Академический Документы
Профессиональный Документы
Культура Документы
Research Policy
journal homepage: www.elsevier.com/locate/respol
a r t i c l e
i n f o
Article history:
Received 9 November 2009
Received in revised form 5 March 2012
Accepted 28 April 2012
Available online 21 May 2012
Keywords:
Intangibles
Technology adoption
Diffusion
Complementarities
a b s t r a c t
Our objective in this paper is to analyse the determinants of the use of advanced manufacturing technologies in manufacturing rms. We go beyond more traditional approaches and consider the role of
complementarities in technology adoption at two levels. First, we adapt Teeces (1986) framework to
study the incentives to use new technology that stem from investments in R&D, human capital and advertising. Second, we analyse whether technology use is conditioned by a system effect that arises from the
use of related technologies. We test our hypotheses on a representative sample of manufacturing rms
in Spain. Our results fully support the idea that R&D investments increase the likelihood of technology
use, but only offer partial support for human capital and advertising investments. Export intensity, being
part of a business group and epidemic effects are also important determinants of adoption.
2012 Elsevier B.V. All rights reserved.
1. Introduction
Our objective in this paper is to understand the role of intangible resources in technology adoption. The idea that intangible
resources contribute to the explanation of technology diffusion
among rms has its roots in some analyses of the impact of information technology (IT) on performance. A part of this literature has
argued that IT are strategic necessities in the sense that they have to
be adopted in order not to suffer from competitive disadvantages,
but that they cannot produce sustainable competitive advantages
(Clemons and Kimbrough, 1986; Clemons and Row, 1991). Recent
research on complementarities has shown that the performance
of IT frequently depends on their interaction with organizational
elements (Powell and Den-Micallef, 1997). In particular, the application of the resource-based view of the rm suggests that, in order
to achieve inimitability, rms have to combine IT with resources
protected by isolating mechanisms such as tacitness, causal ambiguity and time compression diseconomies2 (Barney, 1991; Dierickx
Corresponding author. Tel.: +34 941 299 572; fax: +34 941 299 393.
E-mail addresses: jaime.gomez@unirioja.es (J. Gmez), pilar.vargas@unirioja.es
(P. Vargas).
1
Tel.: +34 941 299 373; fax: +34 941 299 393.
2
Time compression diseconomies make reference to the idea that the returns
diminish when one input, in this case time, remains constant (Dierickx and Cool,
1989). Intangible assets, such as the ones created through investments in R&D or
advertising are likely to be subject to them. Dierickx and Cool (1989, p. 1507) provide
an example for R&D investments: In the case of R&D, the presence of time compression diseconomies implies that maintaining a given rate of R&D spending over a
particular time interval produces a larger increment to the stock of R&D know-how
than maintaining twice this rate over half the time interval.
0048-7333/$ see front matter 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.respol.2012.04.016
1608
3
For example, Cassiman and Veugelers (1999) study the complementarities
between external and internal R&D investments, whereas Belderbos et al. (2006)
assess the complementarities arising in R&D cooperation strategies. A review of the
literature on complementarities and on the types of factors considered may be found
in Ennen and Richter (2010).
may have to use technologies. Following Teece (1986), the capacity of a rm to appropriate the returns from innovation mainly
depends on two types of factors (Teece, 1986): the appropriability
regime and the possession of complementary assets. The appropriability regime is dened by (1) legal and (2) technological factors.
Additionally, the rents accruing to the innovator also depend on
the possession of complementary assets. By complementary assets,
Teece (1986) understands all the resources that need to be jointly
used with the innovation in order to exploit it. They include, for
example, the manufacturing, distribution and service resources
that are needed to operate the different stages of the value chain.
Teeces framework could be easily adapted to understand the
incentives of rms to adopt new technologies. The fact that new
technologies are freely available in the market (there are no legal
barriers protecting the adopter) focuses attention on complementarity assets.4 Despite the weak appropriability regime of new
technologies, the returns from their use could be captured if
the rm possesses complementary resources that are difcult to
obtain, such as the ones created by investments in research and
development and advertising. In fact, research has shown that
the link between IT and protability could be explained through
the consideration of complementary resources under the control
of the rm (Powell and Den-Micallef, 1997). However, their use
as determinants of adoption has been scarce. The existence of
difcult-to-imitate complementary resources could be an indicator of the stimuli that a rm has to adopt a new technology, given
that they could provide the basis for competitive advantage.
Selecting the types of resources more likely to be combined with
new technologies to provide competitive advantages is a difcult
task. Although both tangible and intangible assets play an important role in creating value, intangible assets frequently present time
compression diseconomies and they are typically tacit and hard
to codify (Conner and Prahalad, 1996; Kogut and Zander, 1992).
