Content uploaded by Chee Lip Tee
Author content
All content in this area was uploaded by Chee Lip Tee on Feb 07, 2018
Content may be subject to copyright.
Int. Journal of Economics and Management 9 (S): 41 - 60 (2015)
IJEM
International Journal of Economics and Management
Journal homepage: http://www.econ.upm.edu.my/ijem
R&D Spillovers and the Role of Economic Freedom
TEE CHEE-LIP, W.N.W. AZMAN-SAINI*, SAIFUZZAMAN IBRAHIM
AND NORMAZ WANA ISMAIL
Faculty of Economics and Management, Universiti Putra Malaysia,
43400 UPM Serdang, Selangor
ABSTRACT
This paper examines the role of economic freedom plays in moderating
research and development (R&D) spillovers from developed to
developing countries. Two channels are analyzed namely import and
international student ows. The empirical results based on generalized
method-of-moment system estimation on a panel of 75 developing
countries show that spillover effects through import and international
student ow are signicant, but the latter channel appear to be more
important in term of magnitude. This nding is consistent with view
that technology diffusion via human capital mobility should not be
underestimated. More importantly, the nding reveals that countries
with higher level of economic freedom benet more from R&D
spillovers. This provides further support to the idea that successful
knowledge acquisition requires absorptive capacity.
Keywords: economic freedom, international student ows, R&D
spillovers, total factor productivity, trade
INTRODUCTION
Many economists believe that technological progress is an important determinant
for long-run output growth because it is very fundamental to the economy and
affects all areas of economic activities (Le, 2012). The new growth models (see
* Corresponding Author: E-mail: wazman@upm.edu.my
Any remaining errors or omissions rest solely with the author(s) of this paper.
42
International Journal of Economics and Management
for example, Romer, 1990; Grossman and Helpman, 1991; Aghion and Howitt,
1992) suggest that technological progress is not a free gift from heaven but a direct
outcome of innovation process. This is in contrast to the neo-classical model which
treats technological progress as exogenous. According to the new growth model,
investments in innovation activities would allow country to enjoy technological
progress and greater productivity which ultimately lead to the expansion of the
economy.
Since the pioneering work of Coe and Helpman (1995), many studies have
recognized the importance of international research and development (R&D)
spillovers. Due to the non-rival characteristics of technology, R&D investment
would contribute to the stock of knowledge as it is publicly available to everyone.
Hence, R&D of one country does not only affect domestic rms but also foreign
rms. This suggests that countries which hardly invest in R&D activities would
benet from new knowledge developed by R&D leaders. The theory suggests
that the extent to which local rms can benet from foreign knowledge depends
on many factors such as trade volume (Coe and Helpman, 1995), characteristics
of traded products (Coe, Helpman, and Hoffmaister, 1997), ow of foreign direct
investment (FDI) (van Pottelsberghe and Lichtenberg, 2001), and human capital
mobility (Park, 2004).
Among the factors highlighted above, human capital mobility is a newly
established channel for knowledge spillovers across borders. It is argued that some
knowledge is difcult to be expressed in words or language (Koskinen, Pihlanto
and Vanharanta, 2003) and therefore exchange of goods or investment for spillovers
will not help its diffusion across borders (Lee, 2005). Instead, spillovers of this
type of knowledge require direct communication. Therefore, international students
ow is viewed as conduit for knowledge transmission because students are able to
absorb foreign knowledge when they study abroad or through post schooling job
experience and transfer it back to domestic country when they return (Park, 2004).
R&D via students ow has been hardly investigated. Two exceptions are
Park (2004) and Le (2010). Park (2004) shows that international student ow
is an important spillover channel among developed countries while Le (2010)
complements the nding for spillovers from developed to developing countries.
However, they found that spillover effects through import are relatively stronger
than student ow. Recent literatures show that globalization has led to improved
communication and mobility across border, and this therefore suggests that
disembodied spillovers channel (such as international student ow) today could
be at least as important as embodied channel in past decades (Filatotchev, Liu, Lu
and Wright, 2011). Hence, a study on recent period could lead to different ndings
on the relative importance of various spillover channels.
43
R&D Spillovers and the Role of Economic Freedom
Several recent papers suggest that knowledge spillovers are not automatic
consequences of direct or indirect contact with R&D leaders. They argue that host
countries must have certain quality which allows them to absorb and internalize
the technology generated abroad. For instance, Azman-Saini, et al., (2010) show
that only countries with sufcient freedom of economic activities are able to absorb
and internalize new technologies associated with FDI inow. In an economically
freer environment, rm are more willing to engage in risky investment project,
such as trying out news ideas and new technologies, it will motivates domestic
rm to absorb foreign technology in local market.
The purpose of this paper is therefore to evaluate the role of economic freedom
plays in moderating R&D spillovers from developed to developing countries. Two
channels are analyzed namely import and international student ows. In the face of
increasing globalization, understanding the effective channels of R&D spillovers
across countries is critically important. Evidence on international technology
spillover is equally important for both innovation leader and follower. In the case
of innovation leader, knowing how knowledge is transmitted across countries is
necessary in order to protect the interests of innovators. For the followers, evidence
of spillover effects will provide additional incentives for them to further integrate
with the rest of the world. Greater openness is expected to provide countries with a
better atmosphere for technology acquisition. To achieve this objective, data from
75 developing countries over the 2000-2008 period and a generalized method of
moment (GMM) panel estimator are used. This estimator has several advantages
over other alternatives.
