this model to describe the rate of implementation of technology into
practice. They try to express the importance of human capital for knowledge diffusion.
Applying the model to the concept of technology transfer, we have to focus on the
technology gap, which is the second determining factor in the rate of technology
implementation. The authors assume that the theoretical level of technology, T, grows
exogenously at a constant exponential rate. In the context of this paper, however, T for
a developing country depends on the amount of technologies that are accessible. This
understanding mirrors the concept of the theoretical level of technology in a country that is
developing its own technologies. The growth rate of the level of technology in practice in a
developing country can therefore be described with the model of Nelson and Phelps, taking
as the set of technologies that is accessible by most actors in the developing country.
Arguing that more imports increase the spread of a certain technology within a country,
more imports also increase the set of technologies that are widely accessible (depending on
the technologies that are used in the exporting country). By widening the technology gap,
this leads to a higher growth rate of technology in practice.
)(t
)(t
)(tT
The model of knowledge creation that is presented by Rivera-Batiz and Romer (1991) (see
also Romer (1990)) described knowledge production as depending on the human capital
stock and the already existing knowledge stock.
(4.5) HAA δ=
&
-20-
Technology Transfer Through Trade
This specification, however, relates to the knowledge production of one country. In the
framework at hand that analyses technology transfer from developed to developing
countries, the knowledge stock of the developing country can be expected to be smaller
than the knowledge stock of the developed country and it can be expected to be a subset of
the developed countries knowledge stock. While this might not be completely true due to
adaptations that are made in receiving countries, this assumption can be used as an
approximation. For the trade context, Rivera-Batiz and Romer assume that knowledge
stocks between countries do not overlap, an assumption that is not fulfilled in this paper’s
context. Therefore, to take the specific knowledge distribution into account, I propose to
adopt the model in the following way:
(4.6) ,
)(
21 HFBHAH AAHAHA δδ+=
&
where and represent the knowledge stock in the developing and industrialised
country, respectively, human capital used for domestic innovation, human capital
used in imitative research, and and δ two productivity parameters. All information
that is used in the developing country is also known in the developed country. As
, the measure must be introduced, representing the knowledge stock of
the developed country to which scientists of the developing country have access and which
does not form part of yet. Considering only the portion of foreign knowledge that
scientists have access to lies in line with a comment made by Romer’s (1990) presentation
of this model of knowledge creation. He points out that only the knowledge that scientists
have access to must be considered (p.83). In this context, it has to be stated that also
disembodied spillovers might take place, as Grossman and Helpman (1991a) argue. They
claim that the general knowledge stock is not only influenced by the number of
technologies that the developing country has already acquired but also by the number of
technologies that only the developed country’s firms can produce. In this case, some
information can nevertheless have spilled over to the developing country, through contacts,
blueprints or imported goods, increasing the factor in the adopted Rivera-Batiz-
Romer model, as this factor is determined exogenously.
H
A
F
A
F
A
A
H
F
A
H
A
B
H
1
δ2
H
A)(H
A
)(HF AA
Under this setting, trade constantly increases the set of accessible technologies of the
developing country because technologies that are newly invented also become accessible
by the imitating country. The effect is a rise in the stock variable , as the foreign
accessible knowledge increases while it is not immediately imitated and therefore does
not change. Knowledge creation is greater than under autarky as Rivera-Batiz and Romer
)(HF AA
H
A
-21-
Technology Transfer Through Trade
(1993) point also out in an addendum to their 1991 article, stating that trade policies which
increase the available stock of knowledge increase growth in the economy. The factors δ
and were included to allow that domestic and foreign knowledge influence domestic
knowledge production to differing degrees. As imitation is assumed to be easier than own
knowledge production, it should be expected that δ> . Summarising, it can be said that
rises due to two mechanisms. First, trade leads to a continuous increase in the
accessible knowledge, increasing the imitation of knowledge through the wider set of
available ideas. Second, this increased knowledge production increases the factor ,
leading to more own knowledge production given that human capital is used for own
knowledge production. At the same time, this increase in , ceteris paribus, leads to a
reduction in , the accessible knowledge stock which does not form part of yet,
reducing the scope for imitation.
1
H
A
2
δ
12 δ
H
A
H
A
)(HF AA
2
H
A
Following Connolly (2001), the concept of learning-to-learn could be included in this
specification. Learning-to-learn in knowledge production is a concept similar to learning-
by-doing in the production of goods. Experience in imitating technologies that is gained
through the process of reverse engineering will lead to a more thorough understanding of
the general technological concept and will increase the likeliness of both further reverse
engineering (imitation) as well as the probability of creating new technology in the future.
This insight is also described by Teece (1976, cited in Wang (1989)) who finds that
transfer costs fall for every subsequent application of an existing innovation. This means
that once a new strand of innovation has been transferred, new versions of this strand are
transferred at lower costs (i.e. imitation experience has increased and the imitation
efficiency would increase). I will not further follow the effects of introducing such a
fundamental change to the Rivera-Batiz and Romer model.
δ
In the lab equipment model of Rivera-Batiz and Romer (1991), knowledge only has an
influence on the production of knowledge through the use of more sophisticated
intermediate goods. Nevertheless, trade in goods has an effect on the production of
knowledge. Similar to the model of increasing varieties, also the knowledge production
function as presented by Rivera-Batiz and Romer depends on the set of inputs. As long as
trade brings such inputs into a developing country that are used in knowledge production,
trade will also in this model have a positive effect on knowledge production, as Aghion
and Howitt (1998:374) argue.
-22-
Technology Transfer Through Trade
This model, however, appears ill suited for developing countries. Their specification of
knowledge production might have some justification for countries that are similar to the
technological frontier area, but it appears less appropriate when analysing countries that
are distant to it and which use mainly reverse engineering in order to transfer technology.
This section shortly addresses the working of Manfield’s (1961) model, as presented by
Evenson and Westphal (1995), of knowledge diffusion in an international context. In his
model, the speed of diffusion depends on an external factor, b, describing the degree to
which the benefits and costs of adoption influence the diffusion rate.
(4.8) bt
ae
tp
+
=1
1
)(
Here, describes the fraction of firms that have adopted the new invention at time t
and the factor represents a constant. This model, however, depends on the underlying
costs and benefits of adoption, without explicitly analysing these two factors itself. In his
influential article, Teece (1977) analyses the factors that influenced transfer costs for
multinational enterprises. It can be said that the same factors apply also to the transfer
outside these enterprises even though their effects might be larger or smaller. The three
main factors he finds are i) the age of the technology, ii) the number of manufacturing
experience of the receiving firm, and iii) the number of firms that use a similar or identical
technology. Teece then further analyses the international aspect of the transfer costs and
comes to the conclusion that for machinery transfer the level of host country development
play a significant role in the determination of the transfer costs. In this context, trade can
help making a technology accessible or known and therefore initiate the subsequent
process of diffusion.
)(tp
a
Wang (1989) presents a model of firm-level knowledge stock growth. In this model, both
the effect of R&D spending and the effect of imitation play a role. He argues that a portion
of the foreign existing knowledge, , is available in a country that is participating in
international trade. This size of this portion,
z*A
z, is determined by the size of and a
factor that summarises other factors influencing knowledge transmission (e.g. the degree of
economic integration).
*A
t
z
9 The knowledge stock of a developing countries’ firm, , then
develops as follows:
(4.9) tttttt zgIIzz ),(
1τ++=
+,
-23-
Technology Transfer Through Trade
where represents the amount of R&D spending, the technology gap, τ a
function that represents the capability of absorption, depending on both and , and t a
time index. The absorptive capacity increases with the amount of own research and the
effect of own research increases with the size of the technology gap,
t
It
g),( tt gI
t
g
t
I
0
I>
t
τ and 0>
∂∂
∂∂
tt gI
τ.
When the gap vanishes, so does the possibility of imitation. In this formulation, however,
some additional thoughts have to be incorporated. First, the effect of R&D depends on the
existing knowledge stock and on the quality of the researchers as indicated above. Second,
the absorptive capacity should also depend on the quality of the researchers and on the
cumulative experience in imitating. In order to take these two points into account, it should
be considered to adopt the model to:
(4.9‘) )),(,,,(
0
1i
t
i
ttttttttttt zgIHgIzzHIzz
=
+++τδ ,
where H represents the firm’s human capital stock in period t, δ a parameter indicating
research efficiency of the particular firm, and i
t
i
tt zgI
=0
),(
τ cumulative experience of
imitation. This result, however, is valid for the individual firm and difficult to test for due
to data constraints. Assuming that all n firms are equal in the economy, total technology
stock growth equals and also equals TFP growth. znA &
&=
5 Empirics of Technology Transfer Through Trade
This section deals with the empirical models and findings at the macro-level, putting
emphasis on technology transfer to developing countries. Still, the data situation for
developed countries is much better and countries are much more similar in underlying
conditions, leaving less “noise” in the models. As a result, the empirical models chosen
touch both on technology transfer between industrialised countries and towards developing
countries. No models that test for technology transfer between developing countries or
from Newly Industrialised Countries (NICs) towards Developing Countries exist to my
knowledge. Results are much less clear for models dealing with developing countries,
pointing towards the fact that many underlying determinants of technology transfer have
not yet been incorporated in the models. This can also be seen in the often significant
country-fixed effects.
Models focus normally on embodied technology spillovers, while some also take
disembodied spillovers into account. The determination of the level of technology in the
-24-
Technology Transfer Through Trade
donor country stands central to the analysis. Moreover, the correct description and
functioning of the level and kind of human capital turns out to be problematic and large
differences in results occur, when different measures of human capital are used. We will
now turn towards the specifications and results.
The influential article of Coe and Helpman (1995) started the empirical debate on the
effects of knowledge-spillovers through trade. Starting from the assumption that a
country’s productivity depends on both the domestic and the foreign stock of knowledge
they use cumulative R&D spending in trying to assess the spillover effects through trade
between 21 OECD countries and Israel. Building on Ethier (1982), they assume that the set
of horizontally or vertically differentiated available intermediate goods influences factor
productivity and proxy this set of inputs by accumulated R&D spending. The central
concept of Coe and Helpman is a measure of “foreign R&D stock” for each country. The
R&D capital stocks of a country’s trade partners are weighted by import shares, reflecting
the importance of a trade partner in determining the domestic foreign capital stock, and are
summed up, representing the “foreign R&D capital stock”. Their specification and basic
idea builds the basis for a wide set of models dealing with international knowledge
spillovers.10
A serious drawback of the Coe and Helpman (1995) model is, that it does not include the
theoretically important human capital. This might be valid under the assumption that
human capital levels are quite similar between the 15 non-G7 countries of the sample. In
contrast to this model, Coe, Helpman and Hoffmaister (1997) (henceforth CHH) include a
variable for human capital as they focus on 77 developing countries and the R&D stocks of
the same 22 industrial countries Coe and Helpman used. In their analysis, the authors point
towards the double importance of human capital that directly influences productivity and
increases adoptive capacity. The authors use the secondary school enrolment rate as a
proxy for human capital, but being aware of this measure’s shortcomings, they also
experiment with different constructions of existing data (especially the primary enrolment
rate) without a change in results. They start from the assumption that imported products
embody foreign technology and other information that otherwise would be costly to
acquire. They model is specified first differences:
(5.1)
)log(loglog 0
itit
SM
iit
E
iit
M
iit
S
iiit SmEmSF ++++=∆ααααα
itt
T
ititit
SM
iTSE µαα+++ )log(,
-25-
Technology Transfer Through Trade
where stands for TFP, for the foreign R&D stock, for the secondary school
enrolment rate, for the share of imports of machinery and equipment from industrial
countries relative to GDP and T is a time trend. The foreign knowledge stock is interacted
both with the import share and the level of human capital, in line with the assumption that
human capital increases the rate of transfer and openness increases the access to foreign
technologies. The foreign R&D capital stock in country i, , is calculated as the bilateral
import shares-weighted foreign knowledge stock, , where
i
Fi
Si
E
i
d
k
S
i
m
S
ik
ψ
=
22
1k
=
i
Sik
ψ
represents the
bilateral import shares of country i with respect to industrial country k and the R&D
capital stock of country k.
d
k
S
Their estimation leaves a large and positive coefficient before the -term, indicating the
importance of human capital for TFP. This is not surprising, however, as the authors
calculate TFP as
it
E
)1( αα−
=LK
realGDP
TFP , thus including human capital in their measure of TFP.
A second restriction is the fact that they do not include domestic R&D spending. While
this is conceptually problematic, they implicitly justify their decision by the skewed
distribution of worldwide R&D expenditures.11 Still, even though developing countries
might conduct R&D at a low level, differences in spending between these countries can be
large, having a potentially large effect and should therefore be included.12
CHH drop and as this term turns out to be statistically insignificant. Still
they find both human capital variables jointly significant. While dropping the direct effect
of the foreign knowledge stock appears theoretically justifiable, dropping the interaction
term stands contrary to the theoretical foundations and calls for the use of different human
capital concepts. Neither do they include an interaction term between the three concepts of
foreign knowledge stock, the import share and human capital. Human capital therefore
influences TFP growth only as independent factor. When including a catch-up variable for
the scope of technological learning and fixed effects all coefficients increase. Moreover,
the elasticity of TFP with respect to an increase in the level of human capital becomes
larger than the elasticity of TFP with respect to an increase in the interaction term between
the foreign R&D stock and the import shares. The catch-up variable has the expected
positive effect on TFP growth. From their estimations, they concluded that inventive
activity in developed countries has a large influence on developing countries through trade,
increasing with the ratio of imports to GDP and the foreign R&D capital stock. As a
it
Sitit SE log
-26-
Technology Transfer Through Trade
summarising measure, the authors state their quantified conclusion that the spillover effects
of R&D spending in industrial countries have increased output of developing countries by
US$21 billion, as compared to US$50 billion of official development aid.