Additionally, they are likely to be traded in imperfect factor markets (Barney, 1986) and are not consumed by their use (Collins and
Montgomery, 1998). Teece (2000) suggests that a rms superior
performance depends on its ability to defend and use the intangible
assets it creates. Accordingly, we argue that technology adoption is
conditioned by the possession of certain complementary resources
that are difcult to acquire or copy, namely, technological, human
and marketing resources.5 Similarly, we also analyse the presence
of complementary relationships between technologies. As mentioned in the Introduction, we focus on production technologies
and, in particular, the group of technologies that has been termed
advanced manufacturing technologies.6
4
In fact, this has led some researchers to conclude that, in the long run, technology adoption might only provide competitive parity (Barney, 1991). Some authors
have enunciated this as the strategic necessity hypothesis (Clemons and Kimbrough,
1986; Clemons and Row, 1991). Following this hypothesis, a focal rm would have
incentives to adopt a valuable new technology. However, this would create incentives in rival rms to invest in the same technology in order not to obtain below
normal prots. As long as new technologies are freely available in the market, the
long-run result would be that rival rms would erode the rents accruing to rst
adopters, leading to competitive parity.
5
We are not claiming that difcult-to-imitate complementary resources are the
only ones likely to interact with new technologies to provide competitive advantage.
In fact, Rivkin (2000) points out that complex systems could be inimitable, even if
their elements are not.
6
As suggested by an anonymous referee, we expect differences between the
effects of each of the variables depending on the type of technology considered.
For example, Rogers and Shoemaker (1971) argue that the diffusion of innovations
depends on their attributes, namely, relative advantage, compatibility, complexity,
trialability and observability. For example, we would expect the effect of technological resources to be more important in those cases in which the technology is more
complex. Therefore, the magnitude of the associated effect should be higher for
exible manufacturing than for the other technologies. Similarly, we would expect
a higher effect of the qualications of human resources in those cases in which the
technology is more complex. However, the fact that we do not have information on
all these ve dimensions (or other that could be considered as important) precludes
as from being more precise in our arguments.
7
Some papers in the agricultural economics literature have used education to
explain adoption. However, its use seems to be more related to the concept of
absorptive capacity. For example, Huffman (1974, p. 85) refers to an allocative
effect, understood as the human agents ability to acquire, decode, and sort market and technical information efciently. We attempt to emphasise the idea that
these investments are important for exploiting and appropriating the returns to the
technology.
1609
reading and basic math skills and they should also able to deal
with higher levels of abstraction and to act quickly on complex
information (Parthasarthy and Sethi, 1992). Thus, more automated
plants will employ relatively more educated and skilled workers.
Second, the introduction of a more technology-specic level may
affect the organization of the workforce and will require skilled
operators and technicians who replace skilled craftspeople but also
less skilled workers. Finally, many of these technologies require
qualied support staff to install and maintain them.
It could be argued that these abilities have commodity-like
character, given employee mobility, not constituting a difcult-toimitate resource. However, researchers have maintained that more
qualied workers are more likely to invest in rm-specic investments. The reason is that their mobility opportunities are less than
for employees with low qualications and their activities tend to
imply the use of their intellectual capabilities in idiosyncratic tasks
(Vicente-Lorente, 2000). This makes it highly likely that more qualied workers invest more in rm-specic abilities than less qualied
ones.
More qualied workers also provide other abilities related to
innovation and management that are important in the use of new
technologies. First, as mentioned above, new technologies frequently enable process and product innovations (Koellinger, 2008).
Education contributes to the innovation process by increasing a
persons capacity to think systematically and creatively about
techniques (Wozniak, 1984, p. 71). Second, education may also be
related to management skills, in other words, the increasing ability of educated workers to effectively integrate new technologies
into the activities of the rm. In fact, given that technology-related
knowledge may be contracted and is not rm- (but technology-)
specic, it has been argued that management skills are the only
likely source of competitive advantage: they are path dependent,
they tend to be tacit and rm-specic and they may be socially
complex (Mata et al., 1995).
Hypothesis 2. Firms with more highly qualied personnel are
more likely to use a new technology.
2.3. Marketing resources and technology adoption
Complementarities are also present in marketing activities
(Teece, 1986). The literature on management understands marketing investments as rm-specic assets (see, for example,
Balakrishnan and Fox, 1993; Vicente-Lorente, 2001). Like technological resources, investments in marketing create intangible
resources, such as reputation, brand image and closer relationships
with customers. These resources are frequently difcult to imitate:
they may be subject to time compression diseconomies (Dierickx
and Cool, 1989) or be socially complex (Barney, 1991). Investments
in marketing activities help the rm to develop a good reputation or
a brand image. These assets are quasi-public, in the sense that they
can continue to deliver services after being used. In other words,
once they have been developed, they can be used to support product development, market development and diversication (Ansoff,
1965).
To understand the complementarities between AMT and
marketing investments it is rst necessary to emphasize two characteristics of the former. First, AMT can be used to improve the
information processing capabilities of rms (Kotha and Swamidass,
2000). Firms can use new technologies in order to cope with the
uncertainty surrounding the activities that they perform, given
that they help to collect and manage information. Those rms
that follow a strategy that is intensive in information processing
requirements will benet most from the use of new technology.