This paper lls the gap in the literature in several important ways. First, most
of the previous studies have mainly focused on embodied channel for spillovers
such as import and FDI. On top of embodied channel (i.e. import), this paper also
assess the importance of disembodied channel in R&D spillovers. Specically, it
examines the role of international student ows. Secondly, prior researcher on R&D
spillovers did not account for the role of absorptive capacity in mediating R&D
spillovers. Several recent papers suggest that knowledge acquisition requires the
receiving nations to have some level of absorptive capacity. Therefore, we argue
that absorptive capacity is able to amplify the effects of R&D spillovers. In this
study, we evaluate the role economic freedom plays in enhancing R&D spillovers
from industrial countries to a group of developing countries.
The rest of the paper is structured as follows. The next section presents a
brief literature review. Then, section on research methodology outlines model
specication, methodology and data. Subsequence section discusses empirical
results and their interpretation. Conclusion is reported in the nal section.
44
International Journal of Economics and Management
REVIEW OF LITERATURE
Research and development (R&D) is considered as a major source of technological
progress. According to OECD (2003), R&D can be dened as a ‘creative work
undertaken on a systematic basis in order to increase the stock of knowledge,
including knowledge of man, culture and society, and the use of this stock of
knowledge to devise new applications’. This suggests that R&D is a process of
transforming R&D inputs into R&D outputs which materialize in the forms of
increments to the stock of knowledge and new technologies (i.e. applications of
existing knowledge).
Much of earlier policy debate about technology spillover is based on the
presumption that a country’s productivity depend on domestic investment in R&D.
In line with this emphasis, earlier empirical works on R&D spillover have focused
on the impact of domestic R&D activities on growth (e.g. Griliches, 1988, 1992;
Nadiri, 1993; Mohnen, 1996). Generally, these studies provide convincing evidence
that cumulative domestic R&D is an important determinant of productivity. Indeed,
they nd that the rate of return on R&D investment is high.
However, with a rapid pace of globalization, productivity growth of a country
does not depend only on domestic R&D, but also foreign R&D through interaction
with foreign economies. As a result, a more recent stream of empirical literature
focuses on international R&D spillover1. The pioneering work of Coe and Helpman
(1995) (henceforth, CH) assessed R&D spillover across 21 OECD countries plus
Israel and demonstrate an empirical relationship between R&D expenditures and
total factor productivity (TFP). They nd that not only domestic R&D contributes
signicantly to productivity growth but also foreign R&D incorporated into trade
ows. Trade can boost domestic productivity by making available product that
embodies technological knowledge of trading partners. By enabling a country to
employ larger variety of intermediate product and capital equipment, trade enhances
the productivity of resources. Trade also improves domestic productivity by making
available useful information that would otherwise be costly to acquire.
A number of papers have made progress by examining international R&D
spillover further. Xu and Wang (1999) emphasize that technology diffusion is
associated with trade in differentiated capital goods. They decompose total imports
into the imports of capital and non-capital goods. They nd that about half of the
return on R&D investment in a G7 country spilled over to other OECD countries
and trade in capital goods was found to be a signicant channel for R&D spillovers.
Lumenga-Neso et al. (2005) argue that `indirect’ trade-related R&D spillovers also
1 Keller (2004) provides an in-depth survey of the existing evaluations of international R&D spillovers.
45
R&D Spillovers and the Role of Economic Freedom
take place between countries, even if they do not trade with each other. Country
A may benet from country B’s technology without importing from country B,
if country B exports to country C which in turn exports to country A. Using a
specication that captures such indirect effect of R&D spillovers, they provide
better empirical results than CH. They nd that the `indirect’ trade-related R&D
spillovers are on average 14 times as large as `direct’ spillovers.
The above studies consider international trade as the only channel for
international R&D spillover. They are likely to have underestimated the relative
magnitude of international spillover effects that pass through other channels. Over
the past few decades, foreign direct investment (FDI) by multinational corporations
(MNCs) has grown substantially. The growth rate of world FDI has exceeded
the growth rates of both world trade and GDP (UNCTAD, 2001). FDI has been
an important channel for transferring goods and services across borders (Saggi,
2002). Since MNCs responsible for a large share of global R&D expenditure
(Borensztein et al., 1998), FDI by MNCs could be a potential channel to access
advanced technologies available in the global marketplace. Van Pottelsberghe and
Lichtenberg (2001) (henceforth, PL) extend CH’s analysis by incorporating both
inward and outward FDI ows in addition to the trade ow. Due to limited bilateral
FDI data, PL analyze only 13 out of 22 countries covered in CH’s study. They
nd that foreign R&D can affect domestic productivity through both imports and
outward FDI (i.e. technology sourcing). Although both inward and outward FDI
may facilitate technology acquisition, outward FDI is a more effective channel as it
is more likely to involve ‘total immersion’. By setting up production and research
facilities in countries that have accumulated substantial scientic and technological
capabilities, technology follower can have better access to leading technology.
The nding of technology sourcing practices is consistent with Dunning (1994)
paradigm that companies prefer to invest abroad in order to take advantage of their
own technology base instead of diffusing it internationally. The pioneering works
on technology sourcing by Kogut and Chang (1991) nd that Japanese rms tend
to acquire local U.S. rms when they suffer from technological or comparative
disadvantage but choose to establish new plants when they poses technological
comparative advantages as compared to their U.S. competitors. Evidence on FDI
as a spillover channel has inspired several recent studies in this area (e.g. Bitzer
and Kerekes (2008); Zhu and Jeon (2007); Savvides and Zachariadis, 2005).