Lichtenberg and Pottelsberghe (LP) (1996) attack the setting of the Coe and Helpman (CH)
(1995) model from two different angels. First, they point out that other channels of
technology transfer such as FDI are not taken into account while they are potentially
important. And secondly, they criticise the mechanism by which Coe and Helpman
estimate their foreign R&D capital stock and the way they estimate its impact on TFP.
Both criticisms apply evenly to the CHH model. This section will deal with the second
criticism and its improved estimation results, as in Lichtenberg and Pottelsberghe (1998).
LP argue that CHs specification does not reflect the intensity of research in country .
They therefore propose to include research intensity – a ratio of research to GDP - of the
donor country in the measure to avoid an aggregation bias. A potential merger of two
countries would increase the foreign R&D stock of the importing country, while this bias is
largely reduced in the LP measure. Moreover, LP do not use indexed domestic R&D
stocks, arguing that they lead to an incorrect estimation of the elasticity the foreign R&D
stocks with respect to domestic TFP in a specific year.
j
13 Their specification reads:
(5.2) i
ji
d
j
i
ij
it
it
fd
i
dd
i
d
ii y
S
M
M
y
M
SGSF εαααα θθ
θ
+
+++=
=
21
1
32
1
7
0loglog7loglog .
The variables reflect the before described values and the three parameters θ, θ, and θ
can be used to reduce the model to the version of CH, when θ and θ.
12
3=
3
121 = 0
Repeating CH’s regression without indexation and using their preferred specification with
and θ, the regression with the LP measure of the foreign R&D stock has
a better fit than the CH-specification.
021 =θ13=
14
LP include the import-ratio as a measure for openness within their foreign R&D stock
measure, but also include the import share independently on the right hand side. The
elasticity of TFP with respect to the foreign capital stock interacted with the import share
in the LP specification is much larger than without the interaction but the import share has
a negative influence. The negative impact of the import share, however, indicates that the
foreign R&D capital stock must exceed a certain value for openness to have a positive
effect on TFP growth.15 It remains to be pointed out, that inferences for developing
countries are difficult to make, as no human capital is included and developing countries
are characterised by much less domestic R&D spending.
-27-
Technology Transfer Through Trade
Mayer (2001) does not use a foreign R&D capital stock at all in his regression. He rather
uses the import to GDP share of machinery imports from countries that have a substantial
ratio of R&D to GDP. This is a rather similar construct to the measure proposed by LP.
Research intensity is what appears important to Mayer. The LP measure weights imports
with precisely this research intensity measure of donor countries. The results of both
cannot be compared, however, as LP use developed country data while Mayer uses a set
consisting of developing countries.
Wang and Xu (2000) test three different specifications for the foreign R&D capital stock,
using data on trade in capital goods for 21 OECD countries. Both the measure presented by
Coe and Helpman (1995) including the general import share, , and the preferred
specification of Lichtenberg and Pottelsberghe, , are used. In addition they test a
third, unweighted spillover variable that is supposed to pick up disembodied technology
transfer. They complete their first model by introducing a human capital variable, H. As
they assume the variables to be non-stationary, they construct a second specification. This
is based on first differences and includes a variable for the technology gap between the
countries’ technology level and the world frontier, , that increases with a decrease in
the size of the technology gap. It can be seen that their GAP measure and the unweighted
spillover variable both potentially capture disembodied spillovers.
CHf
S
LPf
S
GAP
)(UWS f
(5.3) )(logloglog7loglog 4321 UWSSSGSF f
it
f
it
d
it
d
ittit ++++=∆ββββα
itit GAPH µββ+++ loglog 65 .
Wang and Xu first run two regressions with a reduced model – leaving out the unweighted
spillover variable, human capital and the technology gap, and defining the model in
levels – first using m (the foreign R&D capital stock, corrected for the share of
imports to GDP, ) and then using log . For both, the coefficient for is
positive and statistically significant at the one-percent level, while the coefficient on
domestic R&D capital is negative but not statistically significant. The coefficient on the
LP-measure is half as large as the coefficient of the CH-measure but the fit of the entire
model is superior with the LP-measure.
CHf
S
log
mLPf
Sd
it
SG log7
Next they estimate the complete model with the m and the measure.
The findings for the coefficients in the reduced model change only minimally, human
capital and the unweighed spillover variable are not significant, while the coefficient on
carries the expected negative sign and is statistically significant at the one-
CHf
S
log LPf
S
log
GAPlog
-28-
Technology Transfer Through Trade
percent level. The similarity of the variables and might result in
the statistically insignificance of the coefficient of in the estimations.
)(log UWS f
log S
log
it
ES
iit
E
iEE αα
+
GAPlog
)
Sm log
itit
Sµα+
(UW
f
S=
log
16
Wang and Xu point out that neither of their three measures of the foreign R&D capital
stock is derived from theoretical models as they are not specified, and call for theoretical
approaches to deriving them.17
Engelbrecht (2002) builds on the same framework, focussing on TFP determination in
developing countries and using a data set of 22 industrial and 61 developing countries.18
He tries to establish a more precise picture of human capital effects by using different
measures for human capital as well as “policy-conditioned” human capital variables. In his
last specification, also a catch-up variable is included in order to test for knowledge
spillovers that are not related to R&D. Engelbrecht “conditions” human capital, E, with the
import share, , , and replaces the individual effect of the foreign R&D capital
stock with a “conditioned” foreign R&D capital stock as .
m mEE =
(5.4) ,
it
S
iit
M
iitit SmF αα +
++= loglog 0
where the variables have already been described. Using different measures for human
capital,19 Engelbrecht finds that the conditioned human capital variable has an significant
and positive effect on TFP growth, especially when using secondary schooling in the total
population. Differentiating for average male and female human capital, he concluding that
female human capital has a stronger effect on TFP growth than male human capital, this
finding being even more pronounced when using “average years of primary schooling in
the female population”. This variable might, however, rather describe the structure of
society, pointing towards the importance of female emancipation for TFP growth.20
The coefficient of is negative in all estimations, possibly due to the double
scaling of the interaction term, indicating the need for another form of foreign R&D
absorption.
itit SE
log
21 Includes a catch-up variable, an interaction term between the distance of GDP
per-capita in the developing country to that of the average OECD country, and using the
non-conditioned human capital variable, he finds a positive effect of human capital, in
particular of secondary education, in the absorption of international knowledge spillovers
other than those associated with R&D. It would be interesting to see whether this finding is
also dependent on the degree of openness as could be tested with an interactive term
.
ititit CmE log
-29-
Technology Transfer Through Trade
Miller and Upadhyay (2000) use a different framework for 83 countries from 1960-1989.
Using two production functions, with and without human capital, they estimate two values
for TFP in the sample countries. They then regress these TFP measures on the stock of
human capital approximated by the average years of schooling of the adult population, the
ratio of exports to GDP, the terms of trade, local price deviations from purchasing power
parity (PPP), and the inflation rate as well as the standard deviations of these variables.
They do not take the foreign R&D capital stock into account but include an interactive
term between exports and human capital. Moreover, they include fixed effects for countries
and six time variables in order to account for time specific effects. In general, their setting
does less well fit the theoretical analysis.
Their results indicate that that the more open an economy is (and the more stable the
openness variable), the higher is its TFP. Human capital, through the interactive term,
increases this effect. Upward deviations from the PPP and a higher inflation rate have a
negative effect on TFP. The last two factors, however, represent a rather indirect effect on
TFP through external demand and investments that are not of direct interest for this paper.
Still, they find that human capital only has a positive effect on TFP when the ratio of
export/GDP ratio exceeds eleven per cent. Also, time effects for the first four periods are
statistically significant. For the TFP2 measure, the results change slightly, indicating a
openness threshold of 50 per cent. Their method of deriving the TFP1 measure assigns
productive effects of human capital to TFP while this is not the case for TFP2. As can be
shown, this largely explains the negative coefficient of human capital in the latter case.
Finally, they divide the countries in the sample into three groups according to their
economic development. Correlations between independent variables and the sets of
countries, however, might lead to problems when estimating as not the complete
relationship is captured within each group. Results are improved for the segregated
regressions and the general model appears to describe the TFP level determination of the
middle income countries (between $US3000-$US10000 on average in 1960-1964) best.
Falvey, Foster and Greenaway (henceforth FFG) (2002) estimate output growth rather than
TFP levels with the help of a static and a dynamic model, using a sample of five OECD
donor countries and 52 developing recipient countries. They argue that knowledge
spillovers have a short- and a long-run impact on growth and include lagged GDP growth
rates. They specify in first differences:
-30-
Technology Transfer Through Trade
(5.5)
it
iittitiit GDP
Inv
ySPILLyyy
++++= 365,212,21,1 lnlnlnlnln βββαα
,
itititiit SACHSTTISECPOP εββββ∆+++++ 7665,54 ln25ln
where represents a change in a variable, ln and ln the lagged values of GDP
growth, the 1965 GDP level, a spillover variable,
1,
ti
y
it
2,
ti
y
65,
ln i
y SPILL
it
GDP
Inv
it
SACHS
the investment
to GDP ratio, the population, the percentage of people over age 25 with
secondary education in 1965, TTI a terms of trade index, and the Sachs and
Werner (1995, as cited in FFG (2002)) index of openness. The static model is similar to
specification (5.5) but does not include the two lagged values of GDP as independent
variables.
it
POP 65,i
25SEC
it
The authors test different measures of the foreign knowledge stock by substituting them as
the variable in their model. They distinguish their measures through the degree of
“publicness” in the donor- and in the receiving country as well as through the way the
imports are weighted in the receiving country. The five specifications are summarised in
table (5.1). When technology is a public good in the donor country, the capital stock enters
directly in the calculation. Assuming that technology in the donor country is a private
good, the capital stock is weighted by the donor country’s GDP, giving a measure of
knowledge-intensity. All else equal, this means that a larger country (a country whose
capital stock is smaller per unit of production) has a smaller effect on the spillover variable
in the recipient. For the recipient country, the “imported knowledge” is weighted with
either total imports (specifications 1M and 4M) or with GDP (specifications 1Q and 4Q).
This leaves, other things equal, the spillover variable of a larger (or more closed) economy
smaller, bringing the amount of “imported” foreign knowledge in relation to the country
size or its openness.
it
SPILL
Both for the static an the dynamic models, including the SACHS openness measure does
not change results, being itself significant at the one-percent level. The catch-up variable,
the measure for human capital, and the investment-GDP ratio are always of the expected
negative sign and significant at the one-percent level. The population variable is mostly
insignificant and carries changing signs. The terms of trade index is only significant in
some of the specifications of the dynamic model (1Q, 2, 3, 4Q) and has a positive sign.
The discussion here focuses on the spillover variable, however, as it represents the effects
of trade on growth through knowledge spillovers.
-31-
Technology Transfer Through Trade
Table (5.1): The five foreign R&D stock specifications of FFG (2002)
Specifi-
cation
Calculation Characteristic
in the
donor/recipient
country
Construct
used by
Effect in
static/
dynamic
model
(1M) ∑∑
==
dd
dtdrt
rt
dtdrt
rt K
M
KM
SKM θ Public/Private CH
(1995,1997)
-0,06***/
-0,04***
(1Q) ∑∑
==
dd
dtdrt
rt
rt
rt
dtdrt
rt K
Q
M
Q
KM
SKQ θ Public/Private CH
f
Sm log
0,02**/
0,01
(2) ∑∑
==
dd
dt
dt
drtrt
dt
dtdrt
rt Q
K
M
Q
KM
SK θ Private/Public LP (1998)
(0-0-1)
0,07***/
0,06***
(3) ∑∑
==
dd
dtdrtrtdtdrtrt KMKMSK θ Public/Public 0,05***/
0,03***
(4M) ∑∑
==
dd
dt
dt
drt
rtrt
dtdrt
rt Q
K
QM
KM
MKS θ Private/Private -0,14***/
-0,17***
(4Q) ∑∑
==
dd
dt
dt
drt
rt
rt
dtrt
dtdrt
rt Q
K
Q
M
QQ
KM
QKS θ Private/Private 0,03**/
0,02
rt
M= total imports of country r
drt
M= share of ’s imports that come from country r d
rt
Q= country ’s GDP r
dt
Q= country ’s GDP d
dt
K= country ’s R&D capital stock d
drt
θ= share of imports from country d in total imports of country r
r stands for the recipient country and d for the donor country.
CH stands for Coe and Helpman (1995)
LP stands for Lichtenberg and Pottelsberghe (1998)
Note: ***,**,* represent significance of the spillover variables in the regressions including the SACHS-openness
measure at the 1, 5, and 10 percent level, respectively
Source: Own summary
Their results find a significant relationship for some but not all of the spillover measures
(see table (5.1)). Only the measures that consider spillovers to be of public nature in the
receiving country are positive and statistically significant in both models. When spillovers
are considered private in the receiving country, the method of deflations has an effect on
the results. Total manufacturing imports used as deflator, taking the distribution of imports
into account but not the total volume of them, leads to a statistically significant negative
sign, contradicting the theoretical concept of spillovers. When GDP is used as a deflator,
the coefficients on the spillover measures become positive but remain non-significant in
the dynamic model. From the dynamic model, FFG estimate the long-run effect of
knowledge spillovers on growth. For the estimations that include the openness measure, an
increase of one percent in the spillover variable (the foreign R&D capital stock) increases
the long-run growth rate by between 0,011 (specification 1Q) and 0,067 (specification 2)
-32-
Technology Transfer Through Trade
percent. The authors point out that long-run effects might be of a longer time horizon than
the two-year lagged variables.