Second, AMT also improve the exibility of rms to adapt to changing demands (Parthasarthy and Sethi, 1992). In particular, they
1610
The usual tests show that the sample is representative of the total population.
In the survey, rms are asked whether their production process uses any of
the following systems: (1) computer numerically controlled (CNC) machines, (2)
robotics, (3) computer assisted design (CAD), (4) combination of some of the above
systems through a central computer (FMS).
10
1611
Table 1
Percentage of adopters and non adopters by technology and year.
CNC
1994
1998
2002
2006
ROBOTICS
CAD
FMS
B2B
B2C
S2B
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
33.1
43.9
48.3
51.5
66.9
56.1
51.7
48.5
18.1
26.3
26.4
31.5
81.9
73.7
73.6
68.5
24.2
35.3
36.9
40.0
75.8
64.7
63.1
60.0
21.4
26.7
22.3
27.1
78.6
73.3
77.7
72.9
5.8
9.9
94.2
90.1
4.1
5.1
95.9
94.9
13.8
25.9
86.2
74.1
The expressions Yes and No indicate whether the rms have adopted or not
complex manufacturing systems may be built that make manufacturing programmable and lead to timely information transfer
across departments, employees, customers and suppliers. This adds
a new element to the study of complementarities: these technologies not only interact with the endowment of resources of the rm,
but they are frequently used as a system. This reinforces the idea
that complementarities have a key role to play in the understanding
of performance differences and adoption behaviour.
Although these four technologies will be the main focus of
our analysis, the Survey also includes information on the use of
e-business in 2002 and 2006, the last two periods for which information has been published. In particular, the available data refer to
the adoption of business-to-business (B2B), business-to-consumer
(B2C) and supplier-to-business (S2B) technologies. Although the
main objective of these technologies is to manage information
electronically, they are also used for enhancing communication
between rms. In other words, similarly to Powell and Den-Micallef
(1997), we distinguish between technologies that are used within
rms (in-rm technologies, namely, CNC, CAD, robotics and FMS)
and those used beyond their limits (beyond-rm technologies,
namely, B2B, B2C and S2B). Following Hypothesis 4, our estimations
should show a stronger relationship between in-rm technologies
than between in-rm and beyond-rm technologies.
Table 1 offers a rst approximation to the data, showing the
distribution of adopters and non-adopters by technology, and year.
As can be seen, the number of adopters is different depending on the
technology. CNC is the most-used technology (51.5% of the rms use
it in 2006), followed by CAD (40.0% in 2006). Robotics and FMS have
Table 2
Percentage of adopters in 2006 by technology and industry.
CNC
ROBOTICS
Meat products
Food and tobacco
Beverages
Textiles and clothing
Leather and footwear
Wood industry
Paper
Edition and graphical arts
Chemical products
Plastic and rubber
Non-metallic minerals
Metallurgy
Metallic products
Machinery for agriculture and industry
Machinery for ofces, data processing, etc.
Electrical material and accessories
Motors and autos
Other transport material
Furniture
Other manufacturing
28.6
41.5
37.5
36.9
34.8
47.5
56.4
37.9
34.8
52.5
46.1
62.0
62.1
64.5
46.2
57.4
75.4
75.0
67.2
42.1
21.4
27.7
43.8
16.7
17.4
25.0
41.0
17.2
18.8
42.6
35.5
42.0
34.7
32.9
30.8
38.9
62.3
43.8
21.3
15.8
Total manufacturing
2
51.5
72.2***
31.5
68.5***
*
***
CAD
3.6
10.6
25.0
42.9
17.4
17.5
46.2
31.0
23.2
42.6
31.6
56.0
45.2
69.7
53.8
57.4
70.5
87.5
39.3
26.3
40.0
160.2***
FMS
B2B
B2C
S2B
7.1
20.2
37.5
21.4
8.7
12.5
38.5
24.1
31.9
29.5
22.4
38.0
29.8
31.6
23.1
31.5
45.9
43.8
19.7
15.8
10.7
5.3
6.3
8.3
0
2.5
23.1
15.5
15.9
11.5
9.2
14.0
7.3
7.9
15.4
13.0
4.9
6.3
13.1
10.5
3.6
2.1
12.5
6.0
0
0
5.1
22.4
4.3
4.9
2.6
4.0
2.4
5.3
0
5.6
6.6
0
6.6
5.3
21.4
17.0
37.5
25.0
13.0
10.0
23.1
39.7
24.6
23.0
17.1
36.0
29.0
30.3
30.8
25.9
32.8
37.5
29.5
21.1
27.1
44.0***
9.9
26.3
5.1
48.4***
25.9
29.3*
1612
Table 3
Probability of adoption of AMT depending on the number of other technologies adopted and unconditional probability of adoption (2006).