Several recent papers suggest that some knowledge do not require exchange
of goods or investment to be transferred (Lee, 2005). They also highlighted that
social engagement like face-to-face interactions would reinforces the knowledge
sharing (Koskinen, Pihlanto and Vanharanta, 2003). Therefore, the relationship
and social connection established between two parties, such as publication, public
46
International Journal of Economics and Management
meeting and conference, information exchange, competitor’s products, patent and
telecommunication is crucial (Almeida and Kogut, 1999, Cohen, Goto, Nagata,
Nelson and Walsh, 2002; Tang and Koveos, 2008). Moreover, Kim and Lee (2004)
argue that embodied technology diffusion (via import and FDI) has a larger impact
on efciency while disembodied technology diffusion affects technical change. With
regard to this issue, direct communication or “disembodied channel” like human
capital is an effective way to transfer knowledge (Song, Almeida and Wu, 2003).
Park (2004) suggests that international student ow is an important mechanism
for technology transfer because students who study abroad would acquire external
knowledge through education or post schooling job experience, and then bring the
knowledge back to home country when they return. International students also
learn the foreign country’s knowledge of technology, material, production method
and organizational structure (Le, 2010). In additional, returnees own the specic
human capital and social capital, therefore act as a bridge between source and host
countries and accelerate the knowledge transfer (Filatotchev et al., 2011). Though
not every international student would returns, migrated workers would still benet
their home country. Foreign workers usually maintain a close connection with their
home country and able to contribute in home countries’ productivity with technology
learned from host country (Le, 2008). Empirical evidence on biotechnology industry
(Zucker, Darby and Brewer, 1998) and semiconductor industry (Almeida and
Kogut, 1999) in the U.S. market show that the mobility of specialists across rms
is found to be one of the major determinants for knowledge transfer. Generally,
human capital mobility was shown to be an important mechanism for knowledge
diffusion and spillovers could be absent without it.
Although research on international R&D spillover has been growing, it
remains limited particularly with respect to R&D spillovers from developed to less
developed countries. It is well known that much of the R&D activity in the world
is concentrated in the industrialized countries. In fact, within the OECD three key
players in R&D activity (i.e. United States, Japan and Germany) accounted for 67%
of R&D expenditure in 2005. This raise concern of whether less developed countries
can benet from high concentration of R&D activity in a handful of developed
countries. The nding of R&D spillovers may have important implications for
less developed countries that lag behind technology frontier and hardly invest in
R&D activities. Analysis on R&D spillover from developed (North) to developing
countries (South) was pioneered by Coe et al. (1997). Following similar approach
as CH, they estimate the elasticity of TFP in 77 developing countries with respect
to R&D stock in developed countries and nd that the R&D spillover from North
to South is substantial. On average, 1% increase in R&D capital stock in developed
countries contributes to 0.06% increase in productivity of developing countries.
47
R&D Spillovers and the Role of Economic Freedom
Among the developed countries, United States is the largest contributor to the
productivity of developing countries owing to its large trade share with developing
countries and also because of its huge R&D capital stock compared to other
developed countries. Due to data limitation, Coe et al. (1997) ignore domestic R&D
capital stock in their analysis.2 Several recent papers that assess North-South R&D
spillovers include Madden et al. (2001), Kwark and Shin (2006), Le (2010), Tang
and Koveos (2008), Le (2012).
In other related development, recent studies show that knowledge diffusion
is not an automatic process. Instead, it requires knowledge recipients to have
certain level of absorptive capacity.3 Specically, the knowledge spillovers may
not be strong in countries with poor absorptive capacity. A number of papers have
tested the absorptive capacity hypothesis in the FDI-growth context. For instance,
Blomstrom et al. (1994) reveal that the growth-effect of FDIs is stronger in
countries with a higher level of development (i.e., when the country is sufciently
rich in terms of per capita income). Meanwhile, Borensztein et al. (1998) found
that the positive impact of FDI on output growth certain level of human capital
to be available in the host countries. Recently, several authors have assessed the
impact of nancial sector development on FDI spillovers (Hermes and Lensink,
2003; Alfaro et al.,2004, 2010; and Durham, 2004). They nd that the success of
technology spillovers from MNCs to local rms required well-functioning nancial
institutions. The development of both banks and stock markets were found to be
important pre-conditions for FDI spillovers. Recently, Azman-Saini et al. (2010)
show that knowledge spillovers via FDI require that host countries to have certain
level economic freedom. The authors argue that the lack of economic freedom can
limit a rm’s (or nation’s) ability to absorb and internalize new technology from
multinational corporations.
RESEARCH METHODOLOGY
This study uses a generalized version of the model employed by Coe and Helpman
(1995), as modified by Lichtenberg and van Pottelsberghe (1998) and van
Pottelsberghe and Lichtenberg (2001). Equation (1) provides the basic econometric
2 Due to underdeveloped nancial market or inappropriate policy, developing countries usually have
limited R&D investment (Grifth, Redding and Reenen, 2003).
3 Cohen and Levinthal (1990) dene absorptive capacity as a rm’s “ability to recognize the value of
new information, assimilate it, and apply it to commercial ends.” This concept differs from learning-
by-doing, which is the automatic process by which rms become more experienced, and hence, more
efcient at current practices. In contrast, with absorptive capacity rms may acquire new knowledge
developed by others that will enable them to do something in different ways.
48
International Journal of Economics and Management
model. It states that the domestic total factor productivity of a country is a function
of different types of foreign R&D capital stocks: 4
TFP = f (SM, SSF) (1)
where TFP is total factor productivity, SM and SSF are respectively import-weighted
and student ow-weighted foreign R&D capital stocks.