The results for various spillover variables leaves the reader with some doubt. Not all
measures are of the correct sign or statistically significant. The assumption on the nature of
spillovers in the donor country do not have a large effect on the results. Modelling them as
private, and keeping the assumptions on the recipient constant, however, leads to larger
values for the coefficients, indicating that their nature is mixed but apparently more private
than public. With respect to the nature of the spillovers in the recipient, the same analysis
indicates that the “public” assumption performs better in the estimations. While the
discussion on the nature of technology in both countries continues, the assumption of
private nature in the donor- and public nature in the recipient country is also intuitively
appealing, while also the measures of public/public perform well. The first case
(specification (2)) leaves us with the measure that is proposed by Lichtenberg and
Pottelsberghe (1998) (henceforth LP) for θ and θ, while the second
(specification (3)) results in a specification with θ, for which LP do not
test, and for which FFG’s results are not appealing. While FFG test for GDP growth, the
measures they use are inherently included in the framework of LP and specification (3) of
FFG performs also especially well in the LP-framework indicating that knowledge be best
considered private in the donor and public in the recipient country.
021 =
1θ
13=
0=32 =
Keller’s (1999; 2002) model, deals with the output effect of R&D stocks on industries’
TFP through using intermediate products that come from other sectors and other countries
in production using a sample of 13 industries in the G7 and Sweden in the time period of
1970-1991. Four different measures are used, the domestic, own sector R&D stock, the
domestic, alien sector R&D, the foreign, own sector R&D stock, and the foreign, alien
sector R&D stock. Measures are calculated using import-export weights (IO), in line with
Coe and Helpman (1995) and with the help of a technology flow matrix (TM), as described
in Evenson et al. (1991). As the first measure performs better, this paper will focus on the
corresponding results.
No measure of human capital in these countries is used. Due to the high homogeneity of
the sample, this might be justifiable, with respect to technology transfer to developing
countries, however, this surely would have to be adjusted. He regresses TFP on the
domestic and foreign (weighted with bilateral import shares), own and alien, sector R&D
stocks, a time-, and country-fixed effect.
-33-
Technology Transfer Through Trade
While Keller’s model does not take other effects than the R&D stocks into account and
deals with a set of developed countries, his basic conclusions are of interest. Under his IO
specification, Keller calculates the relative importance of the four stock variables for TFP
determination. It turns out that the domestic, own sector R&D stock accounts for 51,1
percent of the total effect, the domestic, alien sector R&D for another 29,2 percent, the
foreign, own sector R&D stock for 4,7 percent, leaving 15,0 percent for the foreign, alien
sector R&D stock.22
Hakura and Jaumotte (1999) analyse the effectiveness of inter- and intra-industry trade on
a country’s growth rate for a sample of 24 OECD and 63 developing countries. The paper
includes only trade between OECD countries (technological leaders) and the developing
countries, not touching the issue of trade between Newly-Industrialised countries and
developing countries. Assuming that an increasing gap in TFP between leaders and
followers reduces the cost of adoption, they test for the effect of the import-adjusted TFP
gap and a disembodied convergence factor. This paper strives for quantifying this
parameter with the help of the Grubel-Lloyd intra-industry trade index (IIT). This index
takes a value of zero if no intra-industry trade exists and a value of one if all trade within a
sector is intra-industry. In their regression, sectors are classified as being characterised by
intra- and inter-industry trade depending on the value of the IIT. They explore cut-off
values ranging from 0,1 to 0,9.
Results for TFP-estimations with and without human capital are similar, therefore they
only present the estimations based on TFP . These estimated TFPs are used to
construct the TFP gap which is then used in the regression.
6,04,0
/LKY=
(5.6) i
i
l
IRsIAsi
ils
i
ils
li TFP
TFP
y
m
y
m
TFPcTFP εγβα+
+++= ∑∑
∈∈
lnlnln ,
where indices i represent a developing country, indices the leader countries, indices
sectors which might be either inter-industry ( or intra-industry ( sectors, and
represents the ratio of imports of sector of country i from country to country
’s GDP, representing a measure of openness.
l s
)IR
s
)IA
iils ym /
i
l
First, Hakura and Jaumotte do not include a measure of “adoptive capacity” in
determination of the TFP growth rate. Moreover, they also assume that TFP growth does
depend on foreign TFP growth as a form of disembodied technology spillovers. The
coefficient of foreign TFP growth is close to one as the authors expect, but the mechanisms
through which this technology growth is transferred to the developing countries, is not
-34-
Technology Transfer Through Trade
explored and remains an open question. They come to the conclusion, however, that intra-
industry trade has a much larger effect on TFP growth than does inter-industry trade, even
though some of their estimated coefficients for intra-industry sectors, when disaggregating
for regions, are negative (e.g. South Asia).23
6 Theory, Empirics, and Conclusion
The models presented focus both on embodied and disembodied technology spillovers and
on the role that human capital plays in the adoption of technology. They are predominantly
defined in growth rates. Generally, it has to be said that the estimations test only for a
small amount of the variables that were theoretically presented. While the increase in the
accessible knowledge stock is included in nearly all models and turns out to be significant,
the interaction with domestic conditions are less frequently tested for and often turn out to
be insignificant. Moreover, a long-term effect of openness is never tested for. If one
assumes that knowledge diffuses slowly, a longer openness to trade will increase the effect
of openness on TFP growth. The tested mechanisms differ considerately from the
underlying theory. The concept of the spillover variable and the accessible knowledge
relate to the model of Wang and also to the model of Rivera-Batiz and Romer in its
adopted form. The idea of the gap-variable is closely related to the model of Nelson and
Phelps, even though the interaction term of human capital (or adoptive capacity) that they
use is normally not taken into account. Moreover, the gap variable is used representing
disembodied spillovers, and does not indicate the channel through which these spillovers
reach the country. Besides, the factors that influence the absorptive capacity are not
specifically tested for. An analysis of the empirical findings shows that the empirical
models that were presented cannot profoundly illuminate the underlying functioning of the
true transfer process, but instead rely on the concept of spillovers and catching up. The
most important differences and findings are presented in table (6.1).
The specification of spillovers
A central point of discussion relates to the construction of the spillover variable. As Wang
and Xu (2000) point out, the theoretical concept does not aid in giving guidance for the
empirical construction of this variable. As a result, the precise construction of the
embodied spillover variable remains strongly debated. Both Lichtenberg and Pottelsberghe
(1998) (LP) and Falvey, Foster and Greenaway (2002) (FFG) present a set of different
constructs and test for the one that is most significant. LP state that the foreign R&D stock
-35-
Technology Transfer Through Trade
should take the research intensity of the exporting country into account and weight the
foreign R&D stock by foreign GDP. In contrast to LP, FFG approach the problem of the
spillover variable from a more holistic point of view. They analyse the nature of spillovers
in the donor and the receiving country and construct six variables which they test using an
estimation in output growth rates with lagged variables for a set of developing countries.
They come to the conclusion that spillovers are best modelled when considered as a private
or public good in the donor country and as a public good in the recipient country.
Interestingly, LP’s and FFG’s preferred specifications are similar, even though it is
difficult to compare the two sets of regressions of LP and FFG as they use levels for
OECD countries and growth rates for developing countries, respectively.24 Especially the
private/public spillover variable that FFG construct mirrors the preferred specification of
LP. The public/public spillover variable of FFG is not tested for in any of the other models,
but it can also be described in the LP-framework, with all three θ set to zero. While some
results are obtained, the actual nature of technology in both donor and receiving countries
remains a field for further research.
Related to the construction of the spillover variable is the question of which categories of
imports should be taken into account. Conceptually, intermediate good imports have an
effect on the accessible knowledge stock, while capital good imports represent the direct
channel of technology transfer. Half of the estimations uses trade in intermediate or
manufacturing goods, while the other half uses machinery imports for the construction of
the spillover variable. Coe et al. (1997), Wang and Xu and Engelbrecht use machinery
import data. With respect to this category, Mayer (2001) tests for different kinds of
machinery imports on output growth. His results show that the largest impact is derived
from general machinery imports that can be used in different sectors, followed by sector-
related machinery, while the lowest elasticities of imports on income per-capita growth is
found for the general category of machinery and transport equipment that Coe el al. (1997)
use. Machinery imports, however, are more related to trade in capital goods and have a
different effect on technology transfer than trade in intermediate goods. This understanding
is also shown by Coe et al. (1997) who repeat their estimations also with a spillover
variable based on trade in manufacturing goods. When doing so, the fit of their model is
nearly identical to the one where they use machinery import data.
Following the argumentation of Navaretti and Soloaga (2001) who find that the quality of
imported machinery is correlated with a country’s GDP, the quality of the imported
-36-
Technology Transfer Through Trade
machinery goods should also be taken into account. It would be of interest to investigate
the effect of the complexity of technology on productivity gains. Incompatibilities of the
demanded and existing skill level in a country can reduce the positive effects of machinery
imports on TFP growth. This argumentation is founded in the writings of Basu and Weil
(1998) and Acemoglu and Zilibotti (1998), claiming that technologies have to be
appropriate in order to give rise to the full effect on TFP. Estimation problems are
therefore likely to occur when it is neither accounted for quality nor for the skill level of
the importing country and a countries’ imports are biased towards high-technology. As a
result, the estimated coefficients might not appropriately reflect the effect of the foreign
R&D technology stock and might be biased downward, if the technology is not fully
utilised. For both machinery and intermediate product imports, it would be interesting to
test whether the quantity of imports or the quality of imports has a larger influence. While
the quantity of imported goods increases the spread of knowledge in a country the quality
increases the set of technological knowledge that is accessible by the developing country.
Blyde (2003) includes both a trade and a FDI weighted foreign TFP measure, in order to
account for both channels of technology transfer. The effect related to trade in capital
goods seems to be stronger as it carries a larger coefficient (his estimation is specified in
first differences). Human capital interaction terms are positive and partially statistically
significant.
Also disembodied spillovers are taken into account by Wang and Xu (2000), Coe et al.
(1997), and Engelbrecht (2002) by means of the technological distance between donor and
receiving country. Wang and Xu use the relative TFP level of a country with respect to the
United states, while Engelbrecht and Coe et al. use differences in per capita income
between developing countries and the average OECD country. Both find that the
technology gap variable has a positive and statistically significant effect on the TFP growth
rate. Engelbrecht uses the technology gap variable in interaction with the human capital
term. He finds that in particular secondary education has a positive effect on the absorption
of foreign technology. This finding supports the theoretical approach of Nelson and Phelps.
Wang and Xu and Coe et al., however, do not include the interaction term that is seen as
crucial by Nelson and Phelps. Whereas this paper pointed out that the technology gap
should consist of the difference between applied technology and the set of accessible
technology, the factor of accessibility is not taken into account for disembodied spillovers.
Testing for an interaction term between the technology gap, the import share and human
-37-
Technology Transfer Through Trade
capital would give further insight into the process of technology transfer. Not taking the
import share into account, makes these variables pick up a form of disembodied spillovers
that are not related to openness or trade.
The role of human capital
Engelbrecht (2002) finds that the amount of secondary and tertiary school education have a
significant positive effect on TFP growth, when interacted with a GDP gap variable.
Additionally, he finds that primary education is particularly important in increasing
domestic productivity. This single human capital term should theoretically include
secondary inventions or learning-by-doing, as was pointed out in the theoretical section.
Yet, Engelbrecht does not succeed in modelling both the domestic advancements and the
spillovers in one model, without the human capital variable of one of the two effects
turning negative. Moreover, he does not take the experience with R&D spending into
account when analysing the determinants of the effectiveness of technology transfer in a
country, a factor that is seen by Lall (1992) or Wang (1989) as a crucial determinant of
absorptive capacity.
Miller and Upadhyay (2000) test for the significance of an interaction term between a
human capital variable (average years of schooling of the adult population) and the export
share. They find this interactive term to be positive and statistically significant, meaning
that a higher human capital stock increases the positive effects of openness on TFP. This
finding is very encouraging. Mayer (2001) also finds a positive coefficient for the
interaction term between human capital (proxied by overall educational attainment) and the
un-weighted share of imports from countries that have a high R&D intensity in relation to
total imports. Also Blyde (2003) finds partial evidence for a positive effect of human
capital on technology transfer, particularly when interacted with the FDI-flow weighted
foreign TFP. Taking the interacted and non-interacted FDI coefficient into account, there
appears to be a threshold level of human capital below which positive effects of FDI are no
longer present.
CHH find that even though both the single variable of human capital as well as the
interacted variable with the foreign R&D stock have insignificant coefficients, the two
variables are jointly significant. With respect to their measure of human capital, the
enrolment rate, it should be taken into account that especially in countries where the
enrolment rate is increasing sharply, lagged values of enrolment would more appropriately
reflect the level of human capital for the working population. The positive and statistically
-38-
Technology Transfer Through Trade
significant effect of human capital on output growth that FFG find is not helpful in this
context as it does not assess its impact on technology transfer.
