A comparative analysis of the factors affecting the propensity to innovate in small businesses: The case of Sweden and Norway Humberto López Rizzo

This paper examines a Logit model of the effects of certain factors that hamper the propensity to innovate in the small businesses of Sweden and Norway. This is done by using a Survey (2007) of the Observatory of European SMEs. Our sample consists of 151 and 158 observations respectively. The value of the Odds-Ratio reports that, for a Swedish small business is 0.16 times less probable to innovate if has a problem in access the financing, for a Norwegian small business is 0.41 times less probable to innovate is has a lack of skilled personnel. Both countries will have less chances to innovate if they have the problem of expensive human resources. Finally, if the concurrency level of the market increases, this is a powerful incentive for the Norwegian small businesses to innovate. Their chances are 2.06 times better. This coefficient is not significant for the Swedish small businesses.

factors that inhibit the propensity to

Introduction

innovate in small businesses. This is a “… Our capacity to build on the growth

current issue worldwide, but in the present

and innovation potential of small and

paper, we study the case of Sweden and

medium-sized enterprises (SMEs) will

Norway. The criterion of selection of these

therefore be decisive for the future

countries is based on the following points;

prosperity of the EU”. This quote is one of

first, there is a small number of empirical

the opening lines of the European SBA

studies

(Small Business Act). The European

businesses

Commission signed, in 2008, the European

countries (Jacob et al., 2009); secondly,

SBA with the objective of establishing

even if both countries are similar enough to

programs to invest in the future of the EU,

be compared (Lööf et al. 2001), there is an

bidding particularly on the SMEs and

important element of disparity from the

relying on the innovation pillar to achieve

point of view of the innovation activities

this goal.

and their accomplishment (e.g. number of

on

the and

link

between

innovation

in

small these

patents applications, the R&D expenses at The aim of the present work is precisely to

a firm level, the expenses on R&D as a

extend and clarify the current debate on the

percentage of GDP, and so on). For

Humberto López Rizzo is a current PhD student from the University of Paris 1 – Panthéon Sorbonne. His topic of research is “The place of the SMEs on the growth, the employment and the innovation; an empirical examination”. Contact info: [email protected] ; [email protected]

1

instance, 25.90 percent of the Norwegian

are more Swedish firms that innovate (in

SMEs have in-house activities, and in the

absolute and relative terms), in comparison

case of Sweden, it is 41.80 percent (EC

with the Norwegian ones. This is 54.30

2009). Two questions emerge; why is there

percent versus 46.20 percent. Then, the

a gap in these indicators on innovation

factors

between them? What are the factors that

discussed; the lack of skilled resources is

could explain this discrepancy? To find an

the common obstacle to innovation for

answer to these questions, a Logit model is

both nations. Subsequently, is presented

used, along with an extract of the Survey

the methodology of the Logit regression

(2007) from the Observatory of European

and all the empirical proceeding to

SMEs.. The data set is composed of 151

perform. On this point, there is an

small businesses from Sweden and 158

explanation of how the variables and their

from

is

meaning are coded; in short, all the

considered, in this work, as defined by the

variables are dichotomous. Then, the

European Commission and the OECD in

results of the simulations are presented.

Norway.

A

small

business

that

handicap

innovation

are

2003, a firm with less than 50 employees. The first objective of this paper is to In this research, we initially discuss briefly

determine

some global aspects about innovation in

discouraging

order to get the global picture of this issue.

innovation of small businesses. For this, a

Some of the points considered are the main

Logit model is used a tool of analysis. The

definitions,

results show that the expensive skilled

the

different

types

of

the

impact

factors

have

some

on

labor has

hampering

main

innovation propensity of the Norwegian

references on this point are the ideas of

small businesses, but also the lack of

Schumpeter and some definitions of the

skilled labor does. On the other hand, the

Oslo Manual. After this point, there is a

problems

succinct survey of the empirical evidence

chances of the Swedish small businesses to

of the different elements that may have an

innovate.