TFP measurement used in this paper is different from those in many of the
previous studies. This paper follows a suggestion by Klenow and Rodrigues Clare
(1997) and Hall and Jones (1999) who use human capital augmented labor instead
of only labor. This approach, therefore, also consider the quality of labor. To
highlight the computation of total factor productivity (A), let assume the following
production function:
Y = AK α H 1–α (2)
where Y is output, K is capital stock, α is share of capital income in GDP. Capital
stocks are computed using gross xed capital formation following the perpetual
inventory method (PIM) and H is augmented labor based on Mincerian’s function:
H = exp φ(E) L (3)
where the labor, L, is assumed to be homogenous and each is trained with E years
of schooling.
Equation (3) shows that the labor force is multiplied by efciency, E, which
represents years of schooling and derivative φ’(E) is the return to education where
labor force with no schooling is φ(0) = 0. Years of experience and sum of human
capital with different education and experience level are found to have only little
effect (Klenow and Rodriguez Clare, 1997) and therefore are not used in this paper.
Additionally, following Hall and Jones (1999) several adjustments are made. First,
output measure is adjusted for natural resource so that the countries would not
be ranked as top productivity country due its rich resource. Thus, value added in
the mining industry will be subtracted from GDP. Second, α is set to as standard
neoclassical approach suggests. Third, φ(E) is assumed to be piecewise linear.
The rate of return of education is 13.4 percent for the rst four years (average
of sub Saharan Africa), 10.1 percent for the next four years (average of world),
and 6.8 percent for more than eight years (average of OECD). These gures were
4 Most of the studies which focus on R&D spillovers among developed countries have also included
FDI-weighted foreign R&D stock and domestic R&D stocks in their model. This study focuses on
North-South spillovers and due to the unavailability of data on domestic R&D and FDI for many
developing countries; this study uses a more simplied model as above.
49
R&D Spillovers and the Role of Economic Freedom
suggested by Psacharopoulos (1994) based on survey on return to schooling from
many countries.
Following Le (2010), student ow-embodied capital stocks are computed as
follows:
Sfs n
s
Sd
it jt
ijt
jt
=
cm
/ (4)
where sij is the number of tertiary students originating from country i and studying at
country j, nj is the total number of tertiary students enrolled in country j. Sdj is total
domestic R&D stock in country j. The weight reects the concept where country
i benets from country j’s R&D investments depend on the degree of access by
students from country i to knowledge available in country j.
The import embodied foreign R&D capital stock (Sfmit) is constructed following
van Pottelsberghe and Lichtenberg’s (2001) method as follows:
Sf
m
Sdm y
it jt
ijt
jt
=
cm
/ (5)
where mij is the value of imported goods and services of country i from country j. It
might be interpreted as embodied with R&D intensity of source country (country j),
y is gross domestic product of country j, Sdj is total domestic R&D stock in country j.
This study include as many developing countries as possible but due to data
limitation, only annual data series from 75 developing countries over the 2000-2008
periods are used.5 Data used to compute TFP (i.e. GDP, gross xed capital formation,
labor force) were obtained from the World Development Indicators database except
for human capital which uses average education year for age above 25 as reported
in Barro and Lee (2010). Foreign R&D stocks were constructed based on R&D
spending by G7 (Canada, France, Germany, Italy, Japan, the United Kingdom,
and the United States) and the data were collected from the OECD Main Science
and Technology Indicators database. Bilateral data for import was obtained from
the United Nations Commodity Trade (UN Comtrade) database. The information
on contribution of mining activity to total value added was obtained from the
United Nations Statistics Division National Accounts Main Aggregates Database.
Finally, total number of students enrolled in tertiary level education and number
of international students enrolled were collected from the OECD Education and
Training Database. The economic freedom index was obtained from the Annual
Report of Economic Freedom of the World published by the Fraser Institute.
5 This sample period is dictated by the availability of data on student ows.
50
International Journal of Economics and Management
This paper applies the generalized method-of-moments (GMMs) panel
estimator which was rst proposed by Holtz-Eakin, Newey and Rosen (1988)
and then extended by Arellano and Bond (1991), Arellano and Bover (1995),
and Blundell and Bond (1998). One of the reasons for choosing GMM estimator
is the need to address country-specic effect. Arellano and Bond (1991) suggest
transforming the estimated equation (1) into rst-difference to eliminate country
specic effects as follows:
TFPit – TFPit–s = α (TFPit–1 – TFPit–2) + β1 (Xit – Xit–1) + (εit–s – εit–s) (6)
where X is a vector of independent variables. Within this framework, lagged levels
of the regressors are used as instruments to alleviate bias introduced by possible
endogeneity of regressors and also the correlation between (TFPit-1 – TFPit-2) and
(εit – εit-1). This strategy is valid under two assumptions: (i) the error term is not
serially correlated, (ii) the lag of explanatory variables are weakly exogenous. Then,
following Arellano and Bond (1991) the moment conditions are set as follows:
E[TFPi,t–s � (εi,t – εi,t–1)] = 0 for s ≥ 2; t = 3, ..., T (7)
E[Xi,t–s � (εi,t – εi,t–1)] = 0 for s ≥ 2; t = 3, ..., T (8)
However, Alonso-Borrego and Arellano (1999) and Blundell and Bond (1998)
show that the lagged levels of variables become weak instruments when explanatory
variables are persistent. This problem can result in biased parameter estimates
and inated variance. To address this problem, an alternative system estimator
was proposed by Arellano and Bover (1995) which combines both difference and
level equations in one system of equation. This strategy is known as system GMM
and was shown to be able to reduce bias and imprecision associated with different
estimator (Blundell and Bond, 1998).
