Econ Polit Ind (2015) 42:323–341

DOI 10.1007/s40812-015-0015-4

How do new entrepreneurs innovate?

Gabriele Pellegrino1 • Mariacristina Piva2 •

Marco Vivarelli3,4

Received: 12 November 2014 / Revised: 16 April 2015 / Accepted: 7 June 2015 /

Published online: 24 June 2015
 Associazione Amici di Economia e Politica Industriale 2015

Abstract This paper builds upon Pellegrino et al. (Struct Chang Econ Dyn
23:329–340, 2012) further analysing the determinants of product innovation in
Italian young innovative companies (YICs) by looking at in-house and external
R&D and at the acquisition of external technology in its embodied and disembodied
components. A Tobit approach is applied to study jointly the occurrence of product
innovation and the intensity of such innovation. Results provide evidence that in-
house R&D is linked to product innovation both in mature firms and YICs; however,
YICs turn out to be less in-house R&D-based and, unlike their mature counterparts,
more dependent on external sources of knowledge. While this outcome corroborates
and further reinforce what found—using a different methodology—in Pellegrino
et al. (Struct Chang Econ Dyn 23:329–340, 2012), in this study, other entrepre-
neurial attitudes such as the ability to cooperate with other firms in producing
innovation or the capacity to develop significant organizational changes are also
investigated. The results of the econometric estimations show that these attitudes
turn out to be key innovative strategies in the incumbent firms but, in some specific
cases (such as for the ability to cooperate), appear to be far less important in the
YICs. These results are somehow worrying, since they show that Italian innovative
entrepreneurs are mostly driven by routinized rather than creative strategies.

Keywords YICs  Entrepreneurship  R&D  Product innovation

& Mariacristina Piva

[email protected]
1
SPRU (Science and Policy Research Unit), University of Sussex, Brighton, UK
2
Dipartimento di Scienze Economiche e Sociali, Università Cattolica del Sacro Cuore, Via
Emilia Parmense 84, I-29122 Piacenza, Italy
3
Università Cattolica del Sacro Cuore, Milan, Italy
4
IZA, Bonn, Germany

123
324 Econ Polit Ind (2015) 42:323–341

JEL Classification L26  O31

1 Introduction

Both the academic community and policy makers have shown increased awareness
of the role of young innovative companies (YICs) run by entrepreneurs who should
contribute to the renewal of the industrial structure and ultimately to economic
growth.1 Following this approach, one of the possible explanations of the
productivity gap among US and Europe could depend on the revealed capacity of
the US economy to generate an increasing flow of young innovative firms which
survive introducing new products and gaining a place at the core of emerging
sectors.2 On the contrary, young European firms show a lower innovative capacity
and a higher business failure rate, not contributing to the alleged positive innovative
industrial dynamics (see Bartelsman et al. 2004; Santarelli and Vivarelli 2007;
Vivarelli 2013). In this context, the European economy would appear to lack
innovative and creative founders, who are the core of the so-called ‘entrepreneurial
society’ (Audretsch and Thurik 2000; Audretsch 2007).
When deciding their ‘Knowledge Production Function’ (see Sect. 2), innovative
entrepreneurs face different options: as well as in-house and external R&D
activities, technological acquisition (TA) in its embodied (machinery and equip-
ment) and disembodied components has to be taken into account. Here we build
upon Pellegrino et al. (2012) in which an empirical investigations was carried out to
test whether R&D and TA lead to significant differences in determining innovative
output in firms of different ages, thus providing evidence on whether the KPF of
YICs exhibits some peculiarities in comparison with what emerges in the case of
mature incumbent firms. The results of the empirical analysis showed that in-house
R&D is linked to the propensity to introduce product innovation both in mature
firms and YICs; however, innovation intensity in the YICs appears to be mainly
dependent on embodied technical change from external sources, while in-house
R&D does not play a significant role. Moreover, other interesting differences
between mature and young firms emerged from the econometric analysis. In
particular, while exporting and science-based YICs seem to be more likely to
perform better in terms of innovative intensity, these type of companies, in contrast
with their mature counterparts, do not seem to be established enough to be
responsive to factors such as appropriability conditions and other forms of
innovation activity (such us strategic or managerial form of innovation).
Following these insights, a first research aim of this paper is to further test the
main results of Pellegrino et al. (2012) that YICs are less R&D based and more
dependent on external sources of knowledge than their mature counterparts. The
robustness of this result will be inspected by using a different econometric strategy
1
For example, several EU member states have introduced new measures to support the creation and
growth of YICs, especially by improving their access to funding (see BEPA 2008; Schneider and
Veugelers 2010).
2
For complementary interpretations of the transatlantic productivity gap, see Ortega-Argilés et al. (2011,
2014).

123
Econ Polit Ind (2015) 42:323–341 325

and different specifications. We will then proceed to the main contribution of this
paper, which is to the test whether additional differences emerge between YICs and
mature incumbents with respect to the impact that other entrepreneurial character-
istics have in determining the occurrence of product innovation and the intensity of
such innovation. More in detail, the focus will be on factors such as risk aversion,
attitude towards organizational change and the capacity to develop cooperative
innovation. In this respect, taking into account that innovation is a costly and
uncertain activity, it is reasonable to expect that firms in general, and, YICs—due to
their inherent characteristics—in particular, could be significantly obstructed in
their innovative attempt by their own risk aversion. Indeed, since risk aversion
negatively affects entrepreneurship, this type of analysis will allow us to provide
insights about the role of entrepreneurship in shaping innovative performance.
Furthermore, taking into consideration that cooperative innovation is regarded as a
key factor in enhancing the firm’s innovative performance (see Cassiman and
Veugelers 2002; Cefis et al. 2009) we will test whether this entrepreneurial ability is
more important in determining the firm’s performance of YICs rather than mature
firms.
The rest of the paper is structured as follows: a discussion of the reference
literature is presented in Sect. 2, whereas the description of the data used in the
empirical analysis follows in Sect. 3. Subsequently, the econometric results are
displayed and discussed in Sect. 4. Section 5 concludes the paper by briefly
summarising the main findings and suggesting policy implications.