Two

with

impact

the

innovations, as well as some of the factors innovation.

a deeper

that

financing

on the

reduce

the

impact on innovation (e.g. expenses in R&D,

human

resources

or

market

The second objective is to test the

structure). The next parts includes the

Schumpeterian

database,

some

power, that is to test a dummy that

descriptive statistics which stress that there

measures the perception of the concurrency

its

definition

and

2

hypothesis

of

market

level

of

significance

the

market.

There

the

Swedish

for

is

no

development, where new technologies

small

replace

old

existing

ones.

So,

the

businesses with this variable and the

introduction of radical innovation (a.k.a.

propensity to innovate. However, the

„novelty‟ or „new combinations‟) breaks

Norwegian

a

the stationary equilibrium. According to

positive incentive to innovate when the

Schumpeter, capitalism is precisely a

concurrency of the market increases.

method of economic change that can never

small

businesses

show

be stationary. These continuous disruption Limitations on this research

processes of “creation-destruction” are the

Neither a book nor articles are enough to

drivers of economic growth. According to

cover all the elements that interfere with

Diamond (2004) “Schumpeter‟s central

the innovation issue, because there is most

message is that the process of creative

of the time a large number of considerable

destruction

factors that may contribute to/obstruct the

competition in capitalism that is capable of

innovation capability of small firms. We

dramatic improvements in the quantity and

tried to be explicit in every point, trying to

quality of our lives” (Diamond 2004, p. 3)

describes

the

form

of

keep it short, clear and simple. On most points, sources for further details were

At a firm level, the Oslo Manual

cited as possible. Concerning the results,

underlines the fact that innovation can also

there is one point pertinent to mention;

improve

these results were obtained from a survey

firm‟s ability to innovate. For example,

that

small

improving the capabilities of production

businesses, so the conclusions are from an

processes can make it possible to develop a

average small business.

new

considered

all

kinds

of

performance by increasing the

range

of

products,

and

new

organizational practices can improve the firm‟s ability to gain and create new

Some aspects about innovation

knowledge that can be used to develop The aim of this point is to present the

other innovations (OECD 2005, p. 29).

global picture of some basic elements on

This could be perceived as a kind of

the innovation issue.

virtuous circle of innovation.

Joseph Schumpeter notably influenced the

The definition of innovation, according to

innovation

considered

the Oslo Manual, is “the implementation of

innovation as the driver of economic

a new or significantly improved product

theories.

He

3

(goods or services) or process, a new

There are a vast number of potential

marketing method, or a new organizational

factors that could contribute - at different

method in business practices, workplace

degrees - to the innovation propensity of

organization or external relations” (OECD

small firms. Box (2009) notes the fact that

2005, p. 46).

innovation is boosted by the interaction of both factors: the market and the non-

The Oslo Manual describes four types of

market ones. Among the factors that

innovations1:

hamper the innovation activities, there are

1) Product innovations,

cost factors (e.g. cost too high, lack of

2) Process innovations,

funds within the enterprise and so on),

3) Organizational innovations

knowledge factors (e.g. lack of qualified

4) Marketing innovations.

personnel, lack of information technology, organizational

rigidities

within

the

Quoting the Oslo Manual, we may find

enterprise), market factors (e.g. potential

that:

market

“Product

innovations

involve

dominated

by

established

significant changes in the capabilities of

enterprises) or institutional factors (e.g.

goods and services. Both entirely new

lack

goods

property rights) (for a detailed list of the

and

services,

and

significant

improvements to existing products are

of

infrastructure,

weakness

of

factors see the Oslo Manual, p. 113).

included” (OECD 2005, p. 17). This paper focuses on “product innovation”. This is

Empirical Literature Survey

because its definition fits well with the definition used in the survey of the

The

database to measure.

innovation has been largely explored since

empirical

literature

addressing

the first works of Schumpeter in the 1930s. However, the works studying the factors that influence the innovation capability of 1

a small firm date from a more recent

The ideas of these types of innovations were notably influenced by the works of Schumpeter. He conceives that there are five cases where a novelty or new combinations take place. Schumpeter‟s first novelty type considers “(I) The introduction of a new product - one that consumers are not yet familiar with - or of a new quality of goods” (see Schumpeter, 1967: 66). This is an absolute fact that the definition of product innovation proposed by the OECD matches fine with Schumpeter‟s first novelty type.

period. The aim of this point is to present briefly some of the empirical findings of factors like the expenses in R&D, the skilled human resources or the market structure.