In this approach, lagged first-difference and lagged levels are used as
instruments for equations in levels and rst difference, respectively. Hence, moment
conditions for regression in difference are maintained as in (7) and (8) and additional
moment conditions for regression in levels are set as follows:
[TFPi,t–s – TFPi,t–s–1 � (ηi + εi,t)] = 0 for s ≥ 2; t = 3, ..., T (9)
[Xi,t–s – Xi,t–s–1 � (ηi + εi,t)] = 0 for s ≥ 2; t = 3, ..., T (10)
51
R&D Spillovers and the Role of Economic Freedom
Two specication tests are needed to determine the consistency of GMM
estimator. The rst is Sargan Test which is used to examine over–identifying
restrictions with the null of joint validity of all instruments. The second test
examines the hypothesis of no second–order serial correlation in the error term of
the regression in difference as assumed in Equation (6) (Arellano and Bond, 1991).
If the results fail to reject both null hypotheses, this would indicate that the model
is adequately specied and the instruments are valid.
RESULTS AND DISCUSSION
The main objective of this paper is to estimate R&D spillovers through import and
international student ow and also to investigate the role of economic freedom in
mediating spillover effects. To this end, the GMM estimator outlined in the previous
section is used and results are presented in Table 1. Models (1) and (2) include
import and international student ow as spillover channels, separately. Model (3)
includes both channels simultaneously. As shown in Table 1, TFP elasticities with
respect to both foreign capital stocks have plausible magnitudes, lying in absolute
value between zero and one.
Import-weighted and student ow-weighted foreign capital stocks are found
to be important in all cases. The estimated elasticity for import-weighted capital
stock ranges from 0.06-0.08 while the one for student ow is between 0.016 and
0.1. This suggests that increase in import-weighted foreign R&D capital stock
will increase domestic productivity by 0.06-0.08 percentage point. In the case
of student ows channel, it will increase productivity by 0.016 to 0.1 percentage
point. Additionally, the estimated regression passed both specication tests. The
Table 1 R&D spillovers via student ow and import
(1) (2) (3)
R&Dit-1 0.1981*** 0.0431*** 0.3405***
Sm0.0833*** 0.0658***
Ssf 0.0163*** 0.1055***
Sargan test (p-value) 0.327 0.684 0.921
AR (2) test (p-value) 0.212 0.725 0.916
Number of observation 600 600 600
Notes: All variables are expressed in logarithmic form. Sm, Ssf, are respectively import-
weighted foreign R&D, student ow-weighted foreign R&D. *** indicate statistical
signicance at the 1% level.
52
International Journal of Economics and Management
null of no second-order serial correlation cannot be rejected at the 5% level. Also,
the regression results is not affected by simultaneity bias as the orthogonality
conditions cannot be rejected at the 5% level, as indicated by the Sargan test. This
suggests that the equation is adequately-specied and the instruments employed
in the analysis are valid.
The nding is consistent with Coe and Helpman (1995), Coe et al. (1997),
Lichtenberg and van Pottelsberghe (1998) and van Pottelsberghe and Lichtenberg
(2001) who also nd the importance of imports as an important channel. It is also
in line with Park (2004), Le (2010) and Le (2012) on the role of student ows in
enhancing domestic TFP. However, in term of magnitude of the impact, this nding
is not consistent with Park (2004) who nds import as a more important channel
than student ow across a group of developed countries for the period 1971-1990.
This was further supported by Le (2010) which focuses on spillover effects from
developed to developing countries during the 1998-2005 period. One potential
reason for the difference between our nding and those of Park (2004) and Le
(2010) is because we use recent data during which globalization is prominent.
Generally, our nding support the idea that globalization with advancements in
communication technology promotes a greater role of human capital mobility in
enhancing productivity. As noted by Filatotchev et al. (2011), mobility across border
nowadays is easier than decades before as globalization taking place.
The next step of the analysis is to assess whether economic freedom plays
an important role in mediating R&D spillovers. To this end, we extend Equation
(1) to include interaction term constructed as the product of foreign capital stocks
and the economic freedom (EF) index (i.e. Sm×EF and Sfs×EF). To ensure that the
interaction term does not proxy for Sm, Sfs, and EF, the economic freedom were
included in the regression independently. Within this framework, we rely on the
interaction term to establish the contingency effects. If the term is positive and
signicant, this would imply that the R&D spillovers increase with economic
freedom6. The results of this exercise are tabulated in Table 2. The rst thing to note
is that interaction term Sfs×EF turns out to be positive and statistically signicant at
the 5% level. This result implies that the effect of foreign R&D via student ows
on TFP increases monotonically with EF. However, the same effect could not be
established for spillover effects via import. Additionally, all other variables in
6 It should be noted that the inclusion of interaction term in our model may lead to multicollinearity
problem as the term tends to strongly correlated with original variables. This paper follows Azman-Saini
et al. (2010) suggestion to adopt the following two-step procedure: First, interaction term is regressed on
the foreign capital stock with EF (i.e. Sm×EF and Sfs×EF) and then the residuals from the regressions
in the rst step are saved and used to represent the interaction term.
53
R&D Spillovers and the Role of Economic Freedom
level are positive and statistically signicant. The p-values of both second-order
serial correlation and the Sargan over identication tests suggest that the model is
adequately specied.