2 The literature

The first contributions to introduce the innovative input–output relationship were

put forward by Griliches (1979, 1990), by the means of a three-equation model in
which one of the equations is called Knowledge Production Function (KPF), a
function representing the transformation process from innovative inputs (R&D) to
innovative outputs (patents). The KFP is also included in the models provided by
Crèpon et al. (1998) and Lööf and Heshmati (2001). However, in most of these
previous empirical studies, the KPF is simplified as a link between R&D and
patents.
Historically driven by relative data availability, the relationship between a firm’s
R&D investment and patenting activity leaves room for a more complete approach
to the determinants of innovation. Today, innovation surveys offer more compre-
hensive measures of both innovative inputs and outputs.
Consistently, different innovation outputs, such as product and process innova-
tion, can be seen as the outcomes of several innovation inputs. Beside the formal
R&D investment,3 technological acquisition plays a role through ‘embodied

3
Methodologically, this is well represented by the shift from the R&D-focused Frascati Manual
(‘Guidelines for the collection of R&D data’, first published in 1963) to the Oslo Manual, published in the
1990 s (OECD 1997).

123
326 Econ Polit Ind (2015) 42:323–341

technical change’4 acquired by means of investment in new machinery and

equipment, and through the purchasing of external technology incorporated in
licences, know-how and consultancies (Freeman 1982; Freeman et al. 1982;
Freeman and Soete 1987). Once it has been recognized that innovative inputs are
not confined to formal R&D and that innovative outputs can be measured by
indicators other than patent, we pave the way for a deeper analysis of peculiarities in
the KPF.5 In particular, when innovation is carried out by an entrepreneur leading a
young firm, we can think of R&D as a creative input where endogenous
competences are fully deployed in generating product innovation (Teece et al.
1997; Von Tunzelmann and Wang 2003, 2007), while technological acquisition
appears to be more related to the implementation of external knowledge, with
replication and imitation playing a crucial role. Moreover, those entrepreneurs
relying more on R&D not only create value from their present capabilities but also
pave the way to better absorbing new ideas coming from the external environment
(the so-called ‘absorptive capacity’, see Cohen and Levinthal 1989, 1990).
Hence, as a first issue of investigation we will test again the results of our
previous work (Pellegrino et al. 2012) in which we found that innovative
entrepreneurs are less R&D based than their mature counterparts and basically
dependent on external knowledge provided by other firms and research institutions.
From a theoretical point of view, our previous findings are in contrast with the
Schumpeterian ‘creative destruction’ hypothesis (Schumpeter 1934; the so-called
Schumpeter Mark I) according to which newly established firms are the leading
innovators, while seem to be more in line with a process of ‘creative accumulation’,
where the leading contribution in the innovation process is played by the large and
more experienced incumbent firms (Schumpeter 1942; Schumpeter Mark II). In the
former context, an ‘entrepreneurial regime’ is at work (using an evolutionary
terminology), where innovative entrepreneurs are the main factors of change, while
the latter is a ‘routinized regime’, where larger and older incumbents are the engines
of change leading the innovative process (see Winter 1984; Malerba and Orsenigo
1996; Breschi et al. 2000).
Indeed, when as in this study we focus on all the industrial sectors and not only
on the emerging or the high-tech ones, several arguments support larger mature
firms being more R&D-based than their younger counterparts.6 First of all, mature
incumbents do not suffer from liquidity constraints as they generally have privileged
access to external finance and internal funds to support R&D activities. Secondly,
incumbent firms enjoy a higher degree of ‘appropriability’, as they usually possess
more market power (Gilbert and Newbery 1982). Finally, learning economies (see

4
The embodied nature of technological progress was originally discussed by Salter (1960) and Solow
(1960); in particular, vintage capital models describe an endogenous process of innovation in which the
replacement of old equipment is the main way through which firms update their own technologies (see
also Jorgenson 1966; Hulten 1992; Greenwood et al. 1997; Hercowitz 1998).
5
See Nelson and Winter (1982) and Dosi (1988) for an extended and more articulated view of the
innovative process across firms.
6
A related stream of literature instead focuses the attention on the so-called ‘New Technology Based
Firms’ (NTBFs, see Storey and Tether 1998; Colombo and Grilli 2005), where only YICs in the high-tech
sectors are analysed.

123
Econ Polit Ind (2015) 42:323–341 327

Arrow 1962; Malerba 1992) are often crucial in innovative dynamics, and younger
inexperienced entrepreneurs are obviously at a disadvantage from this perspective.
However, not all innovative firms are large established corporations. Indeed,
economic literature supports the hypothesis that new firms face a different
technological and economic environment from large mature incumbents with
respect to innovative activities (see Acs and Audretsch 1988, 1990; Acs et al. 1994).
Indeed, it may well be the case that entrepreneurial YICs establish their competitive
advantage on the basis of creative and R&D-based product innovation, which
significantly increases both their chances of survival and their economic perfor-
mance in comparison with less innovative start-ups (see Arrighetti and Vivarelli
1999; Michelacci 2003; Cefis and Marsili 2005).
In addition to the investigation of the peculiarities of the KPF in YICs, a second
key issue of interest in this work is to see whether other characteristics can
significantly affect firms’ overall innovative performance. In particular—taking into
account both the previous literature and data availability—we will assume product
innovation (both in terms of its occurrence and its intensity) as an indicator of
innovative performance, and we will assess the role of different determinants in
affecting the level of product innovation. The KPF baseline approach (see above)
will be complemented by the investigation of five additional factors, as follows.
Firstly, we will check the role of a firm’s size to see whether the Schumpeterian
hypothesis, which claims an advantage of larger firms in introducing innovation (see
the classical debate started by Schumpeter 1942; renewed by Arrow 1962 and more
recently continued in Cohen and Klepper 1996), is supported across both the
incumbents and the YICs.
Secondly, the role of sectoral belonging will be studied using Pavitt’s taxonomy
(see Pavitt 1984; Malerba and Orsenigo 1996; Malerba 2005). With regard to YICs,
it will be interesting to see whether the ‘science-based’ and the ‘specialised
supplier’ (the high-technology groups in Pavitt’s taxonomy) young firms enjoy a
relative advantage in developing their innovative products.
Thirdly, we will test the role of risk aversion in deterring entrepreneurial
innovative behaviour (see Kihlstrom and Laffont 1979; Palich and Bagby 1995;
Parker 2004; Kan and Tsai 2006): since innovation is a costly and uncertain activity,
are firms—especially YICs—limited by their own risk aversion? Since risk aversion
and entrepreneurship are inversely correlated, this will be a first direct way to test
the role of entrepreneurship in shaping innovative performance (both in general and
with specific reference to the young innovative companies).
Fourthly, organizational change will be considered as a second indicator of
entrepreneurial capability.7 On the one hand, many scholars have investigated the
complementarity between technological and organizational change (see, for
instance, Bresnahan et al. 2002; Hitt and Brynjolfsson 2002; Piva et al. 2005).
On the other hand, the role of an entrepreneur is precisely that of creatively
combining the different factors of production (see Kirzner 1997). Thus,

7
In this case, in contrast with risk aversion, organizational change is positively correlated with the
entrepreneurial ability.

