The Role of Small Firms in Regional Growth: Evidence from
Regional Data in Iran
Iman Cheratian1
Saeed Ghorbani2
ABSTRACT
Purpose: The main purpose of this paper is to investigate the effect of small firms in Iranian
regional growth at the level of Iranian regions.
Design/Methodology/Approach: This paper by using of endogenous growth theory tests the
hypothesis where or not small firms are important vehicles for regional growth. To get the
results, we divided firms according to their legal status into six types: private, governmental,
cooperative, official, public and finally total number of small firms and analyzed the effect of
each kind of mentioned types on regional growth indicators during the 2001-2013. The
econometric specification is Panel Fixed and Random Effects models.
Findings: The results show that private, governmental and cooperative small firms have
positive and significant impacts on regional growth indicators, while the effect of public small
firms are negative and significant. Overall, the empirical results indicate that the importance of
small firms is positively correlated with regional economic growth and this result is consistent
with previous studies. In addition, our results show that the human capital embodied in small
firms are important for regional economic growth.
KEYWORDS: Entrepreneurship capital, Regional growth, Small firms, Iran.
JEL: D21, R58, O18, O25.
1
International economics PhD student, Economics Department, University of Mazandaran, Babolsar,
Iran.
Center of SMEs studies, Academic Center for Education, Culture and Research (ACECR), Tehran, Iran.
Email: [email protected], Corresponding author, Tel.: +98 9192348218 Fax: +98 88220298.
Center of SMEs studies, Academic Center for Education, Culture and Research (ACECR), Tehran, Iran.
Email: [email protected]
2
1
1. INTRODUCTION
Economic Growth has been a main pensiveness of economist during the last decades.
Ever since Robert Solow (1956) based his model of economic growth on the
neoclassical production function with its key factors of production, physical capital and
unskilled labor, economists have relied upon the model of production function as a basis
for explaining the determinants of economic growth. Solow's approach generally
consisted of relating measures representing these two fundamental factors of production
in trying to explain variations in growth rates typically over time in a single country or
across countries in a cross-sectional context (Audretsch and et al., 2006). Romer's
(1986) critique of the Solow approach was not with the basic model of the neoclassical
production function, but rather with what Romer perceived to be omitted from that
model was knowledge(Audretsch and et al., 2006). He argued it can be possible to
reinterpret (K) as a combination of physical capital and investment in R&D. So (K) in
addition of buildings and machines would include innovation, blueprints, creativity and
ideas for how to produce new goods. Following with Romer, some economists like
Lucas (1988) and others account knowledge, entrepreneurship and innovation as
important factors of production. They believe because it was endogenously determined
as a result of externalities and spillover, so it was particularly important.1
Entrepreneurship encourages innovative activity and puts a region at the forefront of
economic growth.2 Generally, regions that use entrepreneurial capital in a correct way,
may be a winner in a competitive economic growth. So the benefits of entrepreneurship
for regional welfare have, promoted much policy interest in how to favor
entrepreneurship in the regional economy. Entrepreneurship has a central important role
among the process that cause to regional economic change. Entrepreneurs can make
changes and speed up the creation, diffusion and application of new ideas.
This paper considered to entrepreneurship capital as an important factor in regional
economic growth. In other word, the main purpose of our analysis is to provide
empirical evidence on how entrepreneurship capital affect regional economic growth in
Iran by using a Cobb-Douglas function to carry out the analysis. The remainder of the
paper is organized as follows: The second section discusses about existing literature on
the link between firms behavior and regional growth and the mechanisms through
which firms might impact economic performance. Empirical results are presented in the
section 5, based on data and methodology described in the third and fourth sections.
Finally, section 6 provides conclusions for policy as well as for further research.
2. LITERATURE REVIREW
1
For example, Acs et al (2009) suggest that entrepreneurship contributes to economic growth by acting as
a conduit through which knowledge created within incumbent firms’ spillover to agents who set up new
firms.
2
In support of this view, Baumol (1968) considers entrepreneurs the key factor to stimulate economic
growth and suggests that innovations require entrepreneurial initiative to be introduced. (Cravo, Becker
and Gourlay, 2014)
2
The positive relationship between entrepreneurship capital and regional growth argued
in some notable studies. In the most of these studies, the unit of observation is at the
spatial level, city, region and state. For example, Hart and Hanvey (1995) measure the
link of new and small firms to employment generation in the late 1980 for three regions
in the United Kingdom. The results of their study show that although employment
creation came largely from SMEs, but most of the job losses also came from SMEs.