Table 2 Role of Economic Freedom in R&D Spillovers
(4) (5)
R&Dit-1 0.4799*** 0.5063***
Sm0.0643*** 0.0647***
Ssf 0.0987*** 0.1088***
EF 0.3451*** 0.3989***
Sm×EF 0.0028
Ssf×EF 0.5139*** 0.3436***
Sargan test (p-value) 0.830 0.925
AR (2) test (p-value) 0.331 0.447
Number of observation 600 600
Notes: All variables are expressed in logarithmic form. Sm, Ssf, and EF are
respectively import-weighted foreign R&D, student ow-weighted foreign R&D
and economic freedom. *** indicate statistical signicance at the 1% level.
The nding is consistent with several papers who nd that economic freedom
is important in inuencing economic performance. For instance, Doucouliagos and
Ulubasoglu (2006) reveal that economic freedom has an indirect inuence on growth
through physical capital accumulation. Meanwhile Azman-Saini et al. (2010) nd
that economic freedom moderate the impact of FDI on growth. Recently, Farhadi,
Islam and Moslehi (2015) show that improvement in economic freedom is expected
to enhance growth through improved rent of natural resources. The overall nding
supports the view on the importance of promoting freedom of economic activities
to facilitate knowledge spillovers.
Several robustness checks are carried out to ensure that the results we obtain
are robust. First, we compute TFP as Y / (KβL1-β). This measurement was used
in the pioneering study of Coe and Helpman (1995) and many others (e.g. Coe
et al., 1997; van Pottelsberghe and Lichtenberg, 2001; Park, 2004; Le, 2010).
The result of using alternative TFP is reported in Column (1) in Table 3. Second,
we use a different measure of import. Specically, we use import of machinery
and transport equipment and the results are reported in column (2). Third, we use
import of manufactured goods and result is presented in column (3). Finally, we
expand the source countries for foreign R&D as well as countries of destination
54
International Journal of Economics and Management
for student ows. We include 16 OECD countries for this purpose and the results
are reported in column (4) of Table 37.
Table 3 Robustness checks
(1) (2) (3) (4)
Lagged dependent 0.5450*** 0.4672*** 0.4816*** 0.4012***
Sm0.0938*** 0.0781*** 0.0798*** 5.5480***
Ssf 0.0933*** 0.0855*** 0.0961*** 6.7832***
EF 0.2183*** 0.3064*** 0.2958*** 83.0271***
Ssf×EF 0.4122*** 0.4429*** 0.4344*** 51.8342***
Sargan test (p-value) 0.989 0.886 0.830 0.936
AR (2) test (p-value) 0.427 0.424 0.426 0.311
Number of observation 600 600 600 600
Notes: All variables are expressed in logarithmic form. Sm, Ssf, and EF are respectively import-weighted
foreign R&D, student ow-weighted foreign R&D and economic freedom. *** indicate statistical
signicance at the 1% level.
The results in Table 3 show that all variables are statistically signicant at the
1% level and retained their positive signs. Additionally, diagnostic tests for all four
regressions suggest the models are adequately specied. Overall, this suggests that
the results are consistent and robust.
However, it is worth nothing that when traditional TFP measure is used, the
magnitude of coefcients suggest different story about the relative importance
of spillover channel. Result in column (1) shows that the size of coefcients on
import and international student ow are almost the same which suggest that both
channels are equally important for spillover effects. Another observation is that
import of machinery and equipment columns and import of manufactured goods
performs better than total import (i.e. results presented in Table 3). These ndings
are similar to those of Coe et al. (1997). This is consistent with the view that many
consumer goods and services have less technical contents to have any important
impact on productivity.
7 Australia, Belgium, Canada, Denmark, Finland, France, Germany, Ireland, Italy, Japan, Netherland,
Norway, Spain, Sweden, United Kingdom, and United States.
55
R&D Spillovers and the Role of Economic Freedom
CONCLUSIONS
This study examines R&D spillovers from developed to developing countries via
import and student ow channels. It also evaluate whether economic freedom
make a difference to the way knowledge are transmitted across borders. The
results show that international student ow has a greater inuence than import in
transmitting knowledge across borders. Thus, this nding supports the view that
exibility in human capital movement across border would enhance the spillover
of disembodied knowledge or technology. In addition, economic freedom is
found to be able to moderate the spillover effects through international student
ow. Thus, countries that actively promote freedom of economic activity could
gain more in productivity improvement via this channel. Nevertheless, there is no
enough evidence to support its role via import channel. These results cast doubt
on the role of economic freedom in assisting the acquisition of new knowledge
embodied in imported goods. The results are robust to several sensitivity checks
such as different measures of TFP and imports weighted foreign capital stocks.
This suggests that government policies that encourage knowledge acquisition in
foreign country are expected to enhance domestic productivity. Also, countries
that promote freedom of economic activity will provide better environments for
domestic rms to internalize foreign technologies.
REFERENCES
Aghion, P., and Howitt, P. (1992), “A model of growth through creative destruction”,
Econometrica, Vol. 60, No. 2, pp. 323-351.
Alfaro, L, Chanda, A., Kalemli-Ozcan, S. and Sayek, S. (2004), “FDI and economic growth:
The role of local nancial markets”, Journal of International Economics, Vol. 64, No.
1, pp. 113-134
Alfaro, L, Chanda, A., Kalemli-Ozcan, S. and Sayek, S. (2010), “Does foreign direct
investment promote growth? Exploring the role of nancial markets on linkages”, Journal
of Development Economics, Vol. 91, No. 2, pp. 242-256.
Almeida, P., and Kogut, B. (1999), “Localization of knowledge and the mobility of engineers
in the regional networks”, Management and Science, Vol. 45, No. 7, pp. 905-916.