123
328 Econ Polit Ind (2015) 42:323–341

entrepreneurial firms, able to introduce organizational changes (specifically,

entrepreneurial YICs) should be better positioned to generate product innovation.
Fifthly, a third indicator of entrepreneurship adopted in this study is the ability to
cooperate with other firms in joint innovative activities. Cooperative innovation has
indeed been shown to be crucial in determining better innovative performance
across firms (see Cassiman and Veugelers 2002; Piga and Vivarelli 2003, 2004;
Fritsch and Franke 2004; Parker 2008; Cefis et al. 2009). Here we will see whether
this entrepreneurial ability turns out to be significant in explaining the differences in
innovative performance across firms and, more specifically, across YICs.
Summing up, this paper will investigate the innovation strategies adopted by
entrepreneurs in the initial stages of their firm’s life cycle, and will compare them
with what is done by mature older incumbents. In line with our previous work
(Pellegrino et al. 2012) the first hypothesis is that YICs turn out to be less R&D
intensive than their incumbent counterparts. The second hypothesis,—whose
empirical test represents the main contribution of the present work—is that size,
sectoral belonging and various entrepreneurial attitudes may play a different role
when comparing young firms with established companies.

3 Database, variables and methodology

3.1 Database

The empirical analysis was carried out using firm-level data from the third Italian
Community Innovation Survey (CIS3),8 conducted over the 1998–2000 period by
the Italian National Institute of Statistics (ISTAT). This survey is representative at
both sector and firm size level of the entire population of Italian firms with more
than 10 employees (ISTAT 2004).9
The response rate was 53 %, determining a full sample size of 15,512 firms,
9,034 of which (58.24 %) in the manufacturing sector, our focus of attention. The
manufacturing sample was then cleaned of outliers and firms involved in mergers or

8
Given the aims and scope of this paper, attention has been limited to the manufacturing sectors.
9
Firm selection was carried out through a ‘one step stratified sample design’. The sample in each stratum
was selected with equal probability and without reimmission. The stratification of the sample was based
on the following three variables: firm size, sector, regional location. Technically, in the generic stratum h,
the random selection of n_{h} sample observations among the N_{h} belonging to the entire population
was realized through the following procedure:
- A random number in the 0–1 interval was attributed to each Nh population unit;
- Nh population units were sorted by increasing values of the random number;
- Units in the first nh positions in the order previously mentioned were selected.
Estimates obtained from the selected sample are very close to the actual values in the national
population. The weighting procedure follows Eurostat and Oslo Manual (OECD 1997) recommendations:
weights indicate the inverse of the probability that the observation is sampled. Therefore, sampling
weights ensure that each group of firms is properly represented and correct for sample selection.
Moreover, sampling weights help to reduce heteroscedasticity commonly arising when the analysis
focuses on survey data.

123
Econ Polit Ind (2015) 42:323–341 329

acquisitions during the previous 3 years, which would have biased our results.10 We
thus ended up with 7,965 innovating and not-innovating firms.
The sub-sample of innovators was then selected following the standard practice
of identifying innovators as those firms declaring that in the previous 3 years they
had introduced either product or process innovations, or had started innovative
projects (then dropped or still-to-complete at December 31st, 2000). The same
definition was implemented by ISTAT as a filter to single out non-innovators that
were allowed to skip a large number of ‘innovation questions’, leaving us with very
little information about their propensity to innovate or to invest in innovative inputs.
This means that the CIS database provides information relevant to this study only
for innovative firms; therefore only these firms have been considered in the
following analysis,11 ending up with 3,045 firms. This sample was further reduced
to 2,713 firms by keeping only firms investing in at least one of the four innovative
inputs we focus on and whose age was available. Finally, YICs were identified as
innovative firms which had been operational for less than eight years (293 out of
2,713).12

3.2 Innovative variables

Innovative variables capture innovative output and innovative inputs.

With regard to innovative outputs they can be distinguished with respect to their
position in the innovation process. For instance, while patents are better defined as
the outcome of the inventive process, product innovation represents the result of the
market-oriented innovative process. However, even though product innovation is
driven by demand considerations, it represents a pre-market result. In contrast, the
share of sales deriving from innovative products (Lööf and Heshmati 2002;
Mairesse and Mohnen 2002), the intensity of innovation, represents an ex-post result
in which the market has positively welcomed the new products introduced by the
firm (Barlet et al. 2000). This paper uses the ex-post result as the output indicator for
the empirical analysis, i.e. the share of turnover (sales) derived from innovative

10
In fact, mergers and acquisitions may break the link between innovative inputs and outputs (a link that
must be studied within the context of the same economic entity over time).
11
Given that our aim is to analyze the nature of the relationships within the innovative process (and not,
for example, the effect of different inputs in determining the probability of innovating), this data
limitation does not raise a problem of selection bias in our context. Since we are interested in the internal
mechanisms of the innovative process, we have to focus on a randomly-selected sample of innovative
firms (that is, randomness must hold within the innovative sub-sample, not in comparison with the non-
innovative one where such mechanisms are obviously absent). For a study based on a comparison
between innovative and non-innovative Italian firms, see Parisi et al. 2006. Moreover, as Mairesse and
Mohnen (2010) pointed out, since CIS data provides little information about non-innovating firms, in the
absence of additional information about these firms obtained by merging the innovation survey data with
other firm data, not much room is left to distinguish between innovators and non-innovators and to correct
appropriately for potential selectivity biases.
12
As far as the age of the firms in the ‘young firms’ sub-sample is concerned, the threshold of 8 years
was chosen to take into account the trade-off between a lower age and the representativeness of the sub-
sample of YICs (here more than 10 % of the entire sample). However, the estimates discussed in Sect. 4
were replicated using a larger sample of young firms no more than 10 years old. The results, available
from the authors upon request, do not change substantially.