Audretsch and Fritsch (1996) use a database to identifying new business startups and
exit from the social insurance statistics collected for individuals in Germany. They find
that in both the manufacturing and the services sectors, a high rate of turbulence in a
region tends to a lower and not a higher rate of growth. They attribute this negative
relationship to the fact that the startup and death rates are both negatively related to
subsequent economic growth. The areas with higher startups rates tend to experience
lower growth rates in subsequent years. This fact is also true for the death rates. They
conclude that the German regions which experiencing higher death rates also tend to
experience lower growth rates in subsequent years. Callejon and Segarra (1999) use a
data set of Spanish manufacturing industries from 1980 to 1992 to link new firm birth
and death rate to total factor productivity growth in industries and regions. They find
that both new firm startup rate and exit rate contribute positively to the growth of total
factor productivity in regions as well as industries.
Foelster (2000) by using a Layard-Nickell framework provides a link between micro
behavior and macroeconomic performance in Sweden between 1976 and 1995. He
shows that increase in self-employment share have had a positive impact on regional
employment rates in this country. Audretsch and Fritsch (2002) find that different
results emerge for Germany in the 1990s. Regions with higher startup rates exhibit
higher growth rates. This would suggest that Germany is changing over time with the
engine of growth shifting toward entrepreneurship. Holtz-Eakin and Kao (2003) in their
study examine the impact of entrepreneurship on states growth. They use of
productivity change over time as measurement of growth. The results of their study
show that variations in the birth rate and death rate for firms are related to positive
changes in productivity. So in the case of United States, they conclude the
entrepreneurship has a positive impact on productivity growth. Audretsch and Keilbach
(2004a) by introduce a new factor, entrepreneurship capital, and link it to German
regions output in the context of a production function model, explain what is meant by
entrepreneurship capital and why it should influence economic output. The results
indicate that entrepreneurship capital is a significant and important factor shaping
output and productivity. In another study, they suggest that knowledge is not sufficient
to generate the diversity that is the driving force of economic growth. Rather, additional
mechanisms are required to transform knowledge into diversity. One such mechanism is
entrepreneurship. Entrepreneurship is an important mechanism in driving selection
process hence in creating diversity of knowledge, which in turn serves as a mechanism
facilitating the spillover of knowledge. They provide empirical evidence that regions
with higher levels of entrepreneurship indeed exhibit stronger growth in labor
productivity (Audretsch and Keilbach, 2004b). Audretsch and Keilbach (2005)
3
introduce the concept of entrepreneurship capital and link it to the economic
performance of regions. They give a definition of entrepreneurship capital and suggest
different measures of this variable. Economic performance of regions is measured by
stock and the growth rate of regions, labor productivity. They find that entrepreneurship
capital is stronger in urban areas and spatially correlated. Using regressions of
production functions and growth equations, they find evidence that entrepreneurship
capital has a positive and large impact on region's labor productivity. Mueller (2007) in
his study analyzes the relationship between the exploitation of entrepreneurial
opportunities, namely start-up activity, and regional economic growth. In particular, this
paper explores if those regions that increased their new firm formation activity also
experienced an increase in GDP. He find that an increase in innovative start-up activity
is more effective than an increase in general entrepreneurship for economic growth.
Chun-Chien and Chai-Hai (2008) assess how and to what extent knowledge capital and
technology spillover contribute to regional economic growth in China. Moreover, the
absorptive ability played by human capital on acquiring advanced foreign technologies
is also investigated in this study. Empirical results show that knowledge capital both of
R&D capital and technology imports contribute significantly, with similar impact to
regional economic growth. The analysis also suggests the existence of R&D spillovers
as well as international knowledge spillovers. Moreover, a region’s absorptive ability is
considered as the critical capability to absorb external knowledge sources embodied in
FDI and imports, which then contribute to the regional economic growth. Although the
effect of entrepreneurship and knowledge capital on regional economic growth have
been widely discussed in previous studies and different countries, but the role of
mentioned variables in regional economic growth has not been investigated in Iran. So
this paper assesses how and to what extend entrepreneurship capital contribute to
regional economic growth in Iran. Some more studies on the relationship between firms
and regional growth have been listed in Table 1 by authors, country focus, sample size,
variables and empirical results.