Alonso-Borrego, C., and Arrelano, M. (1999), “Symmetrically normalized instrumental-
variable estimation using panel data”, Journal of Business & Economic Statistics, Vol.
17, No. 1, pp. 36-49.
Arellano, M., and Bond, S.R. (1991), “Some tests of specication for panel data: monte carlo
evidence with an application for employment equations”, Review of Economic Studies,
Vol. 58, No. 2, pp. 277-297.
56
International Journal of Economics and Management
Arellano, M., Bover, O. (1995), “Another look at the instrumental-variable estimation of
error-components models”, Journal of Econometrics, Vol. 68, No. 1, pp. 29-52.
Azman-Saini, W.N.W., Baharumshah, A.Z., and Law, S.H. (2010), “Foreign direct
investment, economic freedom and economic growth: International evidence”, Economic
Modelling, Vol. 27, No. 5, pp. 1079-1089.
Barro, R.J. and Lee, J.W. (2010), “A new data set of educational attainment in the world,
1950-2010”, Journal of Development Economics, Vol. 104, pp. 184-198.
Bitzer, J. & Kerekes, M. (2008), “Does foreign direct investment transfer technology across
borders? New evidence”. Economics Letters, Vol. 100, No. 3, pp. 355-358.
Blomstrom, M, Lipsey, R.E. and Zejan, M. (1994), “What explains growth in developing
countries?”, NBER Discussion Paper, No. 1924.
Blundell, R., and Bond, S. (1998), “Initial conditions and moment restrictions in dynamic
panel data models”, Journal of Econometrics, Vol. 87, No. 1, pp. 115-143.
Borenztein, E., De Gregorio, J., and Lee, J. (1998), “How does foreign direct investment affect
economic growth?”, Journal of International Economics, Vol. 45, No. 1, pp. 115-135.
Coe, D.T., and Helpman, E. (1995), “International R&D spillover”, European Economic
Review, Vol. 39, No. 5, pp. 859-887.
Coe, D.T., Helpman E., and Hoffmaister, A.W. (1997), “North South R&D spillovers”, The
Economic Journal, Vol. 107, No. 440, pp. 134-149.
Cohen, W.M., Goto, A., Nagata, A., Nelson, R.R., and Walsh, J.P. (2002), “R&D spillovers,
patent and incentives to innovate in Japan and the United States”, Research Policy, Vol.
31, No. 8-9, pp. 1349-1367.
Cohen, W.M. and Leventhal, D.A. (1990), “Absorptive capacity: A new perspective on
learning and innovation”, Administrative Science Quarterly, Vol. 35, No. 1, pp. 128-152.
Doucouliagos, C., and Ulubasoglu, M.A., (2006), “Economic freedom and economic growth:
Does specication make a difference?”, European Journal of Political Economy, Vol.
22, No. 1, pp. 60-81.
Dunning, J. H. (1994), “Multinational enterprises and the globalization of innovatory
capacity”, Research Policy, Vol. 23, No. 1, pp. 67-88.
Durham, J.B. (2004), “Absorptive capacity and the effects of foreign direct investment and
equity foreign portfolio investment on economic growth”, European Economic Review,
Vol 48, No. 2, pp. 285-306.
Farhadi, M., Islam, M.R., and Moslehi, S. (2015), “Economic freedom and productivity
growth in resource-rich economies”, World Development, Vol. 72, No. C, pp. 109-126.
Filatotchev, I., Liu, X., Lu, J., and Wright, M. (2011), “Knowledge spillovers through human
mobility across national borders: Evidence from Zhongguancun Science Park in China”,
Research Policy, Vol. 40, No. 3, pp. 453-462.
57
R&D Spillovers and the Role of Economic Freedom
Grifth, R., Redding, S., and van Reenen, J. (2003), “R&D and absorptive capacity: Theory
and empirical evidence”, Scandinavian Journal of Economics, Vol. 105, Nno. 1, pp.
99-118.
Griliches, Z. (1988), “Productivity Puzzles and R&D: Another Nonexplanation”, Journal
of Economic Perspectives, Vol. 2, No. 4, pp. 9-21.
Griliches, Z. (1992), “The Search for R&D Spillovers”, Scandinavian Journal of Economics,
Vol. 94, pp. S29-S47.
Grossman, G., and Helpman, E. (1991), Innovation and growth in the global economy, The
MIT Press, Cambridge.
Hall, R.E., and Jones, C.I. (1999), “Why do some countries produce so much more output per
worker than others?”, The Quarterly Journal of Economics, Vol. 114, No. 1, pp. 83-116.
Hermes, N. and Lensink, R. (2003), “Foreign direct investment, nancial development and
economic growth”, Journal of Development Studies, Vol. 40, No. 1, pp. 142-163.
Holtz-Eakin, D., Newey, W., and Rosen, H. (1988), “Estimating vector autoregressions with
panel data”, Econometrica, Vol. 56, No. 6, pp. 1371-1395.
Keller, W. (2004), “International technology diffusion”, Journal of Economic Literature,
Vol. 42, No. 3, pp. 752-782.
Kim, J.W., and Lee, H.K. (2004), “Embodied and disembodied international spillovers of
R&D in OECD manufacturing industries”, Technovation, Vol. 24, No. 4, pp. 359-368.
Klenow, P.J., and Rodriguez-Clare, A. (1997), “The neoclassical revival in growth economics:
Has it gone too far?”, NBER Macroeconomics Annual, Vol. 12, pp. 73-114.