123
330 Econ Polit Ind (2015) 42:323–341

products (TURNIN). This is the only continuous output indicator provided by the
CIS. Finally, it is also important to note that product innovators are a subsample of
the innovative firms considered in this study, since they do not include those firms
only engaged in process innovation or those involved in potential innovative
projects. As a consequence, our TURNIN indicator is a double-censored variable
with a mass of values equal to zero.
Looking at the innovative inputs, four innovative inputs are used in this paper:

– In-house formal Research and Development (intra muros R&D = IR);

– Research and Development outsourced to other firms or research institutes
(extra muros R&D = ER);
– Expenditures in embodied technological change (innovative investment in
equipment and machinery = MAC);
– Expenditures in technology acquisition (disembodied technology such as know-
how, projects and consultancies, licenses and software = TA).

3.3 Other characteristics/variables

Taking into account the reference literature and the hypotheses discussed in Sect. 2,
attention will be paid to the following additional variables:

• Firm size, measured by the number of employees (SIZE), in order to test the
Schumpeterian hypothesis;
• As discussed in Sect. 2, the important role of sectoral belonging will be tested
using Pavitt’s sectoral dummies, controlling for the different sectoral techno-
logical opportunity and appropriability conditions;13
• Turning our attention to the entrepreneurial variables, RISK will measure risk
aversion using a YES/NO (1/0) questionnaire reply centered on the role of
perceived risk as an important obstacle to innovative activities;14
• The entrepreneurial attitude towards organizational change will be implemented
through the dummy ORG, assuming value 1 when the innovative firm has
introduced a significant organizational change at the strategic, management or
shopfloor level;
• Finally, the firm’s attitude towards cooperation will be measured by the dummy
COOP, assuming value 1 when the innovative firm is engaged in innovative
cooperation with other firms.

13
The estimates will include three groups: science-based, specialised supplier and scale intensive firms,
where the default category will be the low-technology group of the supplier dominated firms.
14
We are aware of the limits deriving by the use of this variable as a proxy of risk aversion. However,
taking into account the information provided by the CIS questionnaire, this variable represents the most
accurate proxy at our disposal of the degree of firm’s risk aversion.

123
Econ Polit Ind (2015) 42:323–341 331

Table 1 The variables

TURNIN Share of firm’s total sales due to sale of new products

IR Internal R&D expenditure in 2000, normalized by total turnover

ER External R&D expenditure in 2000, normalized by total turnover
MAC Investments in innovative machinery and equipment in 2000, normalized by total turnover
TA Technological acquisitions in 2000, normalized by total turnover
SIZE Number of employees in 2000
SB Dummy = 1 if science-based firm
SI Dummy = 1 if scale intensive firm
SS Dummy = 1 if specialized supplier firm
RISK Dummy = 1 if firm has perceived high economic risk from the decision to innovate
ORG Dummy = 1 if the firm has realized managerial, strategic or organizational innovation
COOP Dummy = 1 if the firm takes part in cooperative innovative activities

Table 2 Descriptive statistics

All firms Mature firms Young firms (YICs)

2,713 OBS 2,420 OBS 293 OBS

Mean SD Mean SD Mean SD

TURNIN 0.30 0.29 0.30 0.29 0.34 0.32

IR 0.013 0.026 0.013 0.025 0.014 0.032
ER 0.002 0.009 0.002 0.008 0.002 0.011
MAC 0.035 0.078 0.034 0.076 0.042 0.091
TA 0.002 0.018 0.002 0.017 0.004 0.023
SIZE 175.023 633.797 182.629 666.530 112.201 214.542
SB (dummy) 0.116 0.320 0.113 0.316 0.140 0.347
SI (dummy) 0.284 0.451 0.282 0.450 0.300 0.459
SS (dummy) 0.280 0.449 0.282 0.450 0.266 0.443
RISK (dummy) 0.544 0.498 0.545 0.498 0.539 0.499
ORG (dummy) 0.721 0.449 0.714 0.452 0.778 0.416
COOP (dummy) 0.161 0.368 0.162 0.369 0.150 0.358

The summary Table 1 describes the variables used in the empirical analysis,
while Table 2 reports the corresponding descriptive statistics, distinguishing
between all firms, mature firms and YICs.15
Table 3 reports the sectoral compositions of the two subsamples of mature and
young firms; as can be seen, with regard to the four Pavitt (1984) categories, no
15
In the ‘‘Appendix’’, Table 7 reports the correlation matrix; as can be seen, all the correlation
coefficients are less than 0.245, showing that data are not affected by serious collinearity problems.
Finally, Table 8 reports the CIS questions on the basis of which the variables were constructed.

123
332 Econ Polit Ind (2015) 42:323–341

Table 3 Sectoral composition and average employment of the firms belonging to the two subsamples:
Mature and Young firms
Industry pavitt taxonomy Mature firms Young firms (YICs)

N. of firms % Av. Emp N. of firms % Av. emp

Science-based (SB) 273 11.28 296.52 41 14 165.29

Scale intensive (SI) 683 28.22 192.74 88 30.03 95.02
Specialized suppliers (SS) 683 28.22 179.43 78 26.62 131.13
Supplier dominated (SD) 781 32.27 136.77 86 29.35 87.30
Sample 2,420 100 182.63 293 100 112.20

significant differences emerge; indeed, a slight over-representation of science-based

firms in the YIC subsample is compensated for by a lower presence of the
specialised supplier ones. On average, Italian YICs belong to the same sectors as
mature incumbents. Thus NTBFs do not represent the core of Italian YICs, and the
contribution to sectoral renewal by the new and young innovative firms appears
rather limited. Not surprisingly, YICs turn out to be relatively smaller (112
employees on average) than their older counterparts (183 employees).16

3.4 The econometric model

Equation (1) describes the general complete specification of the model:

TURNINi ¼ C X
þ b1 IRinti þ b2 ERinti þ b3 MACinti þ b4 TAinti þ b5 SIZEi
þ ck PAVITTki þ b6 RISKi þ b7 ORGi þ b8 COOPi þ ei
ð1Þ
where C is the constant, i is the firm-index, TURNIN represents the innovative
output in terms of the percentage of sales due to innovative products, IR, ER, MAC
and TA indicate the innovative inputs we are interested in, SIZE, RISK, ORG and
COOP are the variables we want to check for and PAVITT are the sectoral dummies
(k = 3). Consistently with the dependent variable, the four innovative inputs were
normalized by sales; this makes the inputs homogeneous to the output.
Dealing with a zero-inflated censored variable, estimates were run as Tobit
regressions.