4

Expenses in R&D

Lööf et al. (2001) studied the factors that

Acs and Audretsch (1988) found positive

hamper innovation across the Nordic firms

effects of the innovative output (e.g.

by using the CIS (Community Innovation

patents) with the expenses in R&D for the

Survey). They stressed that, overall, the

manufacturing in the United States, in turn

elements that inhibit innovation in Swedish

Albaladejo and Romijn (2000) does it for

firms are the lack of appropriate sources of

the British SME for their innovative

finance and the lack of qualified personnel.

capability. Bhattacharya and Bloch (2004)

For Norwegian firms, the factor that is

for Australian firms counted on their

more discouraging to innovate is the

successive innovations, and Lee (1995) for

organizational rigidities. One interpretation

Korean electronics firms for the setting of

of these results, according to the authors, is

a new technology.

that increasing the supply of skilled personnel would eventually stimulate the

Human Resources

innovation activity in Sweden, whereas in

Albaladejo and Romijn (2000) found that

Norway, the firms that innovate need to

the

continuously change their organizations.

education

background

of

the

founder/manager (mainly if holding a degree in science or engineering), his

The market structure

previous work experience (particularly if

Acs and Audretsch (1988) found a

this is in a multinational and large

negative

domestic company and the public science

concentration and innovation activity for

base), and the education of the firm

247 manufacturing industries in the United

workforce (especially university-trained

States. In the meanwhile, Van Dijk et al.

engineers and/or designers) are factors that

(1997) studied the manufacturing industry

have positive influence on the innovation

of the Netherlands, and found that the

capability of the British SME. Likewise,

more the intensity of the concurrence level,

Van Dijk et al. (1997) noted that the Dutch

the more the propensity to innovate in

small firms are more likely to innovate in

small businesses. The authors pointed out

(skilled) labor-intensive industries. These

that this is because small businesses want

findings stress the fact that the (skilled)

to offset the disadvantage of their size, and

human factor has an important role in the

survive in an industry dominated by large

propensity of a firm to innovate.

firms. Bhattacharya and Bloch (2004)

relationship

between

market

corroborated these findings, but for 1213 business 5

units,

covering

1728

manufacturing locations in the United

percent) corresponds to Norway and 151

States.

(48.9 percent) to Sweden.

The Database

It

may

seem

that

the

number

of

observations used in the present work is The database was obtained from a survey

not large enough to get robust results, but

that was lead by the Observatory of

this kind of methodology (Logit model)

European SMEs on November 2006-

allows the use of a sample of dissimilar

January 2007. The survey is statistical

sizes, being statistically adequate. To

stratified and takes into account economic

illustrate this, for example, Darroch and

activities at the 0-digit level of the

Clover (2005) calculated a Logit model,

European NACE-Nomenclature v1.1. The

using 44 observations. Farid, Silong and

final sample from our countries includes

Sarkar (2010) also calculated a Logit

observations from the sectors D, F, G, H, I,

model,

J, K, N and O, and is composed of a total

Nassimbeni

of 309 observations where 158 (51.1

Logit/Tobit, using 165 observations.

using

88 (2001)

observations, calculated

and a

Measuring Innovation Table 1 shows us the fact that, for our The innovation variable was obtained

survey, more Swedish small business (in

through question 51 of the survey. This

relative and absolute terms) innovate in

point emphasizes the fact that, if a firm

comparison with the Norwegian ones.

introduced a new or significantly improved product or a service in the last two years.