It can be said that the interaction of adoptive capacity with the foreign R&D stock or the
import level appears to be positive even though the coefficient turns out to be insignificant
in some models (Coe et al., 1997). Also in the case of human capital, the precise
construction of the variable remains a field of debate. Engelbrecht points towards the
importance of tertiary education for the absorptive process, finding this relationship for
disembodied spillovers. This follows the model of Nelson and Phelps but does not explain
the channel through which technology is transferred, while pointing towards a channel of
spillovers that is not related to foreign R&D spending. Research experience is not included
in any of the models presented. In their analysis, Crespo, Martín and Velázquez (2002)
find, for a set of OECD countries, that a measure of adoptive capacity, defined as a linear
combination of the human capital- and the domestic R&D capital stock, has a positive
effect on technology transfer. The amount of actual own R&D spending has not been used
for sets of countries exceeding the OECD countries, however, leaving open the question
what the effect for developing countries is. Even though R&D spending remains low in
developing countries, large differences exist. Taking the number of researchers per million
people as a proxy-variable for R&D spending, values range from around 20 researchers in
various African countries to 200 for many other developing countries (UNDP, 2002).
These differences, and also possibly large differences in their experience with adopting
technologies, can have a potentially large effect on TFP growth and should be taken into
account.25
Inter- and Intra-industry trade
Two models that tested for the difference in inter- and intra-industry trade both pointed
towards the large importance of intra-industry trade. These findings are in line with the
obtained theoretical insights. Increased distance to technologies leads to higher costs of
adoption and a lower transfer rate. Secondly, intra-industry trade points towards sourcing
relationships and its positive effects on technology transfer have been pointed out. With
hindsight to the findings of Hakura and Jaumotte, the scope for learning of countries that
do not own certain sectors is reduced while sourcing relationships can help in creating such
industries.
-39-
Technology Transfer Through Trade
Table (6.1): Summary of spillover variables and human capital effects
Countries analysed Concept of H and effects Foreign capital stock Trade in which
goods?
Coe,
Helpman
(1995)
21 OECD +Israel No H included
=
=21
1
j
d
j
i
ij
CHf
iS
M
M
S For weighting:
imported goods and
services to GDP
Coe,
Helpman,
Hoffmaister
(1997)
1971-90.
R&D stocks of 22
industrial countries
and 77 developing
countries
Secondary school enrolment
rate in relation to population
of this age
positive effect on TFP
growth, interaction term with
foreign R&D insignificant,
jointly significant
=
=22
1k
d
k
i
ij
iS
M
M
S Average bilateral
machinery and
equipment import
share, total imports
from industrial
countries to GDP
Lichtenberg,
Pottels-
berghe
(1998)
1971-90
21 OECD +Israel
No H included
=
=21
1
ji
d
jij
LPf
Y
SM
S Imports in goods and
services
Wang,
Xu
(2000)
21 OECD countries
1971-1990
Average years of total school
attainment
positive effects but not
significant
Both
and
CHf
SLPf
S
Capital goods trade
(SITC7)
Engelbrecht
(2002)
1971-90
22 developed and 61
developing countries
Total years of schooling,
secondary, tertiary.
secondary: effect on
interaction term with gap
primary: effect on own
knowledge creation
=
=22
1k
d
k
i
ij
iS
M
M
S Definitions as in
CHH
Miller,
Upadhyay,
(2000)
1960-1989
83 countries, PWT-
data
Average years of schooling
Positive interaction with the
import share,
positive impact alone only
significant at 20 percent
level.
None Exports to GDP ratio
Falvey,
Foster,
Greenaway
(2002)
1976-1990
5 OECD countries
and 52 developing
countries
Secondary education in 1965
positive effect on output
growth
Analysis of public and private
nature in donor and recipient.
a) public/public
=
d
d
r
dr
rr K
M
M
MSK
b) private/public
=
ddt
dt
r
dr
rr Y
K
M
M
MSK
Total manufacturing
imports
Keller
(2002)
1970-1991
8 OECD countries
No H included Trade share weighted R&D
capital stocks
Inter and intra-
industry trade in
intermediate inputs
Hakura,
Jaumotte
(1999)
1970-93.
24 OECD and 63
developing countries
wage weighted schooling
levels (none, primary,
secondary, tertiary)
TFP-gap to OECD average,
weighted with trade shares
Manufacturing trade
Source: Own summary
Conclusion
Theoretically, trade has three types of effects on technology transfer. First, direct effects
result from the import of capital goods including modern technology as well as positive
output effects due to an increasing variety and quality of intermediate inputs into
production. Second, dynamic gains from trade result from an integrated world market that
leads to higher production in developing countries due to lower labour costs. This higher
production that is triggered by FDI, sourcing contracts and import or export competition
-40-
Technology Transfer Through Trade
leads via learning-by-doing to the mastering of better techniques and therefore to an
increase in TFP. Third, trade increases the set of technologies that are available in a
country. This allows imitation to an increasing degree, the more exposed to new
technologies a country is. Besides, knowledge production becomes more effective due to a
broader knowledge base and more experience in imitation and innovation. All channels of
technology transfer are heavily dependent on the country’s or firm’s technological
capability, a factor strongly influenced by the amount of human capital that exists and is
used in the economy.
This paper has presented the theoretical aspects of technology transfer, without analysing
the dynamic economic effects of specialisation in goods. It has focused on the process of
technology transfer and not on its economic, social and political effects. Results are rather
modest, and much work remains to be done until the precise process of spilling-over will
be described correctly. Three basic conclusions remain. First, the degree of openness
appears to have a positive effect on TFP, as does, second, the research intensity of the
trading partners. Third, human capital helps in increasing TFP growth. Its interaction with
imports is theoretically clear but not supported by all empirical studies while some
evidence for the importance of human capital in disembodied spillovers can be found.
Knowledge in the donor and recipient country appears to be private and public,
respectively, and intra-industry spillovers have a larger effect on TFP than inter-industry
spillovers. Still, the mechanisms of how these spillovers precisely take place, however,
have generally been ignored, as Grossman and Helpman (1990) point out. In this context,
the analysis of different empirical estimations leaves many questions unanswered. The
consistent usage of country-fixed effects indicates that other, relevant factors have not been
taken into account. It would be of large interest to specify them more precisely. In
particular, the correct construction of the spillover variable remains a key element for
research. While accessibility is often taken into account for embodied spillovers, this is not
the case for disembodied spillovers, standing in contrast with the theory. While generally
giving support to the assumption that human capital helps in technology adoption, the often
low significance of the human capital variable calls for a more precise measurement of the
elements of human capital that are important in technology transfer. Technology transfer
from newly developed countries to developing countries has not been analysed. While this
paper has looked at the absolute performance of TFP growth, also its relative performance
is of interest. As long as technology transfer occurs at a higher rate then knowledge
-41-
Technology Transfer Through Trade
creation, convergence should take place. The technology level of the developed countries
remains the moving target to aim for and the relative analysis is left for further research.
With regard to further research, some potential improvements to analysing technology
transfer have been presented. First, the exact channels of embodied and disembodied
spillovers remain undetermined while being crucial for a clearer understanding of
technology transfer. In this context, the characteristics of technology as being private or
public in both donor and recipient country need further analysis. Second, the quality of
imported technology and the existing skill level should be taken into account in order to
correct for appropriateness and inefficient technology usage. Long-term effects of trade or
stable trade relationships should be tested for as they can increase compatibility of existing
and new technologies. Third, the effects of past absorptive capacity and therefore of
experience with technology transfer should be assessed as they describe the scope for past
learning-to-learn, thereby increasing the efficiency of the imitative process. Last, and more
generally, it would be desirable to assess factors in technology transfer at the micro level
and to integrate the findings into macro analysis. Available data that would allow to
determine factors that influence productivity remains rare at micro level, however, as
Hasan (2002) points out. The same is valid for the macro level, calling for the construction
of more sophisticated data sets.
One can conclude that openness has a positive influence on TFP-growth. This positive
effect might be supported or counteracted by other effects of openness on growth that are
not considered here. Moreover, increased emphasis on human capital formation appears to
increase the potential for – and the rate of – TFP-growth. Results are more robust for
developed countries than for developing countries, while the potential effects of better
policy recommendations are far larger for the latter group. Results obtained from the sets
of developed countries are not readily transferable to developing countries and should
therefore not be used to develop policy recommendations. While productivity levels in the
world continue to be unequally distributed, explanations for this phenomenon remain basic
and are not very insightful. Many questions have to be answered before the concept of
technology transfer will be thoroughly understood and more accurate policy proposals can
be made.
-42-
Notes
Notes
1 Compared to neoclassical growth theory, models of endogenous technological change are much more
appealing in explaining the underlying reasons for economic growth in the long-run. See the “quality-ladder”
model, including the concept of obsolescence (Aghion and Howitt (1992), Grossman and Helpman (1991b),
Romer’s (1990) model of differentiated capital inputs, or the model of Grossman and Helpman (1993), who
model economic growth as being based on micro-level inventions. Moreover, endogenous technological
progress presents an immediate explanation for differing technology growth rates between nations, a fact that
can be observed in the available data and which could not be explained by exogenous growth theory.
2 Historically, trade was the only way by which ideas could disperse between cultures. While some basic
inventions, such as agriculture and cooking, were probably invented independently by different cultures, they
were also at least locally diffused. For this form of inter-society diffusion of technology, war and trade have
been the basic means, war forming in this respect a particular form of trade (or transfer of ideas). The
moulding of iron or bronze, the Arabic sailing boat with triangular sails, gunpowder, coining, the modern
banking-system, and cotton spinning present some obvious examples.
3 For a survey, the reader is referred to Nordhaus (2001). Here, the author stresses the point that none of the
normally used measures for TFP growth is consistent with the theoretical foundations. He points towards the
importance of differences in productivity levels between sectors, the relative changes of employment
between sectors, and changes in relative importance of sectors within the economy which form three distinct
elements in determining TFP growth as it is usually measured.
4 Profits for local firms will also exist when production costs in developed and developing prices are
identical. The resulting gap in prices has to be taken into account when analysing the profit opportunities for
domestic firms selling their products at home. In the case of export competition, however, their price
advantages will have to be even larger than equalising prices as transport costs must be covered as well in
order to compete successfully.
5 This spending might be R&D spending or product development projects as well as be implicit in the salary
of a worker who is trying to copy a product. It does not have to be officially termed R&D spending.
6 See http://www.mofa.go.jp/region/africa/kenya/
7 Recognising this problem, capital goods were exported in a bundle as early as 1845 in the English textile
industry and this form of exports also developed later in other sectors (Clark and Feenstra, 2001), meaning
that technical assistance, construction expertise and operating personnel were supplied together with the
machinery, a huge success in the particular cases and reducing the risk for the importer.
8 Lucas (1988) argues that learning-by-doing increases the stock of human capital. Due to the reasons that
were given in chapter II, its effects will accrue to TFP within the framework of this paper.
9 It has to be pointed out that for an import competing firm, with most likely being considerably
lower than . In the case of an export competing firm that is selling in foreign markets, the degree of
exposition to foreign technology will equal .
*
Az z
*A
*A
10 Keller (1998) argues that randomly created bilateral import shares actually work better than the original
ones, casting doubt on the results of Coe and Helpman (1995). In a reply to Keller, Coe and Helpman (1999),
however, argue that the shares created by Keller are rather “simple averages with a random error”,
performing worse than the actual import shares.
11 In 1991, the G7 accounted for 92 per cent of all R&D activities.
12 CHH realise that it could be approximated by the number of scientists that work in a country, for example.
They argue, however, that differences are small and that the variable can be ignored. They argue for the
exclusion, knowing that the necessary data does not exist.
13 For a more detailed analysis of the problem, the reader is referred to Lichtenberg and Pottelsberghe (1998).
14 They find an of 0,665 as compared to an of 0,600 for CH while the coefficient of θ is found
not to be significantly different from zero.
2
R2
R2
13,02,11 ==θ
it
H
log )(log UWS f
15 It remains open, however, why LP do not test the specification where θ without
including the import share on the right hand side as independent variable as well. The results of this
specification would be interesting.
16 It should be mentioned that in their first regression (which is not described here), the results are basically
identical to those of Coe and Helpman (1995), although they exclude Israel from their sample. Including
and in their regression improves the fit of the model from an R of 0,598 to an of
0,674.
2 2
R
17 Also Crespo, Martin and Velazques (2001) make this point.
-43-
Notes
18 Engelbrecht re-estimates Coe et al.’s (1997) estimation with the improved education data as presented by
Barro and Lee (1996). Moreover, he drops 16 mainly African countries due to lacking human capital data. It
results that this re-estimation doubles the wellness of fit of Coe et al.’s model.
19 Engelbrecht does not correct for schooling quality differences between countries.
20 It would also be interesting to include measures of human capital distribution within economies as a
variable. A more equal distribution could lead to stronger TFP growth as more people could participate more
intensely in the economic process.
21 Crespo, Martín and Velazquéz (2002) find a significant effect of human capital together with domestic
R&D spending in the absorption of foreign technology for a sample of 28 OECD countries. For all their
specifications, this linearly computed adoption variable has a significant effect on TFP in interaction with the
foreign R&D capital stock.
22 Braconier and Sjöholm (1997) find only intra- but no inter-industry spillovers for a set of nine
manufacturing industries for six large OECD countries.
23 Here, the text says the opposite but from the data it is clear that intra-industry, not inter-industry trade has
the negative effect.
24 Moreover, LP use a measure for imports that contains goods and services, while FFG use manufacturing
imports.
25 In this context, the finding of Coe et al. (1997) deserves special interest. In one of their specifications, they
drop the 15 countries which perform the most R&D among the developing countries, increasing the
coefficients of human capital and the interaction term between the import share and the foreign R&D capital
stock. This indicates that using the amount of R&D spending in developed countries would influence the
estimation results.
-44-
References
References
Acemoglu, Daron and Fabrizio Zilibotti (1998). "Productivity Differences." Institute for
International Economic Studies, Stockholm, Seminar Paper(660).
Aghion, Philippe and Peter Howitt (1992). "A model of growth through creative
destruction." Econometrica 60(2): 323-351.