4 Econometric results

Table 4 reports the econometric results of the Tobit model applied to the entire
sample and separately to the two sub-samples of the mature incumbents and the
YICs. This first baseline specification only reports the four knowledge inputs and

16
As discussed at in Sect. 3, the CIS3 data adopted are collected from a representative sample of Italian
manufacturing firms with more than 10 employees; this means that micro firms (which however are very
rarely innovative) are excluded from the dataset, while SMEs are fully included.

123
Econ Polit Ind (2015) 42:323–341 333

Table 4 Baseline specification

All firms Mature firms Young firms (YICs)

Dependent variable: TURNIN

Constant 0.19*** 0.19*** 0.20***
(14.73) (14.17) (4.60)
IR 2.17*** 2.25*** 1.81**
(7.97) (7.59) (2.54)
ER 0.91 0.56 2.23
(1.12) (0.62) (1.16)
MAC -0.18* -0.29*** 0.42*
(-1.88) (-2.89) (1.68)
TA -0.23 -0.33 0.27
(-0.56) (-0.74) (0.28)
SIZE 0.00 0.00 0.00
(1.22) (1.32) (0.29)
SB 0.12*** 0.10*** 0.22***
(4.78) (3.82) (2.94)
SI -0.03 -0.03 -0.05
(-1.49) (-1.35) (-0.90)
SS 0.10*** 0.10*** 0.09
(5.82) (5.51) (1.45)
N. of firms 2,713 2,420 293
Censored (TURNIN = 0) 615 550 65
Uncensored 2,098 1,870 228

t statistics in parentheses: * Significant at 10 %; ** 5 %; *** 1 %

the size and Pavitt controls. Before moving on to the discussion, it is worth pointing
out that our estimations are based on cross-sectional data, and most of the regressors
considered are simultaneously determined and suffer from common method biases;
therefore interpretation of the results has to be undertaken with caution.
Focusing the attention to the four variables identifying the different innovative
inputs, while some results are in line with the main findings in Pellegrino et al.
(2012) some interesting dissimilarities emerge. In particular, neither external
research (ER) nor technological acquisition (TA) seem to exert a significant role in
enhancing the innovative performance of mature and young firms. Moreover, as
expected, in-house R&D appears to be important in increasing product innovation
for the entire sample, the mature firms and the YICs Italian manufacturing firms;
although the coefficient is smaller in magnitude and less significant in the case of
the YICs. Indeed, R&D input is more directly related to product innovation, while
embodied technological change (MAC) is more linked to process innovation (see
Freeman 1982; Freeman and Soete 1987).17

17
This also explains the negative and significant coefficient of MAC in the estimate referring to the
incumbents (second column of Table 4).

123
334 Econ Polit Ind (2015) 42:323–341

While this evidence is fully in line with our early findings (Pellegrino et al. 2012),
the results of the variable MAC differ notably. Indeed, unlike in Pellegrino et al.
(2012), where this innovative input exerts a relevant role in enhancing the innovative
performance of both mature and (in particular) young firms, in the present study the
coefficient of this variable appears to be negatively and highly significantly in the case
of mature firms and positive and barely significantly in the case of YICs.
All in all, this evidence further corroborates and reinforces (in the case of the variable
MAC) the main conclusion of our previous work (Pellegrino et al. 2012): Italian YICs—
far from being R&D-based NTBFs—are relatively biased in favour of embodied
technological change and less R&D intensive than their older counterparts, clearly
demonstrating a lack of creativity and autonomy in shaping their innovative KPFs.
Turning our attention to size and sectoral controls, the Schumpeterian hypothesis
is not supported by our estimates, the relative coefficient not being significantly
different from zero. Not surprisingly, the science-based and specialised supplier
firms (the two high-tech categories in Pavitt’s taxonomy) are significantly more
inclined to product innovation and this is true both for the entire sample and for the
mature firms. Interestingly enough, with regard to YICs, only the SB dummy turns
out to be significant, with a coefficient that is more than twice the corresponding one
for the mature firms. This means that for YICs it is even more important to belong to
the science-based category, in order to obtain an above-average innovative outcome.
This result makes the descriptive evidence reported in Table 3 even more worrying:
if the majority of Italian YICs were NTBFs belonging to the SB sectors (which is
not the case), their innovative performance would be significantly higher.
Table 5 presents the above specification extended to the entrepreneurial variables
discussed in the previous sections. First of all, all the results deriving from Table 4
above are fully confirmed and so will not be commented on further. As can be seen,
the variable RISK (although negative in sign, as expected) never turns out to be
even barely significant in any of the three regressions; hence, it seems that risk
aversion is not deterring Italian firms from being innovative.18
Shifting our attention to the ability to engage in various forms of organizational
change, no relevant differences emerge between young and mature firms. Indeed,
this variable turns out to be positively and significantly related to firms’ innovative
capacity, in all the three models.
Finally, cooperative agreements (COOP) in general turn out to affect product
innovation positively and significantly; however, this relationship is not significant
with regard to the YICs. This is a further disappointing result concerning the
entrepreneurial profile of Italian YICs; indeed, either they lack the endogenous
capabilities and ‘absorptive capacities’ to engage in effective innovative cooper-
ation, or they are unable to create value (in terms of product innovation) from such
cooperation.
As a further control, Table 6 reports the results from a restricted specification
where the non-significant regressors have been dropped (ER, TA, SIZE, RISK); as
can be noted, all the previous outcomes are fully confirmed.