Measuring the hampering factors

Taking that if a firm responds positively to this question this is considered then as an

The Table 2 summarizes the number of

innovative firm. The following table

observations of the covariates (e.g. the

summarizes the distribution of the number

variables that will explain the propensity to

of firms that are considered as innovators.

innovate through the Logit model) that are

Table 1

present in the firms that innovate and those that do not. At first sight, we can realize that the variable that affects the most in the case of both groups of firms and countries

.

is the lack of skilled human resources. On 6

the other hand, it seems that the high

Most of the firms consider that the

interest rate does not pose any constraint to

concurrency level of their market has

the Nordic small businesses. Anyway, the

increased over the last two years. This is

problem appears to lie in having access to

for both countries. One aspect to note is the

the finance not the payment of the interest

fact that almost the same proportion of

rates. Another point to note is the fact that

Norwegian businesses (innovative and

no Norwegian firm stressed the lack of

non-innovative) has the perception that the

ability to use new technologies, unlike the

conditions of the market have become

Swedish ones. Finally, it may seem

more difficult. Additionally, we could

paradoxical, but some firms of both nations

point out the fact that, for almost five of

are confronted to the restraint of having a

every ten Swedish firms (innovative and

lack of market for innovation.

non-innovative),

the

concurrency

has

increased (this is the 55 percent). So, for Insert here Table 2

the Norwegian ones, it is six of every ten firms (this is the 64.5 percent).

These variables were drawn from question 52 of the poll. It is important to indicate

Methodology

that this question is a multiple-choice one. This means that a firm could have

On the present paper the Logit Model is

experienced no constraint up to seven ones.

used as the econometric method in order to try to explain how different hampering

Measuring the market structure

factors

could

potentially

affect

the

propensity to innovate of the Nordic small The Table 3 summarizes the perception of

businesses. This method was selected for

the concurrency level across the innovative

the following reasons; first, this technique

and non-innovative small businesses.

seems more appropriate to treat the Survey

Table 3

(2007) from the European Observatory of SMEs (e.g. statistical survey at the individual), secondly, the distribution of our sample is positively skewed and the logit transformation should be favored because the data is not normally distributed (Finney 1952), thirdly, there is a scarcity of

7

empirical works using this technique2 on

Then the ratio Pi/(1- Pi) give us as a result

the small business and innovation.

the Odds-Ratio (this is the O.R. or the SPSS present it like Exp(B)) in favor of a

Our starting point is with the following

small firm that innovates. This is the ratio

equation:

of probability that a small firm innovates to the probability that it does not innovate.

(1)

Thus, if we have a Pi= 0.6, this means that the O.R. is 1.5 times favorable to the small We may note that Pi in this case represents

firm that innovate.

the probability of occurrence of an event, let‟s say the probability that a small firm

Transforming Pi, and then taking the

innovates. Inversely (1- Pi) represents the

natural log of the equation (1), we obtain:

probability that a small firm does not innovate.

But this probability Pi is known as the (cumulative) logistic distribution function.

The Li represents the log of the O.R., and

The value of  ranges from -  to +  and

this is called the Logit. In short, the O.R.

ranges between 0 and 1. The graphic 1

after a transformation becomes the Logit.

illustrate this idea.

The 1 represents the constant, and 2 is the covariate that is the variable that

Graphic 1

influences the probability of the event of occurs or not.

If a small business innovates

If a small business does not innovate

The following equation corresponds to the model run here;

2

Hosmer and Lemeshow (2002) note that this technique is applied generally to the field of medicine (such as the epidemiology). Nevertheless Gujarati and Porter (2009) indicate that this technique has been largely used in analyzing various economic cases (such as the money supply).

( )

8

(

)

where, Our procedure is as follows, first we run 1: Lack of ability to use new technologies

the model as mentioned above, and then

2: Too expensive human resources

we present the results and the discussion.