CNN
ROBOTICS
CAD
FMS
Unconditional probability
28.4%
16.9%
9.2%
10.0%
53.4%
30.0%
31.9%
16.8%
78.7%
49.1%
61.6%
39.4%
89.3%
64.3%
81.2%
60.7%
51.5%
31.5%
40.0%
27.1%
Kt =
3
(1 ) Rtk
k=0
11
The survey collects data on (1) external expenses, (2) internal expenses and
(3) total expenses ((1) + (2)) in research and development. To calculate the stock of
technological resources, we have used internal R&D expenses.
12
Firms are asked to classify all the workers according to their qualications.
The options available are: (1) engineers and graduates, (2) middle level engineers,
experts and qualied assistants and (3) other personnel. From this information we
calculated the percentage of workers with a university degree.
3
(1 ) Atk
k=0
13
Firms provide some information on their accounts. For example, they are asked
to report the expenses in publicity, advertisement and public relations.
14
Recent efforts have tried to explain this relationship (see, for example Astebro,
2002), although some confusion about the underlying mechanism still remains.
3.3.4. Ownership
The fourth variable considered in the analysis is the type of
ownership. The argument in favour of a positive effect is that the
subsidiaries of a foreign rm may have access to the resources
of the parent rm. Foreign investment may be a vehicle for the
introduction of superior technology and scientic knowledge. Previous evidence on adoption models revealed a positive impact of
foreign ownership (Baldwin and Diverty, 1995; Bosworth, 1996;
Faria et al., 2003), although this relationship was only signicant in
the case of certain technologies. The presence of foreign capital in
the focal rm is measured through a dummy that takes a value of
one when the presence of foreign investors in the capital is higher
than 30%.
15
1613
1614
16
Following the comments of one of the anonymous referees, we also estimated an
ordered probit model that considers whether the rm adopts 0, 1, 2, 3 or 4 manufacturing technologies. The results are highly consistent with the ones presented here:
rm size, the availability of nancial resources, technological and human resources,
export propensity and integration in a business group are signicant predictors of
adoption. The estimates also show the relevance of epidemic effects.
17
We calculated the variance ination factors (VIF) in order to detect multicollinearity problems. The maximum VIF was 5.02 and the average VIF was 2.47.
Therefore, we do not have reasons to suspect that they could be present.
18
The variable used to test this hypothesis is calculated as the percentage of workers with a university degree. As one anonymous reviewer points out, this has the
problem that some of these employees will be working on R&D activities and this
may lead to double counting. Although we do have information on the number of
employees working on R&D activities, the classication of the workers into different
groups does not match the one for the whole rm. In any case, we rerun the estimations with a measure of human resources from which we subtracted the number of
employees working for the R&D function that were (1) graduates and (2) middle
level technicians. The results were similar to the ones presented here.
19
An anonymous reviewer suggested investigating on the sources of this capital. Unfortunately, our data do not allow us to identify the country in which these
investments originate. We were, however, able to identify the country in which a
rm invests for the two last periods of our sample. Our results are in line with the
work of Grifth et al. (2006), suggesting that investments in OECD countries are
signicant for explaining Robotics and CAD adoption.
20
We also estimated a model that included both the percentage of adopters per
industry and in the economy as a whole. The only variable that remains signicant
in this model is the rst, offering support for the existence of an industry specic
epidemic effect.
1615
Table 4
The effect of technological, human and marketing resources on AMT adoption.