18
However, this may simply be due to possible inaccuracy in the adopted proxy, the only one available
in our dataset.

123
Econ Polit Ind (2015) 42:323–341 335

Table 5 Extended Specification

All firms Mature firms Young firms (YICs)

Dependent variable: TURNIN

Constant 0.12*** 0.13*** 0.12*
(6.75) (6.66) (1.95)
IR 1.99*** 2.05*** 1.74**
(7.32) (6.93) (2.45)
ER 0.50 0.16 1.76
(0.62) (0.17) (0.90)
MAC -0.12 -0.24** 0.44*
(-1.30) (-2.34) (1.78)
TA -0.33 -0.46 0.24
(-0.82) (-1.01) (0.25)
SIZE -0.00 0.00 -0.00
(-0.10) (0.05) (-0.35)
RISK -0.01 -0.01 -0.02
(-0.76) (-0.55) (-0.50)
ORG 0.10*** 0.09*** 0.11**
(6.17) (5.71) (2.03)
COOP 0.08*** 0.08*** 0.09
(3.99) (3.81) (1.32)
SB 0.10*** 0.09*** 0.21***
(4.18) (3.28) (2.77)
SI -0.03 -0.03 -0.06
(-1.62) (-1.47) (-0.95)
SS 0.10*** 0.09*** 0.09
(5.31) (4.99) (1.46)
N. of firms 2,713 2,420 293
Censored (TURNIN = 0) 615 550 65
Uncensored 2,098 1,870 228

t statistics in parentheses: * Significant at 10 %; ** 5 %; *** 1 %

On the whole, our econometric results show that in comparison with the
incumbents, Italian YICs appear to be slightly less R&D-based, more dependent on
external sources of knowledge and lacking the ability to engage into fruitful
innovative agreements.

5 Conclusions and policy implications

This paper analyses the determinants of innovative output in both young and mature
Italian firms, by looking at firms’ internal and external R&D activities as well as at
the acquisition of external technology in its embodied and disembodied

123
336 Econ Polit Ind (2015) 42:323–341

Table 6 Restricted specification

All firms Mature firms Young firms (YICs)

Dependent variable: TURNIN

Constant 0.12*** 0.12*** 0.12*
(6.92) (6.92) (1.91)
IR 2.01*** 2.05*** 1.83***
(7.52) (7.09) (2.62)
MAC -0.12 -0.24** 0.47*
(-1.33) (-2.40) (1.91)
ORG 0.10*** 0.09*** 0.10*
(6.12) (5.67) (1.93)
COOP 0.08*** 0.08*** 0.10
(4.10) (3.89) (1.51)
SB 0.10*** 0.09*** 0.21***
(4.32) (3.35) (2.79)
SI -0.03 -0.03 -0.06
(-1.60) (-1.46) (-0.95)
SS 0.10*** 0.09*** 0.09
(5.37) (5.04) (1.45)
N. of firms 2,713 2,420 293
Censored (TURNIN = 0) 615 550 65
Uncensored 2,098 1,870 228

t statistics in parentheses: * Significant at 10 %; ** 5 %; *** 1 %

components. Moreover, the possible roles of size, three proxies of entrepreneurial

ability and sectoral belonging have been tested.
Overall, our findings further reinforce the main conclusion of Pellegrino et al.
(2012). Indeed, although based on a different econometric strategy and alternative
specifications, our results further support the view that the acquisition of external
sources of knowledge is particularly relevant in enhancing the innovative
performance of YICs. More in detail, looking at the whole sample, it turns out
that in-house R&D is linked to innovative performance, while external sources of
knowledge do not seem to play an important role in Italian manufacturing.
However, when the total sample is split in young and established firms, for the
former internal R&D expenditures play a slightly smaller role in increasing
innovation intensity, while the external acquisition of technology in its embodied
component achieves a certain significance. This evidence is even more relevant
taking into account that we find a negative and significant relationship between this
variable and the measure of firm’s innovative performance in the case of mature
firms.
Turning the attention to the relationship between entrepreneurial characteristics
and firm’s innovative performance, some interesting results emerge. Firstly, the
ability to engage in various forms of organizational changes appears to be equally

123
Econ Polit Ind (2015) 42:323–341 337

relevant in enhancing the firm’s innovative performance for both mature and young
firms. On the other hand, cooperative agreements (variable COOP) turn out to affect
product innovation positively and significantly for the sample of mature firms only;
this relationship being not significant with regard to the YICs. Finally, the variable
measuring firm’s risk aversion (RISK) never shows a significant link with the
dependent variable in any of the three regressions.
These results have relevant implications in terms of policy. Firstly they further
strengthen previous evidence supporting the vision that Italian YICs are far from to
be considered as R&D-based NTBFs. On the contrary, they appear to be rather weak
entrepreneurial entities which need to acquire external knowledge in order to foster
their own innovation activity. Furthermore, the complementary evidence related to
other entrepreneurial attitudes clearly demonstrate the difficulties faced by YICs in
engaging into creative and strategies, such as cooperative innovation.
All in all, these outcomes highlight a potential weakness of Italian YICs, which
seem to lack a fully-fledged endogenous capacity to sustain their own innovative
activities. In turn, this calls for an industrial and innovation policy able to foster
pure NTBFs, that is a policy encouraging a more creative behaviour based on
entrepreneurship and R&D-based innovation strategies.

Acknowledgments The authors would like to thank Andrea Conte, Giovanni Seri and the ADELE
Laboratory at ISTAT in Rome for the provision of CIS 3 data. Comments by the discussant Simon Parker
and the other participants at the ‘1st Joint DIW Berlin/IZA Workshop on Entrepreneurship Research’
(Bonn, February, 25–26, 2010) led to significant improvements to the paper.

Appendix

See Tables 7 and 8.

Table 7 Correlation matrix

TURNIN IR ER MAC TA SIZE RISK ORG COOP

TURNIN 1.000
IR 0.189 1.000
ER 0.086 0.245 1.000
MAC -0.034 -0.070 -0.046 1.000
TA -0.010 0.026 0.044 0.034 1.000
SIZE 0.034 0.052 0.054 -0.041 -0.008 1.000
RISK 0.123 0.173 0.168 -0.074 0.014 0.020 1.000
ORG 0.122 0.077 0.053 -0.086 0.031 0.095 0.090 1.000
COOP 0.006 0.035 0.062 -0.026 0.000 0.209 0.065 0.121 1.000