3: Lack of skilled human resources

Secondly,

4: High interest rates

variable „concurrency level‟ this variable

5: Problems with access to finance, other

measures

we

the

introduce

level

of

the

dummy

concurrency

perceived by the firms (as decreasing = 1,

than interest rates 6: Hard to protect intellectual property

or stable = 2 or increasing = 3). We do this

7: Lack of market demand for innovation

because we want to test the Schumpeterian hypothesis of market power (secondary objective), this will show us if the

The database handles the observation of

conditions

the small firms at the individual, then for

of

the

market

have

any

influence on the Nordic small firms to

each observation we do the following.

innovate. The innovation is measured through a

The results

binary variable (1,0) where it takes the value of 1 where the small business

The software SPSS was used to calculate

introduces an innovation (event occurs)

the rate of innovation of the Nordic small

and 0 when it does not happen (absence of

businesses. Tables 4a and 4b illustrate the

event). Innovation is reefed as the new or

Chi-Squared Goodness of Fit that tests the

improved product or service that was

null hypothesis, and then determines if the

introduced by the small firm in the last two

step is justified. Here, the step shows only

years. This definition corresponds to the

the constant variable, and because the

concept of innovation of the Oslo Manual

significance level is under 0.05, this

(2005). The same principle is applied for

indicates that other (s) variable (s) are

the covariates. For example, a firm that

included in the equation and other (s) are

face the problem of the high interest rates

dropped. Therefore, given the Chi-Squared

and this inhibit it to innovate, then the

test and the significance level (p-value) for

variable HIR for this firm will take the

both groups of countries, we can note that

value of 1 (event occurs) and the value of 0

the outcome model fits, overall, well.

(absence of event) so on for the rest of the variables.

9

Table 4a

Table 5a

Table 5b Table 4b

The subsequent tables (6a and 6b) illustrate The tables 5a and 5b shows the Cox &

the results that are the most pertinent and

Snell R2 test and the Nagelkerke R2. These

applicable to this work.

tests deliver a coefficient of determination R2, explaining the degree of variation

Insert here Table 6a

illustrated by the covariates. In this instance, for the case of the Swedish firms,

Insert here Table 6b

the inclusion of the covariate of too expensive human resources (2) and

These show us the factors than may affect

problems with access to finance (5)

the small businesses in their propensity to

explains between 10.2 and 13.6 percent of

innovate. The factor that handicaps both

the variation in innovation propensity. In

countries is the lack of skilled human

the meanwhile, for the Norwegian firms,

resources (3), but this one has a higher

the inclusion of the covariate of too

impact on the Swedish firms because they

expensive human resources (2) and the

exhibit

lack of skilled human resources (3)

considerable value than the Norwegian

explains between 8.7 and 11.6 percent of

ones. Let‟s say that, for a Swedish firm, it

the variation in innovation. It seems

is 0.21 times less probable to innovate, if it

pertinent to stress the fact that the

has this problem of expensive human

covariates included in the equation of the

resources. In other words, we can say that,

Swedish small businesses have a lighter

for a Swedish firm that has no problem

explaining power, in comparison with the

with expensive human resources, it is 4.76

Norwegian ones.

times more probable to innovate. (This

an

Odd-Ratio

of

a

more

number is obtained by calculating the inverse of the Exp (B) that is 1/0.21=4.76). 10

The second factor that handicaps the

precisely the case of the Norwegian small

innovation capacity of the Swedish firms is

firms; the scarcity of this element has a

the problems with access to finance (5). A

negative impact on their chances to

firm that has this problem is like 0.16 times

innovate. In other words, the absence of

less probable to innovate. In other words, a

skilled human resources handicaps their

firm that does not have this financing

potential to innovate.

problem has 6.25 times better chances to innovate . The second factor that handicaps

We turn now to the inclusion (Table 7) of

the Norwegian firms is the lack of skilled

the variable concurrency level in order to

human resources (3). Then, for a small

test the Schumpeterian hypothesis of the

business that has this problem, it is 0.41

market power - if the concurrency level in

times less probable to innovate, and a firm

a market increases, what does happen with

that does not have this problem of the lack

innovation? Do the economic actors turn to

of human resources has 2.43 times better

innovation in order to stay in the market or

chances to innovate.

maintain a status vis-à-vis the economic profits? The results show a relation that is