CNC (1)
Robotics (2)
CAD (3)
FMS (4)
CNC (5)
Robotics (6)
0.17***
(3.76)
0.08
(0.87)
0.38***
(7.77)
0.29***
(2.92)
0.19***
(4.47)
0.10
(1.11)
0.48***
(9.35)
0.33***
(3.41)
Human resources
Export propensity
0.44***
(5.17)
0.38
(1.32)
0.30***
(5.69)
0.02
(0.01)
0.24***
(3.98)
0.34***
(5.69)
0.40***
(6.48)
Yes
0.87***
(4.32)
0.63***
0.51***
(7.13)
(5.74)
0.26
0.20
(0.84)
(0.64)
***
0.47
0.33***
(8.39)
(6.01)
0.16**
0.04
(2.49)
(0.64)
0.23***
0.35***
(3.36)
(5.37)
0.27***
0.38***
(3.95)
(5.84)
0.39***
0.45**
(5.72)
(6.84)
Yes
Yes
1.44***
2.47***
(6.30)
(9.28)
0.409*** (17.46)
***
0.457 (21.18)
0.292*** (11.80)
0.405*** (16.77)
0.297*** (11.69)
0.413*** (17.13)
4418
1907.70***
0.16***
(3.50)
0.07
(0.83)
1.61***
(2.86)
0.12
(0.71)
0.22
(0.91)
0.41***
(4.81)
0.35
(1.22)
0.29***
(5.37)
0.00
(0.08)
0.24***
(4.02)
0.34***
(5.66)
0.40***
(6.44)
Yes
0.89***
(4.38)
0.36***
(7.28)
0.28***
(2.80)
2.42***
(4.12)
0.14
(0.73)
0.42***
(2.70)
0.60***
(6.65)
0.23
(0.74)
0.45***
(7.95)
0.18***
(2.82)
0.25***
(3.64)
0.29***
(4.17)
0.42***
(5.98)
Yes
1.49***
(6.52)
Firm size
Firm debt ratio
Market concentration
Integrated in a business group
Foreign capital
Year 1998
Year 2002
Year 2006
Industry dummies
Constant
Rho2,1
Rho3,1
Rho4,1
Rho3,2
Rho4,2
Rho4,3
No. observations
Wald test
Comparison test
LR test of Rho2,1 = Rho3,1 = Rho4,
1 = Rho3,2 = Rho4,2 = 0 Rho4,3 = 0
*
**
***
0.33***
(3.75)
0.31
(0.99)
0.39***
(7.10)
0.01
(0.21)
0.14**
(2.16)
0.02
(0.31)
0.13*
(1.93)
Yes
1.74***
(7.32)
1047.75***
CAD (7)
FMS (8)
0.17***
(3.89)
0.12
(1.29)
3.66***
(6.32)
0.66***
(3.64)
0.32*
(1.74)
0.44***
(4.92)
0.23
(0.75)
0.28***
(4.93)
0.01
(0.22)
0.37***
(5.60)
0.39***
(5.87)
0.47***
(6.88)
Yes
2.54***
(9.52)
0.409*** (17.40)
***
0.459 (21.16)
0.290*** (11.67)
0.403*** (16.53)
0.294*** (11.48)
0.404*** (16.54)
4418
1992.44***
111.98***
1025.56***
0.46***
(8.82)
0.33***
(3.36)
2.81***
(4.86)
0.71***
(3.88)
0.33
(0.95)
0.28***
(3.13
0.32
(1.02)
0.35**
(6.21)
0.00
(0.02)
0.14**
(2.17)
0.01
(0.09)
0.11*
(1.68)
Yes
1.76***
(7.36)
also allows us to estimate the correlations between the technologies once the variables included in Table 4 have been controlled
for. These correlations are presented at the bottom of Table 4
and provide us with a way of testing Hypothesis 4. They are all
positive and signicant, suggesting that the arguments leading to
Hypothesis 4 are correct. However, these correlations could be
interpreted in two ways. First, a positive correlation could be due
to the inuence of rm-specic factors that are not included in our
estimations and that determine the propensity of some rms to
adopt all the technologies. Therefore, they provide us with a way
of assessing the inuence of unobservable rm effects. Second, a
positive correlation could mean that the technologies form part of
a system. Hence, the adoption of one technology would increase
the probability of the adoption of the other technologies of the
system.
One possibility to investigate Hypothesis 4 further is to reestimate our full model adding other technologies that we suspect
are not closely related to the four process technologies analysed.
In other words, the correlations between non-related technologies
should capture the inuence of non-observable rm-specic effects
related to the adoption of all the technologies. Fortunately, our data
set also contains information on the adoption of three e-business
technologies. From 2000 on, the survey includes a question on
whether the rm buys goods or services (providers) through the
Internet (S2B), whether it sells to nal consumers through the
Internet (B2C) and whether it sells to rms through the Internet (B2B). Given that the three technologies supporting these
systems are not obviously related to process technologies, a dramatic reduction in the value of the correlations should be observed.
Unfortunately, this signicantly restricts our observation window
Table 5
Correlation matrix between process and information technologies.
CNC
Robotics
CAD
FMS
B2B
B2C
S2B
***
CNC
Robotics
CAD
FMS
B2B
B2C
1
0.414***
0.470***
0.337***
0.105***
0.138***
0.080***
1
0.396***
0.357***
0.090***
0.130***
0.067***
1
0.436***
0.075***
0.009***
0.131***
1
0.122***
0.199***
0.178***
1
0.797***
0.373***
1
0.355***
1616
(to years 2002 and 2006) and the number of available observations
(to 2367).
The correlations resulting from estimating a full multivariate
probit model with the seven technologies are presented in Table 5.
The conclusion is that the correlations between process and ebusiness technologies are clearly lower than those within the
groups. The highest correlation between the two groups is 0.199
(B2B and FMS), the value is non-signicant in two of the cases
(CAD-B2B and CAD-B2C) and three correlations are below 0.10.
In order to make sure that these differences were signicant, we
compared all the correlations between in-rm and beyond-rm
technologies.21 For a given correlation between two technologies belonging to the in-rm technologies (for example, CNC and
Robotics), we tested whether the coefcient was different from that
of the in-rm technology and each of the beyond-rm technologies (i.e., we compared it with CNC-B2B, CNC-B2C and CNC-S2B).