123
338 Econ Polit Ind (2015) 42:323–341

Table 8 The questionnaire

Innovative output variable: TURNIN
Estimate how your turnover in 2000 was distributed between:
New or significantly improved products (goods or services) introduced during the period 1998–2000
Unchanged or only marginally modified products (goods or services) during the period 1998–2000
Innovative input variables
Did your enterprise engage in the following innovation activities in 2000?:
IR: intramural research and All creative work undertaken within your enterprise on a
experimental development (R&D) systematic basis in order to increase the stock of knowledge,
and the use of this stock of knowledge to devise new
applications, such as new and improved products (goods/
services) and processes (including software research)
ER: acquisition of R&D (extramural Same activities as above, but performed by other companies
R&D) (including other enterprises within the group) or other public
or private research organisations
MAC: acquisition of machinery and Advanced machinery, computer hardware specifically
equipment purchased to implement new or significantly improved
products (goods/services) and/or processes
TA: acquisition of other external Purchase of rights to use patents and non-patented inventions,
knowledge licenses, know-how, trademarks, software and other types of
knowledge from others for use in your enterprise’s
innovations
SIZE What was your enterprise’s total number of employees in 1998
and 2000?
RISK During the period 1998–2000, how important were the
following factors as constraints to your innovation activities
or influencing a decision not to innovate? :
Excessive perceived economic risk
ORG Did your enterprise during the period 1998-2000 undertake any
of the following activities?:
Strategy (Implementation of new or significantly changed
corporate Strategies)
Management (Implementation of advanced management
techniques within your enterprise)
Organisation (Implementation of new or significantly changed
organizational structures)
COOP Did your enterprise have any co-operation arrangements on
innovation activities with other enterprises or institutions
during 1998-2000?

References
Acs, Z. J., & Audretsch, D. B. (1988). Innovation in large and small firms: An empirical analysis. The
American Economic Review, 78, 678–690.
Acs, Z. J., & Audretsch, D. B. (1990). Innovation and small firms. Cambridge: MIT Press.
Acs, Z. J., Audretsch, D. B., & Feldman, M. P. (1994). R&D spillovers and recipient firm size. The
Review of Economics and Statistics, 76, 336–340.
Arrighetti, A., & Vivarelli, M. (1999). The role of innovation in the post-entry performance of new small
firms: Evidence from Italy. Southern Economic Journal, 65, 927–939.

123
Econ Polit Ind (2015) 42:323–341 339

Arrow, K. (1962). The economic implications of learning by doing. Review of Economic Studies, 29,
155–173.
Audretsch, D. (2007). The entrepreneurial society. Oxford: Oxford University Press.
Audretsch, D., & Thurik, R. (2000). Capitalism and democracy in the 21st century: From the managed to
the entrepreneurial economy. Journal of Evolutionary Economics, 10, 17–34.
Barlet, C., Duguet, E., Encaoua, D., & Pradel, J. (2000). The commercial success of innovations: An
econometric analysis at the firm level in French manufacturing. In D. Encaoua, B. H. Hall, F.
Laisney, & J. Mairesse (Eds.), The economics and econometrics of innovation (pp. 435–456).
Boston: Kluwer Academic Publishers.
Bartelsman, E., Haltiwanger J., Scarpetta, S. (2004). Microeconometric evidence of creative destruction
in industrial and developing countries. IZA Discussion Paper No. 1374.
BEPA. (2008). Innovation and growth in the EU: The role of SME policy. Brussels: European
Commission.
Breschi, S., Malerba, F., & Orsenigo, L. (2000). Technological regimes and schumpeterian patterns of
innovation. The Economic Journal, 110, 388–410.
Bresnahan, T. F., Brynjolfsson, E., & Hitt, L. M. (2002). Information technology, workplace organization
and the demand for skilled labor: Firm-level evidence. Quarterly Journal of Economics, 117,
339–376.
Cassiman, B., & Veugelers, R. (2002). R&D cooperation and spillovers: Some empirical evidence from
Belgium. American Economic Review, 92, 1169–1184.
Cefis, E., & Marsili, O. (2005). A matter of life and death: Innovation and firm survival. Industrial and
Corporate Change, 14, 1167–1192.
Cefis, E., Rosenkranz, S., & Weitzel, U. (2009). Effects of coordinated strategies on product and process
R&D. Journal of Economics, 96, 193–222.
Cohen, W. M., & Klepper, S. (1996). A reprise of size and R&D. Economic Journal, 106, 925–951.
Cohen, W. M., & Levinthal, D. A. (1989). Innovation and learning: The two faces of R&D. Economic
Journal, 99, S569–S596.
Cohen, W. M., & Levinthal, D. A. (1990). Absorptive capacity: New perspective on learning and
innovation. Administrative Science Quarterly, 35, 128–152.
Colombo, M. G., & Grilli, L. (2005). Founders’ human capital and the growth of new technology-based
firms: A competence-based view. Research Policy, 34, 795–816.
Crépon, B., Duguet, E., & Mairesse, J. (1998). Research, innovation and productivity: An econometric
analysis at the firm level. Economics of Innovation and New Technology, 7, 115–158.
Dosi, G. (1988). Sources, procedures, and microeconomic effects of innovation. Journal of Economic
Literature, 26, 1120–1171.
Freeman, C. (1982). The economics of industrial innovation (2nd ed.). London: Pinter.
Freeman, C., Clark, J., & Soete, L. (1982). Unemployment and technical innovation: A study of long
waves in economic development. London: Pinter.
Freeman, C., & Soete, L. (1987). Technical change and full employment. London: Basil Blackwell.
Fritsch, M., & Franke, G. (2004). Innovation, regional knowledge spillovers and R&D cooperation.
Research Policy, 33, 245–255.
Gilbert, R., & Newbery, D. (1982). Preemptive patenting and the persistence of monopoly. The American
Economic Review, 72, 514–526.
Greenwood, J., Hercowitz, Z., & Krusell, P. (1997). Long-run implications of investment-specific
technological change. American Economic Review, 87, 342–362.
Griliches, Z. (1979). Issues in assessing the contribution of research and development to productivity
growth. The Bell Journal of Economics, 10, 92–116.
Griliches, Z. (1990). Patent statistics as economic indicators: A survey. Journal of Economic Literature,
28, 1661–1707.
Hercowitz, Z. (1998). The ‘embodiment’ controversy: A review essay. Journal of Monetary Economics,
41, 217–224.
Hitt, L. M., & Brynjolfsson, E. (2002). Information technology, organizational transformation, and
business performance. In N. Greenan, Y. L’Horty, & J. Mairesse (Eds.), Productivity, inequality,
and the digital economy. A transatlantic perspective (pp. 55–91). Cambridge: MIT Press.
Hulten, C. R. (1992). Growth accounting when technical change is embodied in capital. American
Economic Review, 82, 964–980.
ISTAT. (2004). Statistiche sull’Innovazione delle Imprese. Settore Industria. Anni 1998–2000, Rome:
ISTAT.