4
5
Netherlands
Italy
Van Stel and Suddle
(2008)
Piergiovanni, Carree
and Santarelli (2009)
Portugal
Baptista, Escaria and
Madruga (2008)
Germany
West
Germany
Audretsch, Bönte and
Keilbach (2008)
Fritsch and Mueller
(2004)
Country
focus
Reference
103 Italian
provinces
Netherland 40
regions
326 West
Germany kreise
Portugal 30
regions
310 WestGerman counties
Sample size
Maximum effect of new businesses on
regional development is reached after 6 years.
Employment impact of new firms is strongest
in manufacturing industries and areas with
higher degree of urbanization
Finding show a positive effect of the increase
in the number of forms active in the creative
industries, net entry and a greater provision of
leisure amenities on regional economic growth
Employment growth, firms startup, population density, wage
growth, degree of urbanization,
degree of rurality
Growth of value added, growth of
employment, creativity,
trademarks, patent, leisure, culture,
human capital, migration, entry rate
of firms
Indirect effects of new firm birth on
subsequent employment growth are stronger
than direct effects
Relative employment change in the
private sector, yearly start-up rate
of businesses
New firms can have both a positive and a
negative effect on regional employment. The
peak of the positive impact of new businesses
on regional development is reached about 8
years after entry
Innovation efforts have both a direct and
indirect effect on regional economic.
Performance. Also innovation efforts of
incumbent firms lead to an increase in regional
technical knowledge.
Entrepreneurship capital, Economic
performance, Technical knowledge,
Innovation efforts
Relative employment change in the
private sector, yearly start-up rate
of businesses
Empirical finding
Variables
Table (1). Pervious literature on firms and regional growth
3.
DATA
Our empirical analysis is based on Iranian regional data. We compiled a database
consisting of 30 provinces in the periods of 2001-2013. The main source of data is
Iranian Statistical Center (ISC) that is the most important data center for information
about manufacturing firms in Iran. The description of the data is as follows: In the first
equation we consider regional GDP per capita as dependent variable and estimate the
effects of independent variables on it. In the second model, regional GDP (without oil)
is considered as dependent variable and the effects of small firms on this variable is
investigated in equation (2). (IVAi) is measured as industries value added in Iranian
regions (in million Rials) and considered as dependent variable in equation (3). (Ki)
used as physical capital and measured as the value of firms’ investment in buying
machines, vehicles, durable tools, administrative equipments, buildings and
installations, land and soft wares. The value of this data is based on million rial. (Li) is
expressed as total number of employees who engaged in Small firms. (Ri) is knowledge
capital and expressed as number of employees with high educational degree (M.S and
Ph.D.) who work in small firms. On the part of entrepreneurship capital, it is the most
interesting explanatory variable for the aim of this paper. (Pri ent) expressed as private
firm and measured as total number of firms that all or more than 50% of their
investment belongs to the private sector (individuals or cooperatives). (Gov ent) is
governmental firm and measured as total number of firms that registered by law and all
or more than 50% of their capital owned by the government. The next explanatory
variable is cooperative firm (Coop ent) that measured as the total number of firms that
at least 51% of their capital owned by the cooperative members. Official firm (Offic ent)
is another independent variable that measured as the number of firms except those
which are cooperative and governmental. (Pub ent) is expressed as public firm and
measured as the number of firms that all or more than 50% of their capital owned by
ministries government agencies, banks, municipalities and other public sector
institutions. Finally (Total ent) is the total number of small firms that are working in
each province.1
Table (2). Summary statistics of variables used in regression
Mean
Median
Std. dev.
Minimum
Maximum
GDP Per capita
37458.87
26073
36785.42
3474.15
2503999
Gross Domestic Product (Y)
81403613
35226321
1.7e+08
2290842
1.77e+09
Industrial value added (IVA)
13125050
4520176
25095842
121684
1.65e+08
Capital (K)
131429.1
71284.5
174730
1328
1227879
Number of employees (L)
8227.9
4344.5
11409.43
455
90997
Number of R&D employees (R&D)
67.199
30.5
97.136
0
707
firms with private status (Pri ent)
154.35
69
301.15
0
3015
firms with governmental status (Gov
2.95
2
3.468
0
19
1
Geographical distribution of small firms (based on their legal status) described in Appendix 3.