Aghion, Philippe and Peter Howitt (1998). Endogenous Growth Theory. Cambridge, MIT.
Allen, T. and A. Thomas (2000). Poverty and Development into the 21st Century. Oxford,
Oxford University Press.
Arrow, K.J. (1962). "The Economic Implications of Learning by Doing." Review of
Economic Studies 29: 155-173.
Barro, Robert J. and Jong-Wha Lee (2000). "International Data on Educational Attainment
Updates and Implications." NBER 7911.
Barro, Robert J. and Xavier Sala i Martin (1995). Economic Growth. New York, McGraw-
Hill.
Basu, Susanto and David N. Weil (1998). "Appropriate Technology and Growth."
Quarterly Journal of Economics 113: 1025-1054.
Benhabib, Jess and Mark M. Spiegel (1994). "The Role of Human Capital in Economic
Development: Evidence From Aggregate Cross-Country Data." Journal of
Monetary Economics 34: 143-173.
Blomstrom, Magnus and Fredrik Sjoholm (1998). Technology Transfer and Spillovers:
Does Local Participation with Multinationals Matter?, Stockholm School of Econ
and NBER; Stockholm School of Econ.
Blomstrom, Magnus and Jian Ye Wang (1992). "Foreign Investment and Technology
Transfer - A Simple Model." European Economic Review 36(1): 137-155.
Blomstrom, Magnus, Mario Zejan and Ari Kokko (1992). Host Country Competition and
Technology Transfer by Multinationals, Stockholm School of Econ and NBER;
Stockholm School of Econ; Stockholm School of Econ.
Blyde, Juan S. (2003). "The Role of Foreign Direct Investment and Imports of Capital
Goods in the North-South Diffusion of Technology." Journal of Economic
Integration 18(3): 545-562.
Braconier, Henrik and Fredrik Sjöholm (1997). "National and International Spillovers
From R&D: Comparing a Neo-classical and an Endogenous Growth Approach."
Working Paper Series in Economics and Finance No 211.
Clark, Gregory and Robert Feenstra (2001). Technology in the great divergence. NBER
Working Paper No. 8596. Cambridge, MA.
Coe, David T. and Elhanan Helpman (1995). "International R&D Spillovers." European
Economic Review 39: 859-887.
Coe, David T., Elhanan Helpman and Alexander W. Hoffmaister (1995). "North-South
R&D Spillovers." NBER Working Paper(5048).
Coe, David T., Elhanan Helpman and Alexander W. Hoffmaister (1997). "North-South
R&D Spillovers." Economic Journal 107(440): 134-49.
Connolly, Michelle (1997). "Learning to Learn: The Role of Imitation and Trade in
Technological Diffusion." Duke University Working Paper no.97-25.
Connolly, Michelle (2001). "The Dual Nature of Trade: Measuring its Impact on Imitation
and Growth." Duke University Working Paper no.97-34.
Crespo, Jorge, Carmela Martín and Francisco J. Velázques (2002). "International
technology diffusion through imports and its impact on economic growth."
European Economy Group Working Paper 12/2002.
-45-
References
Dixit, Avinash K. and Joseph E. Stiglitz (1977). "Monopolistic Competition and Optimum
Product Diversity." The American Economic Review 67(3): 297-308.
Eaton, Jonathan and Samuel Kortum (1995). "Trade in Ideas: Patenting and Productivity in
the OECD." NBER Working Paper(5049).
Eaton, Jonathan and Samuel Kortum (1999). "International Technology Diffusion: Theory
and Measurement." International Economic Review 40(3): 537-570.
Eaton, Jonathan and Samuel Kortum (2001a). "Technology, Trade, and Growth: A Unified
Framework." European Economic Review 45: 742-755.
Eaton, Jonathan and Samuel Kortum (2001b). "Trade in Capital Goods." European
Economic Review 45: 1195-1235.
Engelbrecht, Hans Jurgen (2002). "Human Capital and International Knowledge Spillovers
in TFP Growth of a Sample of Developing Countries: An Exploration of
Alternative Approaches." Applied Economics. May 34(7): 831-41.
Ethier, Wilfried J. (1982). "National and International Returns to Scale in the Modern
Theory of International Trade." The American Economic Review 72(3): 389-405.
Evenson, R., J. Putnam and S. Kortum (1991). Estimating Patent Counts by Industry Using
the Yale-Canada Concordance. Final Report to the National Science Foundation,
Yale University.
Evenson, Robert E. and Larry E. Westphal (1995). Technological Change and Technology
Strategy. Handbook of Development Economics, Volume 3. J. Behrman and T. N.
Srinivasan. Amsterdam, Elsevier.
Falvey, Rod, Neil Foster and David Greenaway (2002). "North-South Trade, Knowledge
Spillovers and Growth." European Economy Group Working Paper 15/2002.
Findley, Ronald (1978). "Relative Backwardness, Direct Foreign Investment and the
Transfer of Technology: A Simple Dynamic Model." Quarterly Journal of
Economics 92(1): 1-16.
Gollop, F.M. and D.W. Jorgenson (1980). U.S. Productivity Growth by Industry, 1947-
1973. New Developments in Productivity Measurement. J. Kendrick and B.
Vaccara, University of Chicaco Press.
Griffith, Rachel, Stephen Redding and John van Reenen (2000). "R&D and Absorptive
Capacity: From Theory to Data." The Institute For Fiscal Studies Working Paper
01/03.
Grossman, Gene M. and Elhanan Helpman (1990). "Trade, Knowledge Spillovers, and
Growth." NBER Working Paper 3485.
Grossman, Gene M. and Elhanan Helpman (1991a). Innovation and growth in the global
economy. Cambridge, MIT.
Grossman, Gene M. and Elhanan Helpman (1991b). "Quality Ladders and Product
Cycles." Quarterly Journal of Economics 106(2): 557-86.
Grossman, Gene M. and Elhanan Helpman (1993). Endogenous Innovation in the Theory
of Growth, Princeton U and NBER; Tel Aviv U.
Gruber, W.H. and D.G. Marquis (1969). Research on the Human Factor in the Transfer of
Technology. Factors in the Transfer of Technology. W. H. Gruber, and Marquis,
D.G. Cambridge, MIT.
Hakura, Dalia and Florence Jaumotte (1999). The Role of Inter- and Intraindustry Trade in
Technology Diffusion, IMF; Harvard U.
Hall, Robert E. and Charles I. Jones (1999). "Why Do Some Countries Produce So Much
More Output Per Worker Than Others?" Quarterly Journal of Economics 114: 83-
116.
Hasan, Rana (2002). "The impact of imported and domestic technologies on the
productivity of firms: panal data evidence from Indian manufacturing firms."
Journal of Development Economics 69: 23-49.
-46-
References
Islam, Nazrul (1995). "Growth Empirics: A Panel Data Approach." Quarterly Journal of
Economics 110: 1127-1170.
Keller, Wolfgang (1999). How Trade Patterns and Technology Flows Affect Productivity
Growth, U TX at Austin and NBER.
Keller, Wolfgang (2002). "Trade and the Transmission of Technology." Journal of
Economic Growth 7(1): 5-24.
Klenow, Peter J. and Andrés Rodríguez-Clare (1997). "The Neoclassical Revival in
Growth Economics: Has it Gone Too Far?" NBER Macroeconomics Annual: 73-
102.
Lall, Sanjaya (1992). "Technological Capabilities and Industrialization." World
Development 20(2): 165-86.
Lichtenberg, Frank R. and Bruno van Pottelsberghe de la Potterie (1998). "International
R&D Spillovers: A Comment." European Economic Review. September 42(8):
1483-91.
Lichtenberg, Frank and Bruno van Pottelsberghe de la Potterie (1996). International R&D
Spillovers: A Re-Examination, Columbia U and NBER; U Libre de Bruxelles.
Lloyd-Ellis, Huw and Joanne Roberts (2002). "Twin Engines of Growth: Skills and
Technology as Equal Partners in Balanced Growth." Journal of Economic Growth
7: 87-115.
Lucas, R.E. (1988). "On the Mechanics of Economic Development." Journal of Monetary
Economics 22(1): 3-42.
Mansfield, Edwin (1961). "Technical Change and the Role of Imitation." Econometrica
29(4): 741-766.
Mayer, Jörg (2001). "Technology Diffusion, Human Capital and Economic Growth in
Developing Countries." UNCTAD Discussion Paper No. 154.
Miller, Stephen M. and Mukti P. Upadhyay (2000). "The Effects of Openness, Trade
Orientation, and Human Capital on Total Factor Productivity." Journal of
Development Economics 63(2): 399-423.
Movery, D.C. and J.E. Oxley (1995). "Inward technology transfer and competitiveness: the
role of national innovation systems." Cambride Journal of Economics 19: 67-93.
Mulligan, C.B. and X. Sala-i-Martin (1995). "A Labour-Income Based Measure of the
Value of Human Capital: An Application to the States of the United States." NBER
Working Paper 5018.
Muysken, Joan, Hannah Rieder and Mombert Hoppe (2002). The impact of education and
mismatch on wages: Germany, 1984 - 2000. Maastricht, University Maastricht.
Navaretti, Giorgio Barba and Isidro Soloaga (2001). "Weightless Machines and Costless
Knowledge: An Empirical Analysis of Trade and Technology Diffusion."
Worldbank Paper 2598.
Nehru, Vikram and Ashok Dhareshwar (1993). "A New Database on Physical Capital
Stocks: Sources, Methodology, and Results." Revista de Análisis Económico 8(1):
37-59.
Nelson, Richard R. and Edmund S. Phelps (1966). "Investment in Humans, Technological
Diffusion, and Economic Growth." American Economic Review: Papers and
Proceedings 56(1): 69-75.
Nordhaus, William D. (2001). "Alternative Methods for Measuring Productivity Growth."
NBER Working Paper 8095.
Pietrobelli, Carlo (1998). "The Multiple Linkages between Technological Change, Growth
and the Economy. A Review of a Research Programme." Labour 12(2): 171-87.
Pietrobelli, Carlo (2000). Technology Transfer for Developing Countries. Technology
Transfer. D. Schroeer and M. Elena. Alsdershot, Asghate.
-47-
References
Radosevic, Slavo (1999). International technology transfer and catch-up in economic
development. Cheltenham, U.K and Northampton, Mass., Elgar.
Rivera-Batiz, Luis A. and Paul M. Romer (1991). "Economic Integration and Endogenous
Growth." Quarterly Journal of Economics 107: 531-556.
Romer, P. (1990). "Endogenous Technological Change." Journal of Political Economy
98(5): 71-102.
Saggi, Kamal (2000). Trade, Foreign Direct Investment, and International Technology
Transfer: A Survey. Dallas, Southern Methodist University.
Teece, D.J. (1976). The Multinational Corporation and Resource Cost of International
Technology Transfer. Cambridge (MA), Ballinger Publishing Co.
Teece, D.J. (1977). "Technology Transfer by Multinational Firms: The Resource Cost of
Transferring Technological Know-How." The Economic Journal 87: 224-261.
UNDP (2002). Human Development Report. New York.
Vernon, Raymond (1966). "International Investment and International Trade in the Product
Cycle." Quarterly Journal of Economics 80(2): 190-207.
Wang, Jian Ye (1989). "Essays on International Trade, Technology Transfer and Growth."
Columbia University, Ph.D.
Wang, Jianmao and Bin Xu (2000). "Trade, FDI, and International Technology Diffusion."
Journal of Economic Integration. December 15(4): 585-601.
Young, A. (1991). "Learning-by-Doing and the Dynamic Effects of International Trade."
Quarterly Journal of Economics 106: 369-405.
Young, Alwyn (1995). "The Tyranny of Numbers: Confronting the Statistical Realities of
the East Asian Growth Experience." Quarterly Journal of Economics 110: 641-680.