These latter results corroborate partially,

not significant for the case of the Swedish

for the case of Sweden, what was noted by

small businesses, but this is not the case

Lööf et al. (2001), that the lack of

with the Norwegian ones. This coefficient

appropriate sources of finance is one

not only was significant (when the

element that inhibits innovation in the

concurrency level increased), but also the

Swedish firms. In addition, this result

Odds-Ratio of this variable showed a

could be linked to what was said in the

notable increase in the propensity to

review of the empirical literature, that is

innovate. The chances to innovate for a

investing in R&D brings out positive

Norwegian small business are in the order

effects on innovation. In this case, if the

of 2.06 times. In other words, when the

Swedish small businesses do not have

concurrency level increases in Norway, the

enough access to funds, their innovative

propensity to innovate is 2.06 times better

propensity will be inhibited. Insert here Table 7 Concerning the case of Norway, different empirical works note the fact that qualified labor has a positive impact on the disposition of a firm to innovate. This is 11

point out that the covariant of the

Conclusions

Swedish equation has a slightly better The main findings of this research paper

power

of

prediction

than

the

are recapitulated as a general conclusion.

Norwegian's. They exhibit a value of 10.2 and 13.6 percent, and for Norway

1) The descriptive statistics showed that

a value of 8.7 and 11.6 percent.

there are more innovative Swedish

4) The value of the Odds-Ratio calculated

small businesses (in absolute and in

shows that the chances to innovate in

real terms) than Norwegian ones; this

the firms of both countries decrease

is 54.30 percent against 46.30 percent.

when the firms have the problem of

This tendency is corroborated with the

expensive skilled human resources, but

stylized

fact

beginning; Norwegian

presented

25.90 SMEs

percent have

in

the

this has a deeper impact on Norwegian

of

the

small businesses.

in-house

5) For a Swedish small business, it is 0.16

activities, and in the case of Sweden, it

times less probable to innovate if it has

is 41.80 percent. So, there are more

a problem with access to financing.

Swedish small businesses innovating

Otherwise stated, a firm without this

because there are more Swedish SMEs

problem of financing has 6.25 times

involved in in-house activities.

better chances to innovate. These

2) The descriptive statistics of the factors

results

for

the

Swedish

small

hampering innovation reveal that, for

businesses corroborate the findings of

both groups of firms, the human factor

Lööf et al. (2001), who point the fact

handicaps their propensity to innovate.

that for the Swedish firms the problems

The factors of the lack of skilled

of access to financing are traduced in a

human resources and the expensive

diminution of the chances to innovate.

human resources, however, more

important

impact

have a on

6) For a Norwegian small business, it is

the

0.41 times less probable to innovate if

innovative small businesses. Finally, it

it has the problem of the lack of skilled

seems that the Swedish small business

human resources. In turn, a firm not

have more difficulties in getting access

having this problem increases its

to financing than the Norwegian small

chances to innovate by 2.43 times. In

businesses.

the review of the empirical literature,

3) The coefficients of Cox & Snell R2 and

the positive effects of the skilled labor

2

Nagelkerke R (on tables X and Y)

on the innovation were accentuated, so 12

this

is

what

needed for the

Albaladejo, Manuel and H. Romijn (2000).

Norwegian small businesses to increase

Determinants of Innovation Capability in

the probabilities to innovate.

Small UK Firms: An Empirical Analysis,

7) Concerning

is

the

Schumpeterian

Eindhoven Centre for Innovation Studies,

hypothesis of market power innovation,

The Netherlands, Working Paper 00.13,

after testing the dummy variable of the

Arvanitis, Spyros (1997). “The Impact of

concurrency

no

Firm Size on Innovative Activity – an

significance for the Swedish small

Empirical Analysis Based on Swiss Firm

businesses.

Data”, Small Business Economics, 9, 473-

level,

we

found

Consequently,

the

increasing concurrency does not give

490.

any incentive

Bhattacharya, Mita, and H. Bloch (2004).

to

these

firms

to

introduce innovation. Contrariwise, the

“Determinants

increase in the concurrency level for

Business Economics, 22, 155-162.

the Norwegian small businesses works

Box, Sarah (2009), OECD Work on

as a considerable incentive to innovate.