6
Table (2). Summary statistics of variables used in regression
Mean
Median
Std. dev.
Minimum
Maximum
15.07
12
11.474
0
56
firms with official status (Offic ent)
198.63
116.5
234.58
3
1488
firms with public status (Pub ent)
2.77
2
3.03
0
17
Total small firms (Total ent)
373.88
215.5
522.8
16
4351
ent)
firms with cooperative status (Coop
ent)
4. METHODOLOGY
The main hypothesis of this paper suggests that there is a positive relationship between
entrepreneurship capital and regional economic growth. More specifically, this paper
takes a look at the impact of entrepreneurial firms’ performance on regional GDP
(without oil), regional GDP per capita and regional value added in industrial sectors. A
common approach to measure the impact of entrepreneurship capital on regional
economic performance is on the basis of a production function of the Cobb-Douglas
type (Audretsch and Keilbach, 2005). In this function, (Yi) refers to regional economic
performance that we choose regional GDP (without oil), regional GDP per capita and
regional value added in industrial sectors as indicators of this variable. (Ki) refers to
small firms' physical capital, (Li) represents labor force who works in small firms, (Ri)
represents knowledge capital and (Ei) represents entrepreneurship capital that measured
as the number of small firms according to their legal status. The subscript i refer to
regions. As discussed earlier. In neoclassical economics, regional growth is shaped by
the traditional factors of labor and capital. The contribution of the new endogenous
growth theory was to suggest an additional source of growth, Knowledge. However in
different studies argued that the presence of knowledge is not sufficient to generate
growth. Rather, what matters is not knowledge, per se, but rather the propensity for that
knowledge to create diversity, which, in turn, result in growth. Entrepreneurship is one
mechanism, although certainly not the only one (Audretsch and Keilbach, 2004b).
Entrepreneurship capital can be expected to exert a positive impact on economic output
for a number of reasons. The first2 is that it is a mechanism for knowledge spillovers.3 A
second way that entrepreneurship capital exerts a positive influence on economic output
is through increased competition by an increased number of enterprises.4 A third way
that entrepreneurship capital generates economic output is by providing diversity among
2
See Grossman and Helpman (1991) and Cohen and Levinthal (1989) for a further discussion of this
approach.
3
Although the literature identifying mechanisms of transmitting knowledge spillovers is few and
underdeveloped, however one important area where such transmission mechanisms have been identified
involves entrepreneurship. Entrepreneurship involves the start-up and growth of new enterprises
(Audretsch and Keilbach, 2004a).
4
See Jacobs (1969); Porter (1990); Feldman and Audretsch (1999) and Gleaser et al. (1992) for a further
discussion of this approach.
7
the firms.5 Not only does entrepreneurship capital generate a greater number of
enterprises, but also it increases the variety of enterprises in the location (Audretsch and
Keilbach, 2004a).
Yi=αKiβ1.Liβ2.Riβ3.Eiβ4.eεi
(1)
We take equation (1), make a logarithmic transformation and then we have the
following equation:
Ln (Y)it= Ln (a) + β1Ln (K)it + β2Ln (L)it + β3Ln (R)it + β4Ln (E)it + εit
(2)
According to above function, a positive coefficient on physical capital (β1) and labor
force (β2) is consistent with the neoclassical growth theory as asserted by Solow (1956).
Also a positive coefficient on knowledge capital (β3) and entrepreneurship capital (β4) is
consistent with the endogenous growth theory, as posited by Romer (1986) and Lucas
(1988). A positive β3 and β4 would support the hypothesis that entrepreneurial capital
has a positive impact on regional economic performance that expressed as regional
GDP, regional GDP per capita and industries value added in this paper. Table 2 lists
summary statistics for regional indicators and small firms' variables that distributed in
Iranian provinces. As can be seen, there is a wide difference in the number of firms'
establishment in Iranian provinces that the mean number of official firms are higher
than other kinds of small firms. In addition, we can also observe that, on average, there
are about 8227 labor forces who work in small firms during the sample period. Table 3
provides the correlations among the variables considered in the regressions and the
results show high cross-correlations among dependent and independent variables. In
appendix 1 definition of variables and source of data is presented.