-48-
NOTE DI LAVORO DELLA FONDAZIONE ENI ENRICO MATTEI
Fondazione Eni Enrico Mattei Working Paper Series
Our Note di Lavoro are available on the Internet at the following addresses:
http://www.feem.it/Feem/Pub/Publications/WPapers/default.html
http://www.ssrn.com/link/feem.html
NOTE DI LAVORO PUBLISHED IN 2004
IEM 1.2004 Anil MARKANDYA, Suzette PEDROSO and Alexander GOLUB: Empirical Analysis of National Income and
So2 Emissions in Selected European Countries
ETA 2.2004 Masahisa FUJITA and Shlomo WEBER: Strategic Immigration Policies and Welfare in Heterogeneous Countries
PRA 3.2004 Adolfo DI CARLUCCIO, Giovanni FERRI, Cecilia FRALE and Ottavio RICCHI: Do Privatizations Boost
Household Shareholding? Evidence from Italy
ETA 4.2004 Victor GINSBURGH and Shlomo WEBER: Languages Disenfranchisement in the European Union
ETA 5.2004 Romano PIRAS: Growth, Congestion of Public Goods, and Second-Best Optimal Policy
CCMP 6.2004 Herman R.J. VOLLEBERGH: Lessons from the Polder: Is Dutch CO2-Taxation Optimal
PRA 7.2004 Sandro BRUSCO, Giuseppe LOPOMO and S. VISWANATHAN (lxv): Merger Mechanisms
PRA 8.2004 Wolfgang AUSSENEGG, Pegaret PICHLER and Alex STOMPER (lxv): IPO Pricing with Bookbuilding, and a
When-Issued Market
PRA 9.2004 Pegaret PICHLER and Alex STOMPER (lxv): Primary Market Design: Direct Mechanisms and Markets
PRA 10.2004 Florian ENGLMAIER, Pablo GUILLEN, Loreto LLORENTE, Sander ONDERSTAL and Rupert SAUSGRUBER
(lxv): The Chopstick Auction: A Study of the Exposure Problem in Multi-Unit Auctions
PRA 11.2004 Bjarne BRENDSTRUP and Harry J. PAARSCH (lxv): Nonparametric Identification and Estimation of Multi-
Unit, Sequential, Oral, Ascending-Price Auctions With Asymmetric Bidders
PRA 12.2004 Ohad KADAN (lxv): Equilibrium in the Two Player, k-Double Auction with Affiliated Private Values
PRA 13.2004 Maarten C.W. JANSSEN (lxv): Auctions as Coordination Devices
PRA 14.2004 Gadi FIBICH, Arieh GAVIOUS and Aner SELA (lxv): All-Pay Auctions with Weakly Risk-Averse Buyers
PRA 15.2004 Orly SADE, Charles SCHNITZLEIN and Jaime F. ZENDER (lxv): Competition and Cooperation in Divisible
Good Auctions: An Experimental Examination
PRA 16.2004 Marta STRYSZOWSKA (lxv): Late and Multiple Bidding in Competing Second Price Internet Auctions
CCMP 17.2004 Slim Ben YOUSSEF: R&D in Cleaner Technology and International Trade
NRM 18.2004 Angelo ANTOCI, Simone BORGHESI and Paolo RUSSU (lxvi): Biodiversity and Economic Growth:
Stabilization Versus Preservation of the Ecological Dynamics
SIEV 19.2004 Anna ALBERINI, Paolo ROSATO, Alberto LONGO and Valentina ZANATTA: Information and Willingness to
Pay in a Contingent Valuation Study: The Value of S. Erasmo in the Lagoon of Venice
NRM 20.2004 Guido CANDELA and Roberto CELLINI (lxvii): Investment in Tourism Market: A Dynamic Model of
Differentiated Oligopoly
NRM 21.2004 Jacqueline M. HAMILTON (lxvii): Climate and the Destination Choice of German Tourists
NRM 22.2004 Javier Rey-MAQUIEIRA PALMER, Javier LOZANO IBÁÑEZ and Carlos Mario GÓMEZ GÓMEZ (lxvii):
Land, Environmental Externalities and Tourism Development
NRM 23.2004 Pius ODUNGA and Henk FOLMER (lxvii): Profiling Tourists for Balanced Utilization of Tourism-Based
Resources in Kenya
NRM 24.2004 Jean-Jacques NOWAK, Mondher SAHLI and Pasquale M. SGRO (lxvii):Tourism, Trade and Domestic Welfare
NRM 25.2004 Riaz SHAREEF (lxvii): Country Risk Ratings of Small Island Tourism Economies
NRM 26.2004 Juan Luis EUGENIO-MARTÍN, Noelia MARTÍN MORALES and Riccardo SCARPA (lxvii): Tourism and
Economic Growth in Latin American Countries: A Panel Data Approach
NRM 27.2004 Raúl Hernández MARTÍN (lxvii): Impact of Tourism Consumption on GDP. The Role of Imports
CSRM 28.2004 Nicoletta FERRO: Cross-Country Ethical Dilemmas in Business: A Descriptive Framework
NRM 29.2004 Marian WEBER (lxvi): Assessing the Effectiveness of Tradable Landuse Rights for Biodiversity Conservation:
an Application to Canada's Boreal Mixedwood Forest
NRM 30.2004 Trond BJORNDAL, Phoebe KOUNDOURI and Sean PASCOE (lxvi): Output Substitution in Multi-Species
Trawl Fisheries: Implications for Quota Setting
CCMP 31.2004 Marzio GALEOTTI, Alessandra GORIA, Paolo MOMBRINI and Evi SPANTIDAKI: Weather Impacts on
Natural, Social and Economic Systems (WISE) Part I: Sectoral Analysis of Climate Impacts in Italy
CCMP 32.2004 Marzio GALEOTTI, Alessandra GORIA ,Paolo MOMBRINI and Evi SPANTIDAKI: Weather Impacts on
Natural, Social and Economic Systems (WISE) Part II: Individual Perception of Climate Extremes in Italy
CTN 33.2004 Wilson PEREZ: Divide and Conquer: Noisy Communication in Networks, Power, and Wealth Distribution
KTHC 34.2004 Gianmarco I.P. OTTAVIANO and Giovanni PERI (lxviii): The Economic Value of Cultural Diversity: Evidence
from US Cities
KTHC 35.2004 Linda CHAIB (lxviii): Immigration and Local Urban Participatory Democracy: A Boston-Paris Comparison
KTHC 36.2004 Franca ECKERT COEN and Claudio ROSSI (lxviii): Foreigners, Immigrants, Host Cities: The Policies of
Multi-Ethnicity in Rome. Reading Governance in a Local Context
KTHC 37.2004 Kristine CRANE (lxviii): Governing Migration: Immigrant Groups’ Strategies in Three Italian Cities – Rome,
Naples and Bari
KTHC 38.2004 Kiflemariam HAMDE (lxviii): Mind in Africa, Body in Europe: The Struggle for Maintaining and Transforming
Cultural Identity - A Note from the Experience of Eritrean Immigrants in Stockholm
ETA 39.2004 Alberto CAVALIERE: Price Competition with Information Disparities in a Vertically Differentiated Duopoly
PRA 40.2004 Andrea BIGANO and Stef PROOST: The Opening of the European Electricity Market and Environmental
Policy: Does the Degree of Competition Matter?
CCMP 41.2004 Micheal FINUS (lxix): International Cooperation to Resolve International Pollution Problems
KTHC 42.2004 Francesco CRESPI: Notes on the Determinants of Innovation: A Multi-Perspective Analysis
CTN 43.2004 Sergio CURRARINI and Marco MARINI: Coalition Formation in Games without Synergies
CTN 44.2004 Marc ESCRIHUELA-VILLAR: Cartel Sustainability and Cartel Stability
NRM 45.2004 Sebastian BERVOETS and Nicolas GRAVEL (lxvi): Appraising Diversity with an Ordinal Notion of Similarity:
An Axiomatic Approach
NRM 46.2004 Signe ANTHON and Bo JELLESMARK THORSEN (lxvi): Optimal Afforestation Contracts with Asymmetric
Information on Private Environmental Benefits
NRM 47.2004 John MBURU (lxvi): Wildlife Conservation and Management in Kenya: Towards a Co-management Approach
NRM 48.2004 Ekin BIROL, Ágnes GYOVAI and Melinda SMALE (lxvi): Using a Choice Experiment to Value Agricultural
Biodiversity on Hungarian Small Farms: Agri-Environmental Policies in a Transition al Economy
CCMP 49.2004 Gernot KLEPPER and Sonja PETERSON: The EU Emissions Trading Scheme. Allowance Prices, Trade Flows,
Competitiveness Effects
GG 50.2004 Scott BARRETT and Michael HOEL: Optimal Disease Eradication
CTN 51.2004 Dinko DIMITROV, Peter BORM, Ruud HENDRICKX and Shao CHIN SUNG: Simple Priorities and Core
Stability in Hedonic Games
SIEV 52.2004 Francesco RICCI: Channels of Transmission of Environmental Policy to Economic Growth: A Survey of the
Theory
SIEV 53.2004 Anna ALBERINI, Maureen CROPPER, Alan KRUPNICK and Nathalie B. SIMON: Willingness to Pay for
Mortality Risk Reductions: Does Latency Matter?
NRM 54.2004 Ingo BRÄUER and Rainer MARGGRAF (lxvi): Valuation of Ecosystem Services Provided by Biodiversity
Conservation: An Integrated Hydrological and Economic Model to Value the Enhanced Nitrogen Retention in
Renaturated Streams
NRM 55.2004 Timo GOESCHL and Tun LIN (lxvi): Biodiversity Conservation on Private Lands: Information Problems and
Regulatory Choices
NRM 56.2004 Tom DEDEURWAERDERE (lxvi): Bioprospection: From the Economics of Contracts to Reflexive Governance
CCMP 57.2004 Katrin REHDANZ and David MADDISON: The Amenity Value of Climate to German Households
CCMP 58.2004 Koen SMEKENS and Bob VAN DER ZWAAN: Environmental Externalities of Geological Carbon Sequestration
Effects on Energy Scenarios
NRM 59.2004 Valentina BOSETTI, Mariaester CASSINELLI and Alessandro LANZA (lxvii): Using Data Envelopment
Analysis to Evaluate Environmentally Conscious Tourism Management
NRM 60.2004 Timo GOESCHL and Danilo CAMARGO IGLIORI (lxvi):Property Rights Conservation and Development: An
Analysis of Extractive Reserves in the Brazilian Amazon
CCMP 61.2004 Barbara BUCHNER and Carlo CARRARO: Economic and Environmental Effectiveness of a
Technology-based Climate Protocol
NRM 62.2004 Elissaios PAPYRAKIS and Reyer GERLAGH: Resource-Abundance and Economic Growth in the U.S.
NRM 63.2004 Györgyi BELA, György PATAKI, Melinda SMALE and Mariann HAJDÚ (lxvi): Conserving Crop Genetic
Resources on Smallholder Farms in Hungary: Institutional Analysis
NRM 64.2004 E.C.M. RUIJGROK and E.E.M. NILLESEN (lxvi): The Socio-Economic Value of Natural Riverbanks in the
Netherlands
NRM 65.2004 E.C.M. RUIJGROK (lxvi): Reducing Acidification: The Benefits of Increased Nature Quality. Investigating the
Possibilities of the Contingent Valuation Method
ETA 66.2004 Giannis VARDAS and Anastasios XEPAPADEAS: Uncertainty Aversion, Robust Control and Asset Holdings
GG 67.2004 Anastasios XEPAPADEAS and Constadina PASSA: Participation in and Compliance with Public Voluntary
Environmental Programs: An Evolutionary Approach
GG 68.2004 Michael FINUS: Modesty Pays: Sometimes!
NRM 69.2004 Trond BJØRNDAL and Ana BRASÃO: The Northern Atlantic Bluefin Tuna Fisheries: Management and Policy
Implications
CTN 70.2004 Alejandro CAPARRÓS, Abdelhakim HAMMOUDI and Tarik TAZDAÏT: On Coalition Formation with
Heterogeneous Agents
IEM 71.2004 Massimo GIOVANNINI, Margherita GRASSO, Alessandro LANZA and Matteo MANERA: Conditional
Correlations in the Returns on Oil Companies Stock Prices and Their Determinants
IEM 72.2004 Alessandro LANZA, Matteo MANERA and Michael MCALEER: Modelling Dynamic Conditional Correlations
in WTI Oil Forward and Futures Returns
SIEV 73.2004 Margarita GENIUS and Elisabetta STRAZZERA: The Copula Approach to Sample Selection Modelling:
An Application to the Recreational Value of Forests
CCMP 74.2004 Rob DELLINK and Ekko van IERLAND: Pollution Abatement in the Netherlands: A Dynamic Applied General
Equilibrium Assessment
ETA 75.2004 Rosella LEVAGGI and Michele MORETTO: Investment in Hospital Care Technology under Different
Purchasing Rules: A Real Option Approach
CTN 76.2004 Salvador BARBERÀ and Matthew O. JACKSON (lxx): On the Weights of Nations: Assigning Voting Weights in
a Heterogeneous Union
CTN 77.2004 Àlex ARENAS, Antonio CABRALES, Albert DÍAZ-GUILERA, Roger GUIMERÀ and Fernando VEGA-
REDONDO (lxx): Optimal Information Transmission in Organizations: Search and Congestion
CTN 78.2004 Francis BLOCH and Armando GOMES (lxx): Contracting with Externalities and Outside Options
CTN 79.2004 Rabah AMIR, Effrosyni DIAMANTOUDI and Licun XUE (lxx): Merger Performance under Uncertain Efficiency
Gains
CTN 80.2004 Francis BLOCH and Matthew O. JACKSON (lxx): The Formation of Networks with Transfers among Players
CTN 81.2004 Daniel DIERMEIER, Hülya ERASLAN and Antonio MERLO (lxx): Bicameralism and Government Formation
CTN 82.2004 Rod GARRATT, James E. PARCO, Cheng-ZHONG QIN and Amnon RAPOPORT (lxx): Potential Maximization
and Coalition Government Formation
CTN 83.2004 Kfir ELIAZ, Debraj RAY and Ronny RAZIN (lxx): Group Decision-Making in the Shadow of Disagreement
CTN 84.2004 Sanjeev GOYAL, Marco van der LEIJ and José Luis MORAGA-GONZÁLEZ (lxx): Economics: An Emerging
Small World?
CTN 85.2004 Edward CARTWRIGHT (lxx): Learning to Play Approximate Nash Equilibria in Games with Many Players
IEM 86.2004 Finn R. FØRSUND and Michael HOEL: Properties of a Non-Competitive Electricity Market Dominated by
Hydroelectric Power
KTHC 87.2004 Elissaios PAPYRAKIS and Reyer GERLAGH: Natural Resources, Investment and Long-Term Income
CCMP 88.2004 Marzio GALEOTTI and Claudia KEMFERT: Interactions between Climate and Trade Policies: A Survey
IEM 89.2004 A. MARKANDYA, S. PEDROSO and D. STREIMIKIENE: Energy Efficiency in Transition Economies: Is There
Convergence Towards the EU Average?