The results (not shown) of all the comparisons are signicantly
different, showing that within-group correlations are higher than
inter-group correlations. In other words, our results conrm that
complementarities between related technologies exist and that
they are very different from the ones corresponding to unrelated
technologies, supporting Hypothesis 4.
5. Conclusions and implications
In this paper, we have focused on the different incentives
that rms have to adopt advanced technologies depending on the
possibilities of appropriating their returns. We have argued that
complementary resources are one of the mechanisms that rms
use to prot from the use of new technologies. One implication of
this hypothesis is that the use of AMT should be more likely in rms
that possess more complementary assets. The evidence presented
in this paper seems to conrm our conjecture: complementary
resources are associated with a rms technology use.
Clear differences in adoption behaviour are identied depending on the technologies analysed. More precisely, only technological resources are unequivocally related to the use of new
technologies in manufacturing rms. The fact that we have studied the use of AMT could help us to understand this result. These
technologies are related to product manufacturing and design.
Therefore, investments in the knowledge base of the rm could
have served to understand and rene both processes, detecting the
need and the role of new technologies. Our results seem to suggest that more complex technologies benet to a larger extent from
these investments, as proposed by Rogers and Shoemaker (1971).
Human and Marketing resources are also important, although they
present a weaker relation and different effects depending on the
technologies analysed. The quality of the services provided by
human resources is positively related to computer aided design
and exible manufacturing, whereas marketing investments are
only related to robotics and computer aided design. These results
are consistent with previous literature (Dunne and Troske, 2005;
Kotha and Swamidass, 2000). The interpretation of these results
has, however, to be taken with some caution, given that we lack
data on other characteristics of the technologies (see, for instance,
Rogers and Shoemaker, 1971) and that, at the time of the study, the
diffusion process had already started.
We also explain the use of advanced technologies through the
complementarities that arise within systems of technologies. In
fact, the interrelations between some of the technologies are the
main reason that we observe positive correlations between them,
once we control for the determinants of adoption. This supports the
21
We followed Buis (2011) in order to get access to the correlations. We thank
Maarten Buis and Stephen Jenkins for their advice.
1617
Table A1
Descriptive statistics of dependent and independent variables.
1. CNC
2. Robotics
3. CAD
4. FMS
5. B2B
6. B2C
7. S2B
8. Firm size
9. Firm debt ratio
10. Stock of technological resources
11. Human resources
12. Stock of marketing resources
13. Export propensity
14. Market concentration
15. Integrated in business group
16. Foreign capital
Mean
Observations
St. Dev.
Min.
Max.
0.45
0.26
0.35
0.24
0.08
0.05
0.19
0.23
0.56
0.01
0.11
0.03
0.19
0.41
0.33
0.18
4418
4418
4418
4418
2367
2367
2367
4418
4418
4418
4418
4418
4418
4418
4418
4418
0.50
0.44
0.48
0.43
0.27
0.21
0.39
0.62
0.23
0.04
0.13
0.13
0.26
0.13
0.47
0.39
0
0
0
0
0
0
0
0.001
0
0
0
0
0
0.20
0
0
1
1
1
1
1
1
1
14.20
1
0.65
1
4.32
1
0.95
1
1
technologies could imply that all the rms reduce their initial levels
of protability if they are not able to compensate for them.
Acknowledgements
We acknowledge nancial support from the Spanish Ministry
of Science and Technology and FEDER (projects ECO2008-04129
and ECO2011-22947), the Regional Government of Aragn (project
S09/PI138-08) and Universidad de La Rioja (projects API 19/08, API
11/25 and EGI 11/29). We are grateful for the comments and sugges
tions provided by Dolores Ann,
Sergio Palomas, two anonymous
referees, the Editor of this Journal, participants at the Conferences
of ACEDE, IAMB, EURAM and at seminars held at the Universities
of Zaragoza, La Rioja and Rey Juan Carlos. A previous version of
this paper was included in the Working Paper Series of the Spanish Fundacin de las Cajas de Ahorros (FUNCAS) (document number
505/2010).
Appendix A. Descriptive statistics
(See Table A1)
References
Adams, J.D., 1999. The structure of rm R&D, the factor intensity of production, and
skill bias. The Review of Economics and Statistics 81 (3), 499510.
Amara, N., Landry, R., Becheikh, N., Ouimet, M., 2008. Learning and novelty of innovation in established manufacturing SMEs. Technovation 28, 450463.
Ansoff, I., 1965. Corporate Strategy: An Analytic Approach to Business Policy for
Growth and Expansion. McGraw-Hill, New York.
Argote, L., Beckman, S.L., Epple, D., 1990. The persistence and transfer of learning in
industrial settings. Management Science 36, 140154.
Arvanitis, S., Hollenstein, H., 2001. The determinants of the adoption of advanced
manufacturing technology. An empirical analysis based on rm-level data
for Swiss manufacturing. Economics of Innovation and New Technology 10,
377414.