5
See Hannan and Freeman (1989) for a further discussion of this approach. 8
9
0.18 ***
0.66 ***
0.71 ***
0.61 ***
0.13 *
0.19 ***
0.41 ***
0.10
-0.15 ***
0.24 ***
0.15 ***
-0.09
0.18 ***
Ln (L)
Ln (R&D)
Ln (Pri ent)
Ln (Gov ent)
Ln (Offic ent)
Ln (Coop ent)
Ln (Pub ent)
Ln (Total ent)
0.70 ***
0.17 ***
0.24 ***
0.72 ***
0.76 ***
0.73 ***
0.76 ***
0.17 ***
0.47 ***
0.79 ***
0.03
0.68 ***
0.82 ***
0.77 ***
1.00
Ln (K)
0.99 ***
0.29 ***
0.45 ***
0.96 ***
0.29 ***
0.93 ***
0.88 ***
1.00
Ln (L)
Note 1: N= 391. Note 2: Significance levels: ***p< 0.01, **p<0.05, *p<0.1
0.64 ***
0.20 ***
0.25 ***
0.63 ***
0.64 ***
0.74 ***
0.51 ***
Ln (K)
0.77 ***
0.92 ***
0.64 ***
Ln (IVA)
1.00
1.00
0.68 ***
Ln (GDP)
Ln (IVA)
1.00
Ln (GDP)
Ln (GDP per
cap)
Ln (GDP
per cap)
0.87 ***
0.20 ***
0.42 ***
0.91 ***
0.09
0.76 ***
1.00
Ln (R&D)
0.94 ***
0.24 ***
0.35 ***
0.83 ***
0.29 ***
1.00
Ln (Pri
ent)
Table (3). Correlation matrix
0.30 ***
0.34 ***
0.16 ***
0.24 ***
1.00
Ln (Gov
ent)
0.96 ***
0.29 ***
0.49 ***
1.00
Ln (Offic
ent)
0.47 ***
0.33 ***
1.00
Ln (Coop
ent)
0.29 ***
1.00
Ln (Pub
ent)
1.00
Ln (Total
ent)
5.
EMPIRICAL RESULTS
Empirical results in this paper start by reporting the descriptive statistics in Table 2 and
correlation matrix in table 3. Appendix 2 shows the results of panel unit root tests to
examine the stationary of model variables. Panel unit root tests are a generalization of
the augmented Dickey-Fuller (ADF) individual country (state) unit root tests to a
common panel unit root test. The basic panel unit root test regression can be written as
follows:
yit= ρiyi,t-1 + Xitδi + εit
(3)
where i= 1,2,…,N cross-section units or series that are observed over periods t=
1,2,…,Ti. The Xit is the exogenous variables in the model, including any fixed effects or
individual trend, ρi are the autoregressive coefficients, and the error εit are assumed to be
mutually independent of individual disturbance. In this section we apply the Levin, Lin
& Chu, PP-Fisher Chi2, Im-Pesaran-Shin, Breitung t-test and ADF-Fisher Chi2 tests to
examine the stationary situation of model variables. The null hypothesis of this test is
that each series in the panel contains a unit root test, while the alternative hypothesis
allow for some of the individual time series to have unit roots. The results of panel unit
root test in appendix 2 provide evidence to reject the null hypothesis of unit root for
entire variables in 1% and 5% level of significant. Tables 4 and 5 show the effects of
small firms in regional GDP per capita. In Tables 6 and 7 we estimate the effects of
small firms in GDP (without oil) and finally in Tables 8 and 9 the effects of small firms
in regional industries value added are analyzed. In estimation of the production function
model, the first equation estimates the traditional Solow model of the production
function. In the second model, the factor knowledge capital is added. The rest models
show results when entrepreneurship capital is included in the production function
model.
5.1.Impact of Small Firms on Regional GDP Per Capita
Estimation of the production function model of equation (2) produces the results
displayed in tables 4 to 9. In the regression results, the first equation estimates the
traditional Solow model of production function. As the positive and statistically
significant coefficient suggest that both physical capital and labor force are important
factors of production in economic growth of Iranian regions. In the model 2, the factor
knowledge capital is added. The positive and statistically significant coefficient of this
variable supports the Romer's argument that knowledge-intensive inputs matter as a
factor of production. In the 3 to 8 columns, the number of small firms based on their
legal status included in the model. The results show that private (8.9%), governmental
(29.2%) and cooperative (11.7%) small firms have positive and significant effects on
regional GDP per capita, although governmental small firm has higher significant effect
than other mentioned types. The coefficient of public (-12.2%) small firms is negative
and indicates this type of firms are negatively related to regional GDP per capital
growth. The coefficient of official small firm is positive but insignificant. But as can be
seen, total number of small firm has strong positive and significant coefficient (96.3%)
10
and indicates that small firms can effect on regional GDP per capita positively and
significantly in general.