GG 90.2004 Rolf GOLOMBEK and Michael HOEL : Climate Agreements and Technology Policy
PRA 91.2004 Sergei IZMALKOV (lxv): Multi-Unit Open Ascending Price Efficient Auction
KTHC 92.2004 Gianmarco I.P. OTTAVIANO and Giovanni PERI: Cities and Cultures
KTHC 93.2004 Massimo DEL GATTO: Agglomeration, Integration, and Territorial Authority Scale in a System of Trading
Cities. Centralisation versus devolution
CCMP 94.2004 Pierre-André JOUVET, Philippe MICHEL and Gilles ROTILLON: Equilibrium with a Market of Permits
CCMP 95.2004 Bob van der ZWAAN and Reyer GERLAGH: Climate Uncertainty and the Necessity to Transform Global
Energy Supply
CCMP 96.2004 Francesco BOSELLO, Marco LAZZARIN, Roberto ROSON and Richard S.J. TOL: Economy-Wide Estimates of
the Implications of Climate Change: Sea Level Rise
CTN 97.2004 Gustavo BERGANTIÑOS and Juan J. VIDAL-PUGA: Defining Rules in Cost Spanning Tree Problems Through
the Canonical Form
CTN 98.2004 Siddhartha BANDYOPADHYAY and Mandar OAK: Party Formation and Coalitional Bargaining in a Model of
Proportional Representation
GG 99.2004 Hans-Peter WEIKARD, Michael FINUS and Juan-Carlos ALTAMIRANO-CABRERA: The Impact of Surplus
Sharing on the Stability of International Climate Agreements
SIEV 100.2004 Chiara M. TRAVISI and Peter NIJKAMP: Willingness to Pay for Agricultural Environmental Safety: Evidence
from a Survey of Milan, Italy, Residents
SIEV 101.2004 Chiara M. TRAVISI, Raymond J. G. M. FLORAX and Peter NIJKAMP:A Meta-Analysis of the Willingness to
Pay for Reductions in Pesticide Risk Exposure
NRM 102.2004 Valentina BOSETTI and David TOMBERLIN: Real Options Analysis of Fishing Fleet Dynamics: A Test
CCMP 103.2004 Alessandra GORIA e Gretel GAMBARELLI: Economic Evaluation of Climate Change Impacts and Adaptability
in Italy
PRA 104.2004 Massimo FLORIO and Mara GRASSENI: The Missing Shock: The Macroeconomic Impact of British
Privatisation
PRA 105.2004 John BENNETT, Saul ESTRIN, James MAW and Giovanni URGA: Privatisation Methods and Economic Growth
in Transition Economies
PRA 106.2004 Kira BÖRNER: The Political Economy of Privatization: Why Do Governments Want Reforms?
PRA 107.2004 Pehr-Johan NORBÄCK and Lars PERSSON: Privatization and Restructuring in Concentrated Markets
SIEV 108.2004 Angela GRANZOTTO, Fabio PRANOVI, Simone LIBRALATO, Patrizia TORRICELLI and Danilo
MAINARDI: Comparison between Artisanal Fishery and Manila Clam Harvesting in the Venice Lagoon by
Using Ecosystem Indicators: An Ecological Economics Perspective
CTN 109.2004 Somdeb LAHIRI: The Cooperative Theory of Two Sided Matching Problems: A Re-examination of Some
Results
NRM 110.2004 Giuseppe DI VITA: Natural Resources Dynamics: Another Look
SIEV 111.2004 Anna ALBERINI, Alistair HUNT and Anil MARKANDYA: Willingness to Pay to Reduce Mortality Risks:
Evidence from a Three-Country Contingent Valuation Study
KTHC 112.2004 Valeria PAPPONETTI and Dino PINELLI: Scientific Advice to Public Policy-Making
SIEV 113.2004 Paulo A.L.D. NUNES and Laura ONOFRI: The Economics of Warm Glow: A Note on Consumer’s Behavior
and Public Policy Implications
IEM 114.2004 Patrick CAYRADE: Investments in Gas Pipelines and Liquefied Natural Gas Infrastructure What is the Impact
on the Security of Supply?
IEM 115.2004 Valeria COSTANTINI and Francesco GRACCEVA: Oil Security. Short- and Long-Term Policies
IEM 116.2004 Valeria COSTANTINI and Francesco GRACCEVA: Social Costs of Energy Disruptions
IEM 117.2004 Christian EGENHOFER, Kyriakos GIALOGLOU, Giacomo LUCIANI, Maroeska BOOTS, Martin SCHEEPERS,
Valeria COSTANTINI, Francesco GRACCEVA, Anil MARKANDYA and Giorgio VICINI: Market-Based Options
for Security of Energy Supply
IEM 118.2004 David FISK: Transport Energy Security. The Unseen Risk?
IEM 119.2004 Giacomo LUCIANI: Security of Supply for Natural Gas Markets. What is it and What is it not?
IEM 120.2004 L.J. de VRIES and R.A. HAKVOORT: The Question of Generation Adequacy in Liberalised Electricity Markets
KTHC 121.2004 Alberto PETRUCCI: Asset Accumulation, Fertility Choice and Nondegenerate Dynamics in a Small Open
Economy
NRM 122.2004 Carlo GIUPPONI, Jaroslaw MYSIAK and Anita FASSIO: An Integrated Assessment Framework for Water
Resources Management: A DSS Tool and a Pilot Study Application
NRM 123.2004 Margaretha BREIL, Anita FASSIO, Carlo GIUPPONI and Paolo ROSATO: Evaluation of Urban Improvement
on the Islands of the Venice Lagoon: A Spatially-Distributed Hedonic-Hierarchical Approach
ETA 124.2004 Paul MENSINK: Instant Efficient Pollution Abatement Under Non-Linear Taxation and Asymmetric
Information: The Differential Tax Revisited
NRM 125.2004 Mauro FABIANO, Gabriella CAMARSA, Rosanna DURSI, Roberta IVALDI, Valentina MARIN and Francesca
PALMISANI: Integrated Environmental Study for Beach Management:A Methodological Approach
PRA 126.2004 Irena GROSFELD and Iraj HASHI: The Emergence of Large Shareholders in Mass Privatized Firms: Evidence
from Poland and the Czech Republic
CCMP 127.2004 Maria BERRITTELLA, Andrea BIGANO, Roberto ROSON and Richard S.J. TOL: A General Equilibrium
Analysis of Climate Change Impacts on Tourism
CCMP 128.2004 Reyer GERLAGH: A Climate-Change Policy Induced Shift from Innovations in Energy Production to Energy
Savings
NRM 129.2004 Elissaios PAPYRAKIS and Reyer GERLAGH: Natural Resources, Innovation, and Growth
PRA 130.2004 Bernardo BORTOLOTTI and Mara FACCIO: Reluctant Privatization
SIEV 131.2004 Riccardo SCARPA and Mara THIENE: Destination Choice Models for Rock Climbing in the Northeast Alps: A
Latent-Class Approach Based on Intensity of Participation
SIEV 132.2004 Riccardo SCARPA Kenneth G. WILLIS and Melinda ACUTT: Comparing Individual-Specific Benefit Estimates
for Public Goods: Finite Versus Continuous Mixing in Logit Models
IEM 133.2004 Santiago J. RUBIO: On Capturing Oil Rents with a National Excise Tax Revisited
ETA 134.2004 Ascensión ANDINA DÍAZ: Political Competition when Media Create Candidates’ Charisma
SIEV 135.2004 Anna ALBERINI: Robustness of VSL Values from Contingent Valuation Surveys
CCMP 136.2004 Gernot KLEPPER and Sonja PETERSON: Marginal Abatement Cost Curves in General Equilibrium: The
Influence of World Energy Prices
ETA 137.2004 Herbert DAWID, Christophe DEISSENBERG and Pavel ŠEVČIK: Cheap Talk, Gullibility, and Welfare in an
Environmental Taxation Game
CCMP 138.2004 ZhongXiang ZHANG: The World Bank’s Prototype Carbon Fund and China
CCMP 139.2004 Reyer GERLAGH and Marjan W. HOFKES: Time Profile of Climate Change Stabilization Policy
NRM 140.2004 Chiara D’ALPAOS and Michele MORETTO: The Value of Flexibility in the Italian Water Service Sector: A
Real Option Analysis
PRA 141.2004 Patrick BAJARI, Stephanie HOUGHTON and Steven TADELIS (lxxi): Bidding for Incompete Contracts
PRA 142.2004 Susan ATHEY, Jonathan LEVIN and Enrique SEIRA (lxxi): Comparing Open and Sealed Bid Auctions: Theory
and Evidence from Timber Auctions
PRA 143.2004 David GOLDREICH (lxxi): Behavioral Biases of Dealers in U.S. Treasury Auctions
PRA 144.2004 Roberto BURGUET (lxxi): Optimal Procurement Auction for a Buyer with Downward Sloping Demand: More
Simple Economics
PRA 145.2004 Ali HORTACSU and Samita SAREEN (lxxi): Order Flow and the Formation of Dealer Bids: An Analysis of
Information and Strategic Behavior in the Government of Canada Securities Auctions
PRA 146.2004 Victor GINSBURGH, Patrick LEGROS and Nicolas SAHUGUET (lxxi): How to Win Twice at an Auction. On
the Incidence of Commissions in Auction Markets
PRA 147.2004 Claudio MEZZETTI, Aleksandar PEKEČ and Ilia TSETLIN (lxxi): Sequential vs. Single-Round Uniform-Price
Auctions
PRA 148.2004 John ASKER and Estelle CANTILLON (lxxi): Equilibrium of Scoring Auctions
PRA 149.2004 Philip A. HAILE, Han HONG and Matthew SHUM (lxxi): Nonparametric Tests for Common Values in First-
Price Sealed-Bid Auctions
PRA 150.2004 François DEGEORGE, François DERRIEN and Kent L. WOMACK (lxxi): Quid Pro Quo in IPOs: Why
Bookbuilding is Dominating Auctions
CCMP 151.2004 Barbara BUCHNER and Silvia DALL’OLIO: Russia: The Long Road to Ratification. Internal Institution and
Pressure Groups in the Kyoto Protocol’s Adoption Process
CCMP 152.2004 Carlo CARRARO and Marzio GALEOTTI: Does Endogenous Technical Change Make a Difference in Climate
Policy Analysis? A Robustness Exercise with the FEEM-RICE Model
PRA 153.2004 Alejandro M. MANELLI and Daniel R. VINCENT (lxxi): Multidimensional Mechanism Design: Revenue
Maximization and the Multiple-Good Monopoly
ETA 154.2004 Nicola ACOCELLA, Giovanni Di BARTOLOMEO and Wilfried PAUWELS: Is there any Scope for Corporatism
in Stabilization Policies?
CTN 155.2004 Johan EYCKMANS and Michael FINUS: An Almost Ideal Sharing Scheme for Coalition Games with
Externalities
CCMP 156.2004 Cesare DOSI and Michele MORETTO: Environmental Innovation, War of Attrition and Investment Grants
CCMP 157.2004 Valentina BOSETTI, Marzio GALEOTTI and Alessandro LANZA: How Consistent are Alternative Short-Term
Climate Policies with Long-Term Goals?
ETA 158.2004 Y. Hossein FARZIN and Ken-Ichi AKAO: Non-pecuniary Value of Employment and Individual
Labor Supply
ETA 159.2004 William BROCK and Anastasios XEPAPADEAS: Spatial Analysis: Development of Descriptive and Normative
Methods with Applications to Economic-Ecological Modelling
KTHC 160.2004 Alberto PETRUCCI: On the Incidence of a Tax on PureRent with Infinite Horizons
IEM 161.2004 Xavier LABANDEIRA, José M. LABEAGA and Miguel RODRÍGUEZ: Microsimulating the Effects of Household
Energy Price Changes in Spain
NOTE DI LAVORO PUBLISHED IN 2005
CCMP 1.2005 Stéphane HALLEGATTE: Accounting for Extreme Events in the Economic Assessment of Climate Change
CCMP 2.2005 Qiang WU and Paulo Augusto NUNES: Application of Technological Control Measures on Vehicle Pollution: A
Cost-Benefit Analysis in China
CCMP 3.2005 Andrea BIGANO, Jacqueline M. HAMILTON, Maren LAU, Richard S.J. TOL and Yuan ZHOU: A Global
Database of Domestic and International Tourist Numbers at National and Subnational Level
CCMP 4.2005 Andrea BIGANO, Jacqueline M. HAMILTON and Richard S.J. TOL: The Impact of Climate on Holiday
Destination Choice
ETA 5.2005 Hubert KEMPF: Is Inequality Harmful for the Environment in a Growing Economy?