Innovation- A Stocktaking of Existing

The

variable

Work, OECD Science, Technology and

showed a value of 2.06, which means

Industry Working Papers, 2009/2, OECD

that,

publishing.

Odds-Ratio

given

an

of

this

increase

in

the

of

Arthur,

Innovation”,

M.

Small

concurrency level, the firms have 2.06

Diamond

(2004).

times better chances to innovate.

“Schumpeter‟s Creative Destruction: A Review of the Evidence”, paper resubmitted to Journal of Private Enterprise), Departament of Economics, University of Nebraska at Omaha.

References

Darroch M.A.G. and T.A. Clover (2005). Acs, Zoltan J. and D.B. Audretsch (1987).

“The Effects of Entrepreneurial Quality on

“Innovation, Market Structure, and Firm

the Success of Small, Medium and Micro

Size”, The Review of Economics and

Agri-Businesses in Kwazulu-Natal, South

Statistics, 69(4), 567-574.

Africa”, Agrekon, 44(3), 321-343.

Acs Zoltan J. and D.B. Audretsch (1988).

Van Dijk, Bob, R. Den Hertog, B.

“Innovation in Large and Small Firms: An

Menkveld, and R. Thurik (1997), “Some

Empirical

New Evidence of the Determinants of

Analysis”,

The

American

Large-and Small-Firm Innovation”, Small

Economic Review, 78(4), 678-690.

Business Economics, 9, 335-343. 13

European Commission (2009). SBA Fact

a Comparison of the Nordic Countries,

Sheet 2009, European Commission -

SSE/EFI

Enterprise and Industry.

Economics and Finance 457.

Jacob, Merle., Johansson M., Hellström T.,

Nassimbeni, Guido (2001). “Technology,

Iversen E., Kutinlahti P., Bikeholm Munk

Innovation Capacity, and the Export

K., Knudusen L.G. and Wennerberg S.B.

Attitude of Small Manufacturing Firms: a

(2009),

Logit/Tobit Model”,

“SMEs and the new role of

academics

research

in

four

Nordic

Working

Paper

Series

in

Research Policy,

30(2001), 245-262.

countries”, Nordisk InnovationsCenter,

OECD (2005), Oslo Manual: Guidelines

Farid, Hadi, A.D. Silong, and S.K. Sarkar

for Collecting and Interpreting Innovation

(2010). “Application of Logit Model in

Data, Third Edition, OECD Publishing, A

Innovation

joint publication of OECD and Eurostat,

Adoption:

a

Study

on

Biotechnology Academic Researchers in

OECD (2010). Measuring Innovation

Malaysia”, American-Eurasian Journal of

Symeonidis, George (1996). Innovation,

Agricultural & Environmental Sciences,

Firm

9(3), 282-287.

Schumpeterian Hypotheses and Some New

Finney, D.J., Ed. (1952). Probit Analysis,

Themes, OECD Economic Studies 27(II),

Cambridge University Press, Cambridge,

35-70.

England.

Schumpeter, Joseph A. (1967). The Theory

Gujarati, Damodar N. and D.C. Porter

of the Economic Development: An Inquiry

(2009), Basic Econometrics, McGraw-Hill

into Profits, Capital, Credit, Interest, and

Irwin, United States of America, Fifth

the Business Cycle, New York, Oxford

Edition.

University Press, Fifth printing.

Hosmer, David W. and S. Lemeshow

Tether, B.S. (1998). “Small and Large

(2000). Applied Logistic Regression, A

Firms: Sources of Unequal Innovations?”,

Wiley-Interscience

Research Policy, 27(1988), 725-745.

Publication,

John

Wiley & Sons, Inc., United States of America. Lee, Jangwoo (1995).

“Small

firms‟

innovation in two technological settings”, Research Policy, 24(1995), 391-401. Lööf, Hans, A. Heshmati, R. Asplund, and S-O

Nåås

(2001).

Innovation

and

Performance in Manufacturing Industries: 14

Size

and

Market

Structure:

Table 2

15

Table 6a

Table 6b

16

Table 7

17