5.2. Impact of Small Firms on Regional GDP (without oil)
To extend our estimations, in the second model, we use of regional GDP (without oil)
growth as dependent variable and regression the model. The regression results show that
the effects of physical capital (K) and human capital (R&D) are positive and significant
in all estimations and support traditional Solow and Romer's view. The coefficients of
labor force are significant and positive in most cases and support our initial assumption
too. From columns 3 to 8, we use of different types of small firms in equation 2 and
estimate the model. As can be seen in tables 5 and 6, private (8.9%) and governmental
(33.11%) small firms have positive and public (-12.6%) small firm has negative and
significant effects on regional GDP growth. The coefficient of cooperative and official
small firms are insignificant, but similar to previous equation estimation, the result
shows a positive (92.7%) and significant effect of total number of small firms on
regional growth and suggesting that small firms promote growth in the Iranian regions.
5.3. Impact of Small Firms on Regional Industries Value Added
In the third part of model estimation, we use of industries value added (IVA) as
dependent variable. The positive and statistically significant coefficient of physical
capital (K) and labor force (in most of the cases) is consistent with the Solow model and
indicates that a region's industrial value added is positively related to the traditional
factors of production function. Model (2) reports that use of skilled labor in small firms
affect regional growth positively and significantly. In the columns 3 to 8,
entrepreneurship capital indicators include in equation 2. The results show that by using
of regional industries value added as dependent variable, the coefficients of private
(10.0%), Governmental (30.9%) and cooperative (18.6%) small firms on regional
growth are positive and significant strongly. The negative signs of the coefficients of
official and public firms are insignificant and similar to former estimations. Finally, the
estimation reported in column 8 indicates the positive and important role of small firms
in regional growth and this result keeps consistent for all regional growth indicators. It
should be noticed that positive relationship between regional growth and small firms
can possibly be related to the quality of institutions. As discussed in other studies, the
institutional quality might be an important factor explaining the economic performance
across regions. Better institutions also create the necessary conditions to encourage
more human capital formation that will be better used by productive entrepreneurs (Dias
and McDermott, 2006).
6. SUMMARY AND CONCLUSION
Following of Solow's (1956) neoclassical model of production function as a basis for
analyzing economic growth, many of new policy directions were developed to enhance
the two traditional factors of production, physical capital and unskilled labor. On the
other hand, endogenous growth theory which presented by Romer (1986) and Lucas
11
(1988) started a new direction focusing on improving knowledge capital through
investment in R&D, education and human capital. This paper indicates that these
approaches should consider an important factor that also shapes the economic
performance of a region: the entrepreneurship capital of that region. In this paper, we
measure the entrepreneurship capital by the number of small firms with private,
governmental, cooperative, official and public legal status. Based on a Cobb-Douglas
production function, empirical results found that entrepreneurship capital exerts positive
and significant impacts on regional GDP per capita, GDP and industries value added
growth. The results hold for private, governmental and cooperative small firms. The
results suggest that in the case of Iran, government policies should direct to
enhancement of entrepreneurship capital. It should be consider that our findings
certainly do not contradict the conclusions of earlier studies linking growth to factors
such that labor, physical capital and knowledge, whereas the evidence of this paper
points to an additional factor that named entrepreneurship capital. In other word, our
findings indicate that entrepreneurship capital plays an important role in the model of
production function. Although our findings cannot provide detailed policy
recommendations, but it can be stated that under certain conditions, policies focusing on
enhancing entrepreneurship capital and small firms performance can prove to be more
effective than those targeting more traditional factors. Whether these results hold for
other countries or for other time periods can only be ascertained through subsequent
researches.
12
13
0.4420 ***
0.5837 ***
(7.771)
(-1.847)
0.6815 ***
-0.3189 *
Ln(K)t
Ln (L)t
29.397 ***
0.6793
0.6509
23.9241 ***
382
R2
R2adjusted
F-value
Observation
Note 1: * p< 0.10, ** p< 0.05, *** p< 0.01.