CCMP 6.2005 Valentina BOSETTI, Carlo CARRARO and Marzio GALEOTTI: The Dynamics of Carbon and Energy Intensity
in a Model of Endogenous Technical Change
IEM 7.2005 David CALEF and Robert GOBLE: The Allure of Technology: How France and California Promoted Electric
Vehicles to Reduce Urban Air Pollution
ETA 8.2005 Lorenzo PELLEGRINI and Reyer GERLAGH: An Empirical Contribution to the Debate on Corruption
Democracy and Environmental Policy
CCMP 9.2005 Angelo ANTOCI: Environmental Resources Depletion and Interplay Between Negative and Positive Externalities
in a Growth Model
CTN 10.2005 Frédéric DEROIAN: Cost-Reducing Alliances and Local Spillovers
NRM 11.2005 Francesco SINDICO: The GMO Dispute before the WTO: Legal Implications for the Trade and Environment
Debate
KTHC 12.2005 Carla MASSIDDA: Estimating the New Keynesian Phillips Curve for Italian Manufacturing Sectors
KTHC 13.2005 Michele MORETTO and Gianpaolo ROSSINI: Start-up Entry Strategies: Employer vs. Nonemployer firms
PRCG 14.2005 Clara GRAZIANO and Annalisa LUPORINI: Ownership Concentration, Monitoring and Optimal Board
Structure
CSRM 15.2005 Parashar KULKARNI: Use of Ecolabels in Promoting Exports from Developing Countries to Developed
Countries: Lessons from the Indian LeatherFootwear Industry
KTHC 16.2005 Adriana DI LIBERTO, Roberto MURA and Francesco PIGLIARU: How to Measure the Unobservable: A Panel
Technique for the Analysis of TFP Convergence
KTHC 17.2005 Alireza NAGHAVI: Asymmetric Labor Markets, Southern Wages, and the Location of Firms
KTHC 18.2005 Alireza NAGHAVI: Strategic Intellectual Property Rights Policy and North-South Technology Transfer
KTHC 19.2005 Mombert HOPPE: Technology Transfer Through Trade
(lxv) This paper was presented at the EuroConference on “Auctions and Market Design: Theory,
Evidence and Applications” organised by Fondazione Eni Enrico Mattei and sponsored by the EU,
Milan, September 25-27, 2003
(lxvi) This paper has been presented at the 4th BioEcon Workshop on “Economic Analysis of
Policies for Biodiversity Conservation” organised on behalf of the BIOECON Network by
Fondazione Eni Enrico Mattei, Venice International University (VIU) and University College
London (UCL) , Venice, August 28-29, 2003
(lxvii) This paper has been presented at the international conference on “Tourism and Sustainable
Economic Development – Macro and Micro Economic Issues” jointly organised by CRENoS
(Università di Cagliari e Sassari, Italy) and Fondazione Eni Enrico Mattei, and supported by the
World Bank, Sardinia, September 19-20, 2003
(lxviii) This paper was presented at the ENGIME Workshop on “Governance and Policies in
Multicultural Cities”, Rome, June 5-6, 2003
(lxix) This paper was presented at the Fourth EEP Plenary Workshop and EEP Conference “The
Future of Climate Policy”, Cagliari, Italy, 27-28 March 2003
(lxx) This paper was presented at the 9th Coalition Theory Workshop on "Collective Decisions and
Institutional Design" organised by the Universitat Autònoma de Barcelona and held in Barcelona,
Spain, January 30-31, 2004
(lxxi) This paper was presented at the EuroConference on “Auctions and Market Design: Theory,
Evidence and Applications”, organised by Fondazione Eni Enrico Mattei and Consip and sponsored
by the EU, Rome, September 23-25, 2004
2004 SERIES
CCMP Climate Change Modelling and Policy (Editor: Marzio Galeotti )
GG Global Governance (Editor: Carlo Carraro)
SIEV Sustainability Indicators and Environmental Valuation (Editor: Anna Alberini)
NRM Natural Resources Management (Editor: Carlo Giupponi)
KTHC Knowledge, Technology, Human Capital (Editor: Gianmarco Ottaviano)
IEM International Energy Markets (Editor: Anil Markandya)
CSRM Corporate Social Responsibility and Sustainable Management (Editor: Sabina Ratti)
PRA Privatisation, Regulation, Antitrust (Editor: Bernardo Bortolotti)
ETA Economic Theory and Applications (Editor: Carlo Carraro)
CTN Coalition Theory Network
2005 SERIES
CCMP Climate Change Modelling and Policy (Editor: Marzio Galeotti )
SIEV Sustainability Indicators and Environmental Valuation (Editor: Anna Alberini)
NRM Natural Resources Management (Editor: Carlo Giupponi)
KTHC Knowledge, Technology, Human Capital (Editor: Gianmarco Ottaviano)
IEM International Energy Markets (Editor: Anil Markandya)
CSRM Corporate Social Responsibility and Sustainable Management (Editor: Sabina Ratti)
PRCG Privatisation Regulation Corporate Governance (Editor: Bernardo Bortolotti)
ETA Economic Theory and Applications (Editor: Carlo Carraro)
CTN Coalition Theory Network
... As argued in Mukherjee (2016;, trade liberalization helps in productivity gains for the host country via a reduction in tariff, access to better quality goods, increased efficiency of production process, and the innovation capacity of domestic firms through international spillovers of R&D (research & development). Similarly, firm level empirical evidence about productivity gains has also been extensively researched with conflicting implications (Hoppe, 2005). Moreover, empirical studies demonstrate that there is a sectoral bias in productivity gains and industries differ in their ability to exploit prevailing technological trajectories or problem-solving processes (Nelson and Winter, 1977;Dosi, 1982). ...
... (iv) Lajeri and Nielsen (2000) and Eichner and Wagener (2003;2005) have demonstrated that Kimball's (1990) notion of "decreasing absolute prudence" is tantamount to the decreasing slope of the curve for negative of the marginal utility from the business regarding mean (i.e., ) in expected return -∂ ( , ) ∂ for the -models. ...
Article
We build a two-moment decision-theoretic framework to study how firms in the food-processing industry negotiate between risk and return while relying on imported inputs for production at an intensive margin. Two possibilities emerge: either a co-movement or a trade-off in risk and return under various industry and economic conditions. Building on our theoretical setting, we design a testable empirical framework that considers a panel of 316 firms in the Indian food-processing industry between 1993-2009. We find strong evidence of a decrease in the absolute risk aversion preference, although the magnitude varies measurably across firms.
... Very often, international trade agreements include technology transfer cooperation, technical assistance, and joint R&D projects (Martínez-Zarzoso and Chelala 2021). The trade openness that focusses on imports of capital goods and openness to export markets promotes the learning abilities of the trading countries, hence promoting technology transfer that often is measured in terms of total factor productivity (Hoppe 2005). The Trade Openness (TO) is measured by the summation of exports and imports as a ratio of GDP. ...
... Ayrıca anlaşmanın ihlal edilmesi durumunda alınacak tebdirler cezalar fikri mülkiyet hakları çerçevesinde ortaya konulmuştur. Hoppe (2005) teknoloji açıklığı konusuna ticaret ile teknoloji arasındaki ilişki bakış açısıyla bakmıştır. Sermaye mallarının ithalatı ve ihracatının pazarların açık olmasını sağladığını ve bu sayede yaparak öğrenme sağlanarak toplam faktör verimliliğinin arttırıldığını belirtmiştir. ...
Article
Ekonomik Entegrasyon Anlaşmalarının Yüksek Teknolojili Ürün İhracatına olan Etkisinin Türkiye özelinde Avrupa Birliği ile gerçekleştirdiği Gümrük Birliği'nin Etkilerinin Analizini İçermektedir.
... An excellent summary of the literature on indirect spillovers can be found in Hoppe (2005), who identifies three main factors that determine technology transfer: direct efforts to make the transfer successful, the capacity to adopt new technologies, and the differences between the trading countries. The author concludes that trade enables technology transfer mainly through imports of capital goods and openness to export markets, which enable learning-by-doing, thus increasing total factor productivity (TFP). ...
Article
Full-text available
This is the first paper that analyzes for a global sample of countries how trade agreements that include technology-related provisions impact exports of goods, and how this impact differs depending on the technology content of the goods. It includes estimations of a structural gravity model for a panel of 176 countries over the period 1995–2015. The model differentiates between provisions relating technology transfer, technical cooperation, research and development, and patents and intellectual property rights. It also estimates the differences in these effects depending on whether the trade flow in question is between countries with similar or different levels of development. The main results indicate that regional trade agreements (RTAs) that contain technology provisions generate a significantly higher volume of trade than RTAs that do not, after controlling for the depth of the RTAs. For countries that ratify RTAs that include such provisions, it is exports of technology-intensive goods that increase the most. Trade agreements including such provisions have a heterogeneous effect that varies by income level of the trading partners and depends on the extent to which the RTA incorporates other provisions.
... 2-Local labor force employment Employment Will to employ locally trained individuals to benefit from knowledge transfer with employees relocated. 3-Collaborations with local firms Collaboration Research of collaborations/partnerships to benefit from local knowledge 9. Govindarajan & Ramamurti, 2011;Govindarajan & Trimble, 2012;Agarwal, Gupta, & Dayal, 2007;Belderbos, Van Roy, & Duvivier, 2013;Blomström & Kokko, 1998;Hoppe, 2005;Javorcik & Spatareanu, 2005;Keller, 2004;Kokko et al., 1996;Maskus, 2003;Vaidyanathan, 2008;Wang, 2005;Z. Wei & Youssef, 2012;Young & Lan, 1997 States ( Hughes, 2010b). ...
Article
The recent phenomenon of multinational companies localizing their R&D centers into emerging countries requires a new analytical framework. This phenomenon pushes the knowledge frontier beyond its traditional limits and creates a paradigm change in terms of innovation and technology transfer. Our study aims to enrich the R&D internationalization literature in a twofold manner. Firstly, while we confirm the globalization trend of knowledge sources as already observed in previous studies, we show that multinational companies might now choose emergent countries as a strategic place to externalize their R&D. Secondly, we empirically illustrate the phenomena of reverse innovation and reverse technology transfer in the pharmaceutical sector. Although these phenomena are still difficult to quantify, our analytical framework highlights this paradigm change in this global industry, even in a sector where intellectual property is very sensitive. This confirms that sources of knowledge are now more and more sought in emerging countries (in particular China) because of the necessity to fulfill local needs and to conquer these huge markets.
Article
Full-text available
Trade has a huge influence on the economic growth of a country, due to globalization no country could fulfil their national needs and desire without international trade. The aims of this study are to measuring and analyzing the consequences of the United States and China trade war on the United States and China economic growth and global economic growth. To achieve this objective, the study employed Autoregressive Distributed Lag (ARDL) model depending on monthly data from 2016 M1 to 2019 M12. The data were obtained from Federal Reserve Economic Database (FRED) and The United States Census Bureau (USCB), Economic indicators. The result of the study shows that: Dummy variable (D) indicates the tariff rate has a positive and statistically significant effect on United States GDP. However, the effect of tariff rate (D) on the Chines GDP is negative and statistically significant. Likewise, tariff rate(D) harms global GDP, thus if tariff rate (D) changes by 1%, the global GDP decreases by (0.005). In conclusion, increasing tariff rates from both countries (the U.S and China) has negative consequences on their economic growth and the global GDP.
Article
This paper examines the impact of agricultural trade liberalization on agricultural total factor productivity (TFP) growth in Africa using panel data for 13 countries from 2005 to 2016. Our contribution is two-fold. Firstly, we analyse the impact of domestic agriculture support in the spirit of the Agreement on Agriculture. Secondly, we draw attention to the South–South versus South–North debate to the agriculture sector. We examine the impact of trade by source, split between trade within and outside Africa. We compute TFP growth for maize and rice using the Malmquist-data envelopment analysis approach. We then use the dynamic fixed effects approach to estimate panel auto-regressive-distributed-lag models. TFP computations show falling growth rates for both maize and rice. Evidence suggests that domestic agriculture support measures have positive output effects but negative productivity effects. We find that reducing trade-distorting agriculture support coupled with good governance significantly increases TFP growth. Accordingly, we appeal that domestic agriculture support is refocused from producer payments to infrastructure development. Furthermore, we document that South–South trade productivity gains match and can surpass South-North Trade. Hence we emphasize increasing intra-Africa agriculture trade.
Article
In recent years, though total factor productivity (TFP) convergence phenomenon has gained tremendous importance yet further deliberations for identification of catalytic factors that can help developing countries to achieve their steady developmental paths, are under way. Against this backdrop, present study investigates the principal determinants of TFP convergence by employing data of 91 developing countries over the period 1960–2015 and with USA being the frontier country. In concordance with the existing literature, main focus remains on technology diffusion for the catch-up process and is measured by means of trade openness (TO) and foreign direct investment (FDI) with introduction of their interaction terms. However, TFP is computed by incorporating the Growth Accounting Model while empirical results are drawn from the 2-step GMM estimation technique. It is surfaced that though high degree of openness benefits TFP growth and convergence but FDI has a dominating role. Therefore, governments can play a competent role via unflagging efforts in ensuring that the right kind of policies are enacted, promoting trading activities and FDI flows.
Article
Empirical analyses of knowledge spillovers from foreign direct investment (FDI) offer mixed results; they find positive, neutral and negative FDI spillover effects. This lack of evidence mainly comes from the results of firm-level panel data analysis. This is important since this approach seems to be the most appropriate for estimating FDI spillovers. The paper takes a look at recent substantive and methodological developments in FDI spillover analysis, which have brought some more optimistic results with regard to FDI spillovers, and can help in further development in this field. The main substantive development relates to the introduction of a broad variety of sources of firm heterogeneity (foreign affiliates as well as local firms) in the analysis. Others include differentiation between vertical (inter-industry) and horizontal (intra-industry) spillovers, and host country absorptive capacity for knowledge spillovers. Methodological developments relate to distinguishing between technological/knowledge and productivity spillovers, improvement of modelling and estimation methods, and an increased amount and quality of data.
Chapter
According to new growth theory, technological progress is endogenous and driven by an intentional investment of resources by profit-seeking firms. Still, innovation activities in firms depend heavily on external sources (Fagerberg, 2005). For most countries foreign sources of technology are of dominant importance for productivity growth (Eaton and Kortum, 1999; Keller, 2002). Therefore, economic analysis of innovation recognizes international knowledge flows (through FDI, trade, licensing and international technological collaborations) as important determinants of the development and diffusion of innovations. Here, the notion of technology and knowledge spillovers is central. It is based on theories of endogenous technical change of the early 1990s (Romer, 1990; Grossman and Helpman, 1991; Aghion and Howitt, 1998), claiming that the return to technological investments is partly private and partly public (Keller, 2004). Because of the non-rival character of technology, an innovation that is produced by one firm may also be used by another firm, without incurring very much additional cost (Smolny, 2000). These are technology or knowledge spillovers.