Arvanitis, S., Louikis, E., 2009. A comparative study of the effect of the ICT, organizational and human capital on labour productivity in Greece and Switzerland.
Information Economics and Policy 21, 4361.
Astebro, T., 2002. Noncapital investment costs and the adoption of CAD and CNC in
U.S. metalworking industries. Rand Journal of Economics 33 (4), 672688.
Astebro, T., Colombo, M.G., Seri, R., 2005. The Diffusion of Complementary
Technologies: An Empirical Test (downloaded on 8 November 2009 from
http://ssrn.com/abstract=690981).
Balakrishnan, S., Fox, I., 1993. Asset specicity, rm heterogeneity and capital structure. Strategic Management Journal 14, 316.
Baldwin, J., Diverty, B., 1995. Advanced Technology Use in Canadian Manufacturing Establishments. Research Paper, 85. Analytical Studies Branch.
Statistics Canada, Ottawa (downloaded on 8 November 2009 from
http://publications.gc.ca/collections/Collection/CS11-0019-85E.pdf).
Baptista, R., 2000. Do innovations diffuse faster within geographical clusters? International Journal of Industrial Organisation 18, 515535.
Barbosa, N., Faria, A.P., 2008. Technology adoption: does labour skill matter? Evidence from Portuguese rm-level data. Empirica 35, 179194.
1618
Barney, J., 1991. Firm resources and sustained competitive advantage. Journal of
Management 17 (1), 99120.
Barney, J.B., 1986. Strategic factors markets: expectations, luck and business strategy. Management Science 32 (10), 12311241.
Bartoloni, E., Baussola, M., 2001. The determinants of technology adoption in Italian
manufacturing industries. Review of Industrial Organization 19, 305328.
Battisti, G., Canepa, A., Stoneman, P., 2009. E-Business usage across and within rms
in the UK: protability, externalities and policy. Research Policy 38 (1), 133
143.
Beaumont, N., Schroder, R., Sohal, A., 2002. Do foreign-owned rms manage
advanced manufacturing technology better? International Journal of Operations
and Production Management 22 (7), 759771.
Beede, D.N., Young, K.H., 1996. Patterns of advanced technology adoption and manufacturing performance: employment growth, labor productivity and employee
earnings. Research Paper: Economics and Statistic Administration, Ofce of Policy Development, Washington, DC.
Belderbos, R., Carree, M., Diederen, B., Lokshin, B., Veugelers, R., 2004. Heterogeneity
in R&D cooperation strategies. International Journal of Industrial Organization
22, 12371263.
Belderbos, R., Carree, M., Lokshin, B., 2006. Complementarity in R&D cooperation
strategies. Review of Industrial Organization 28, 401426.
Bernard, A.B., Jensen, J.B., 1999. Exceptional exporter performance: cause, effect, or
both. Journal of International Economics 47, 125.
Blois, K.J., 1985. Marketing strategies and the new manufacturing technologies. International, Journal of Operations and Production Management 6 (1),
3441.
Bosworth, D., 1996. Determinants of the use of advanced technologies. International
Journal of the Economics of Business 3 (3), 269293.
Buis, M., 2011. Stata tip 97: getting at s and s. The Stata Journal 11 (2), 13.
Canepa, A., Stoneman, P., 2005. Financing constraints in the inter rm diffusion of new process technologies. The Journal of Technology Transfer 30,
159169.
Cappellari, L., Jenkins, S.P., 2003. Multivariate probit regression using simulated
maximum likelihood. The Stata Journal 3 (3), 278294.
Cassiman, B., Veugelers, R., 1999. Make and buy in innovation strategies: evidence
from Belgian manufacturing rms. Research Policy 28 (1), 6380.
Castellaci, F., 2002. Technology gap and cumulative growth: models and outcomes.
International Review of Applied Economics 16 (3), 333346.
Clemons, E.K., Kimbrough, S., 1986. Information systems, telecommunications, and
their effect on industrial organization. In: Proceedings of the 7th International
Conference on Information Systems, San Diego, CA.
Clemons, E.K., Row, M.C., 1991. Sustaining IT advantage. The role of structural differences. MIS Quarterly 15 (3), 275292.
Cohen, B., 1975. Multinational Firms and Asian Exports. Yale Univ. Press, New Haven,
CT.
Cohen, W.M., Levinthal, D.A., 1990. Absorptive capacity: a new perspective on learning and innovation. Administrative Science Quarterly 35, 128152.
Collins, D.J., Montgomery, C.A., 1998. Creating Corporate Advantage Harvard Business Review 76 (3), 7083.
Colombo, M.G., Mosconi, R., 1995. Complementary and cumulative learning effects
in the early diffusion of multiple technologies. Journal of Industrial Economics
43, 1348.