123
340 Econ Polit Ind (2015) 42:323–341

Jorgenson, D. W. (1966). The embodiment hypothesis. Journal of Political Economy, 74, 1–17.
Kan, K., & Tsai, W. D. (2006). Entrepreneurship and risk aversion. Small Business Economics, 26,
465–474.
Kihlstrom, R. E., & Laffont, J. J. (1979). A general equilibrium entrepreneurial theory of firm formation
based on risk aversion. Journal of Political Economy, 87, 719–748.
Kirzner, I. (1997). Entrepreneurial discovery and the competitive market process: An Austrian approach.
Journal of Economic Literature, 35, 60–85.
Lööf, H., & Heshmati, A. (2002). Knowledge capital and performance heterogeneity: A firm-level
innovation study. International Journal of Production Economics, 76, 61–85.
Lööf, H., & Heshmati, A. (2001). On the relationship between innovation and performance: A sensitivity
analysis, In ECIS—Stockholm school of economics, working paper series in economics and finance
No. 446.
Mairesse, J., & Mohnen, P. (2002). Accounting for innovation and measuring innovativeness: An
illustrative framework and an application. American Economic Review Papers and Proceedings, 92,
226–230.
Mairesse, J., & Mohnen, P. (2010). Using innovations surveys for econometric analysis. NBER Working
Paper 15857.
Malerba, F. (1992). Learning by firms and incremental technical change. Economic Journal, 102,
845–859.
Malerba, F. (2005). Sectoral systems: How and why innovation differs across sectors. In J. Fagerberg, D.
C. Mowery, & R. R. Nelson (Eds.), The oxford handbook of innovation (pp. 380–406). Oxford:
Oxford University Press.
Malerba, F., & Orsenigo, L. (1996). Schumpeterian patterns of innovation. Cambridge Journal of
Economics, 19, 47–65.
Michelacci, C. (2003). Low returns in R&D due to the lack of entrepreneurial skills. The Economic
Journal, 113, 207–225.
Nelson, R. R., & Winter, S. G. (1982). An evolutionary theory of economic change. Harvard: Harvard
University Press.
OECD. (1997). Oslo manual: The measurement of scientific and technological activities. Paris: OECD,
Proposed Guideline for Collecting and Interpreting Technological Innovation Data.
Ortega-Argilés, R., Potters, L., & Vivarelli, M. (2011). R&D and productivity: Testing sectoral
peculiarities using micro data. Empirical Economics, 41, 817–839.
Ortega-Argilés, R., Piva, M., & Vivarelli, M. (2014). The transatlantic productivity gap: Is R&D the main
culprit? Canadian Journal of Economics (forthcoming).
Palich, L. E., & Bagby, D. R. (1995). Using cognitive theory to explain entrepreneurial risk taking:
Challenging conventional wisdom. Journal of Business Venturing, 10, 425–438.
Parisi, M. L., Schiantarelli, F., & Sembenelli, A. (2006). Productivity, innovation and R&D: Micro
evidence for Italy. European Economic Review, 50, 2037–2061.
Parker, S. C. (2004). The Economics of Self-employment and Entrepreneurship. Cambridge: Cambridge
University Press.
Parker, S. (2008). The economics of formal business networks. Journal of Business Venturing, 23,
627–640.
Pavitt, K. (1984). Sectoral patterns of technical change: Towards a taxonomy and a theory. Research
Policy, 13, 343–373.
Pellegrino, G., Piva, M., & Vivarelli, M. (2012). Young firms and innovation: A microeconometric
analysis. Structural Change and Economic Dynamics, 23, 329–340.
Piga, C. A., & Vivarelli, M. (2003). Sample selection in estimating the determinants of cooperative R&D.
Applied Economics Letters, 10, 243–246.
Piga, C. A., & Vivarelli, M. (2004). Internal and external R&D: A sample selection approach. Oxford
Bulletin of Economics and Statistics, 66, 457–482.
Piva, M., Santarelli, E., & Vivarelli, M. (2005). The skill bias effect of technological and organisational
change: Evidence and policy implications. Research Policy, 34, 141–157.
Salter, W. E. G. (1960). Productivity and technical change. Cambridge: Cambridge University Press.
Santarelli, E., & Vivarelli, M. (2007). Entrepreneurship and the process of firms’ entry, survival and
growth. Industrial and Corporate Change, 16, 455–488.
Schneider, C., & Veugelers, R. (2010). On young highly innovative companies: Why they matter and how
(not) to policy support them. Industrial and Corporate Change, 19, 969–1007.
Schumpeter, J. A. (1934). The theory of economic development. Cambridge: Harvard University Press.

123
Econ Polit Ind (2015) 42:323–341 341

Schumpeter, J. A. (1942). Capitalism, socialism and democracy. New York: Harper and Brothers.
Solow, R. M. (1960). Investment and technical progress. In K. J. Arrow, S. Karlin, & P. Suppes (Eds.),
Mathematical methods in the social sciences (pp. 89–104). Stanford: Stanford University Press.
Storey, D. J., & Tether, B. S. (1998). New technology-based firms in the European Union: An
introduction. Research Policy, 26, 933–946.
Teece, D. J., Pisano, G., & Shuen, A. (1997). Dynamic capabilities and strategic management. Strategic
Management Journal, 18, 509–533.
Vivarelli, M. (2013). Is entrepreneurship necessarily good? Microeconomic evidence from developed and
developing countries. Industrial and Corporate Change, 22, 1453–1495.
Von Tunzelmann, N., & Wang, Q. (2003). An evolutionary view of dynamic capabilities. Economie
Appliquée, 16(3), 33–64.
Von Tunzelmann, N., & Wang, Q. (2007). Capabilities and production theory. Structural Change and
Economic Dynamics, 18, 192–211.
Winter, S. G. (1984). Schumpeterian competition in alternative technological regimes. Journal of
Economic Behavior and Organization, 5, 287–320.

123