Koskinen, K.U., Pihlanto, P., and Vanharanta, H. (2003), “Tacit knowledge acquisition and
sharing in a project work context”, International Journal of Project Management, Vol.
21, No. 4, pp. 281-290.
Kogut, B. and Chang, S.J. (1991), “Technological Capabilities and Japanese Foreign Direct
Investment in the United States”, Review of Economics and Statistics, Vol. 73, No. 3,
pp. 401-413.
Kwark, N.S. and Shin, Y.S. (2006), “International R&D spillovers revisited: Human capital
as an absorptive capacity for foreign technology”, International Economic Journal, Vol.
20, No. 2, pp. 179-196.
Le, T. (2008), “Brain drain or brain circulation: Evidence from OECD’s international
migration and R&D spillovers”, Scottish Journal of Political Economy, Vol. 55, No. 5,
pp. 618-636.
Le, T. (2010), “Are student ows a signicant channel of R&D spillovers from the north to
south?”, Economic Letters, Vol. 107, No. 3, pp. 315-317.
Le, T. (2012), “R&D spillovers through student ows, institutions, and economic growth:
What can we learn from African countries?”, Scottish Journal of Political Economy,
Vol. 59, No. 1, pp. 115-130.
58
International Journal of Economics and Management
Lee, G. (2005), “Direct versus indirect international R&D spillovers”, Information Economics
and Policy, Vol. 17, No. 3, pp. 334-348.
Lichtenberg, F.R., and van Pottelsberghe de la Potterie, B. (1998), “International R&D
spillovers: A comment”, European Economic Review, Vol. 42, No. 8, pp. 1483-1491.
Lumenga-neso, O., Olarreaga, M. and Schiff, M. (2005) “On `indirect’ trade-related R&D
spillovers”, European Economic Review, Vol. 49, No. 7, pp. 1785-1798.
Madden, G., Savage, S.J., and Bloxham, P. (2001), “Asian and OECD international
spillovers”, Applied Economics Letters, Vol.8, No. 7, pp. 431-435.
Mohnen, P. (1996), “R&D Externalities and Productivity Growth”, STI Review, OECD,
No. 18, pp. 39-66.
Nadiri, M.I. (1993), “Innovation and Technological Spillovers”, NBER working paper No.
4423.
OECD (2003), Proposed Standard Practice for Surveys of Research and Development
(Frascati Manual), OECD, Paris.
Park, J. (2004), “International student ows and R&D spillovers”, Economic Letters, Vol.
82, No. 3, pp. 315-320.
Psacharopoulos, G. (1994), “Returns to investment in education: A global update”, World
Development, Vol. 22, No. 9, pp. 1325-1343.
Romer, P.M. (1990), “Endogenous technological change”, Journal of Political Economy,
Vol. 98, No. 5, pp. S71-S102.
Saggi, K. (2002), “Trade, foreign direct investment, and international technology transfer:
a survey”, World Bank Research Observer, Vol. 17, No. 2, pp. 191-235.
Savvides, A. and Zachariadis, M. (2005), “International technology diffusion and the growth
of TFP in the manufacturing sector of developing economies”, Review of Development
Economics, Vol. 9, No. 4, pp. 482-501.
Song, J., Almeida, P., and Wu, G. (2003), “Learning by hiring: When is mobility more likely
to facilitate interrm knowledge transfer?”, Management Science, Vol. 49, No. 4, 351-365.
Tang, L., and Koveos, P.E. (2008), “Embodied and disembodied R&D spillovers to developed
and developing countries”, International Business Review, Vol. 17, No. 15, pp. 546-558.
UNCTAD (2001), World investment report: promoting linkages, United Nations.
Van Pottelsberghe de la Potterie, B., and Lichtenberg, F. (2001), “Does foreign direct
investment transfer technology across border”, The Review of Economics and Statistics,
Vol. 83, No. 3, pp. 490-497.
Xu, B. & Wang, J. (1999), “Capital goods trade and R&D spillovers in the OECD”, Canadian
Journal of Economics, Vol. 32, No. 5, pp. 1258-74.
59
R&D Spillovers and the Role of Economic Freedom
Zhu, L. & Jeon, B.N. (2007), “International R&D Spillovers: Trade, FDI, and Information
Technology as Spillover Channels”, Review of International Economics, Vol. 15, No.
5, pp. 955-976.
Zucker, L.G., Darby, M.R., and Brewer, M.B. (1998), “Intellectual human capital and the
birth of U.S. biotechnology enterprises”, American Economic Review, Vol. 88, No. 1,
pp. 290-306.
60
International Journal of Economics and Management
APPENDIX
List of Countries
Albania Fiji Pakistan
Algeria Gabon Panama
Argentina Ghana Papua New Guinea
Bahrain Guatemala Paraguay
Bangladesh Guyana Peru
Barbados Honduras Philippines
Belize Hungary Poland
Benin India Romania
Bolivia Indonesia Russian Federation
Botswana Iran, Islamic Rep. Rwanda
Brazil Jordan Senegal
Bulgaria Kenya Sierra Leone
Burundi Kuwait South Africa
Cameroon Latvia Sri Lanka
Central African Rep. Lithuania Syrian Arab Rep.
Chile Malawi Thailand
Colombia Malaysia Togo
Congo, Rep. Mali Tunisia
Costa Rica Mauritius Turkey
Cote d’Ivoire Mexico Uganda
Croatia Morocco Ukraine
Dominican Rep. Namibia Uruguay
Ecuador Nepal Venezuela, RB
Egypt, Arab Rep. Nicaragua Zambia
El Salvador Niger Zimbabwe