Conner, K.R., Prahalad, C.K., 1996. A resource-based theory of the rm: knowledge
versus opportunism. Organizational Science 7 (5), 477501.
Denrell, J., Fang, C., Winter, S.G., 2003. The economics of strategic opportunity.
Strategic Management Journal 24, 977990.
Dierickx, I., Cool, K., 1989. Asset stock accumulation and sustainability of competitive
advantage. Management Science 35 (12), 15041511.
Doms, M., Dunne, T., Troske, K., 1997. Workers, wages and technology. The Quarterly
Journal of Economics 112 (1), 253290.
Dunne, T., Troske, K., 2005. Technology adoption and the skill mix of US manufacturing plants. Scott Journal of Political Economy 52 (3), 387405.
Eaton, J., Kortum, S., 2002. Technology, geography, and trade. Econometrica 70,
17411779.
Ennen, E., Richter, A., 2010. The whole is more than the sum of its parts- or is
it? A review of the empirical literature on complementarities in organizations.
Journal of Management 36 (1), 207233.
Faria, A., Fenn, P., Bruce, A., 2003. A count data model of technology adoption. Journal
of Technology Transfer 28, 6379.
Fuentelsaz, L., Gmez, J., Palomas, S., 2009. The effects of new technologies on
productivity: an intrarm diffusion-based assessment. Research Policy 38,
11721180.
Gmez, J., Vargas, P., 2009. The effect of nancial constraints, absorptive capacity and
complementarities on the adoption of multiple process technologies. Research
Policy 38 (1), 106119.
Greve, H., 2009. Bigger and safer: the diffusion of competitive advantage. Strategic
Management Journal 30 (1), 123.
Grifth, R., Harrison, R., Van Reenen, J., 2006. How special is the special relationship?
Using the impact of U.S. Spillovers on U.K. rms as a test of technology sourcing.
The American Economic Review 96 (5), 18591875.
Griliches, Z., 1981. Market value, R&D and patents. Economic Letters 7, 183187.
Griliches, Z., Hall, B.H., Pakes, A., 1981. R&D, patents and market value revisited: is
there a second technological opportunity factor? Economics of Innovation and
New Technology 1, 183202.
Grossman, G., Helpman, E., 1991. Innovation and Growth in the World Economy.
MIT Press, Cambridge, MA.
Hannan, T.H., McDowell, J.M., 1987. Rival precedence and the dynamics of technology adoption: an empirical analysis. Economica 54, 155171.
Hill, C.W.L., 1988. Differentiation versus low cost or differentiation and low cost: a
contingency framework. Academy of Management Review 13, 401412.
Hirchey, M., Weygandt, J.J., 1985. Amortization policy for advertising and
research and development expenditures. Journal of Accounting Research 23,
326335.
Huffman, W.E., 1974. Decision making: the role of education. American Journal of
Agricultural Economics 56 (1), 8597.
Karshenas, M., Stoneman, P., 1993. Rank, stock, order, and epidemic effects in the
diffusion of new process technologies: an empirical model. Rand Journal of
Economics 24 (4), 503528.
Keller, W., 2004. International technology diffusion. Journal of Economic Literature
42, 752782.
Koellinger, P., 2008. The relationship between technology, innovation, and rm performance empirical evidence form e-business in Europe. Research Policy 37,
13171328.
Kogut, B., Zander, U., 1992. Knowledge of the rm, combinative capabilities, and
replication of technology. Organization Science 3 (3), 383397.
Kotha, S., Swamidass, P.M., 2000. Strategy, advanced manufacturing technology
and performance: empirical evidence from U.S. manufacturing rms. Journal
of Operations Management 18 (2), 257277.
Lal, K., 1999. Determinants of the adoption of Information Technology: a case study
of electrical and electronic goods manufacturing rms in India. Research Policy
28, 667680.
Lal, K., 2002. E-business and manufacturing sector: a study of small and mediumsized enterprises in India. Research Policy 31, 11991211.
Luque, A., 2002. An option value approach to technology adoption in U.S. manufacturing: evidence from microdata. Economics of Innovation and New Technology
11 (6), 543568.
1619
Villalonga, B., 2004. Intangibles resources, Tobins q, and the sustainability of performance differences. Journal of Economic Behaviour and Organization 54 (2),
205230.
Wernerfelt, B., 1984. A resource based view of the rm. Strategic Management Journal 5 (2), 171180.
Winter, S., 1987. Knowledge and competence as strategic assets. In: Teece, D. (Ed.),
The Competitive Challenge. Center for Research in Management, Berkeley, CA,
pp. 159184.
Wozniak, G.D., 1984. The adoption of interrelated innovations: a human capital
approach. The Review of Economics and Statistics 66 (1), 7079.
Wuyts, S., Dutta, S., Stremersch, S., 2004. Portfolios of interrm agreements in
technology-intensive markets: consequences for innovation and protability.
Journal of Marketing 68, 88100.