0.7056
Yes
Fixed effect
380
0.7352
Yes
14.2605 ***
26.0700 ***
0.6543 ***
(10.704)
(-1.442)
(10.553)
(3.006)
(11.096)
(6.868)
T-value
377
27.1295 ***
0.6987
0.7252
Yes
29.7298 **
-0.0830
0.4647 ***
0.4407 ***
7.2461 ***
Coefficient
Model (3)
(4.738)
(11.295)
(8.964)
T-value
Chi2 stat for Hausman Test
Ln (Gov ent)t
Ln (Pri ent)t
0.6581 ***
7.8912 ***
(4.173)
5.2837 ***
Constant
Ln (R&D)t
Coefficient
T-value
Coefficient
Model (2)
Variable
Model (1)
Table (4). Estimation results of empirical model (dependent variable: GDP per capita)
(5.259)
(6.021)
(3.870)
(6.969)
(10.827)
T-value
284
36.8182 ***
0.8068
0.8293
Yes
18.8347 ***
0.2918 ***
0.6205 ***
0.6562 ***
0.4144 ***
9.0967 ***
Coefficient
Model (4)
14
(8.769)
(12.31)
(4.584)
(11.564)
(2.035)
7.9617 ***
0.4576 ***
0.5869 ***
0.6756 ***
0.1173 **
Constant
Ln(K)t
Ln (L)t
Ln (R&D)t
Ln (Coop ent)t
0.7048
28.4241 ***
380
0.7636
0.7406
33.2819 ***
374
R2
R2adjusted
F-value
Observation
Note 1: * p< 0.10, ** p< 0.05, *** p< 0.01.
Yes
Yes
Fixed effect
0.7305
25.5337 ***
(0.0197)
(10.669)
(3.489)
(11.279)
(8.144)
T-value
24.2979 ***
0.0023
0.6580 ***
0.5860 ***
0.4421 ***
7.8992 ***
Coefficient
Model (6)
Chi2 stat for Hausman Test
Ln (Total ent)t
Ln (Pub ent)t
Ln (Offic ent)t
T-value
Coefficient
Variable
Model (5)
(-2.509)
(12.886)
(7.110)
(10.609)
(10.181)
T-value
256
31.0929 ***
0.7835
0.8096
Yes
9.4032 ***
-0.1218 **
0.7956 ***
1.0147 ***
0.4427 ***
11.149 ***
Coefficient
Model (7)
Table (5). Estimation results of empirical model (dependent variable: GDP per capita) (continued)
(2.661)
(10.686)
(1.221)
(11.084)
(2.091)
T-value
380
29.2205 ***
0.7107
0.7360
Yes
31.8398 ***
0.9634 ***
0.6518 ***
0.5345
0.4320 ***
3.7391 **
Coefficient
Model (8)
15
0.4840 ***
0.6608 ***
(15.999)
(-1.433)
0.672 ***
-0.2006
Ln(K)t
Ln (L)t
51.846 ***
0.7716
0.7514
38.160 ***
382
R2
R2adjusted
F-value
Observation
Note 1: * p< 0.10, ** p< 0.05, *** p< 0.01.
0.8110
Yes
Fixed effect
380
0.8270
Yes
11.049 ***
17.733 ***
0.6825 ***
(10.591)
(1.975)
(10.584)
(-3.3495)
(11.394)
(13.931)
T-value
377
48.900 ***
0.8078
0.8247
Yes
20.529 ***
0.0899 *
-0.519 ***
0.4734 ***
14.584 ***
Coefficient
Model (3)
(5.106)
(11.776)
(16.531)
T-value
Chi2 stat for Hausman Test
Ln (Gov ent)t
Ln (Pri ent)t
0.6838 ***
15.2823 ***
(11.827)
11.6830 ***
Constant
Ln (R&D)t
Coefficient
T-value
Coefficient
Model (2)
Variable
Model (1)
(7.220)
(9.421)
(4.763)
(9.634)
(17.755)
284
63.331 ***
0.879
0.8931
Yes
37.268 ***
0.3311 ***
0.6277 ***
0.6104 ***
0.4220 ***
16.1129 ***
T-value
Model (4)
Coefficient
Table (6). Estimation results of empirical model (dependent variable: Gross Domestic Production (without oil))
16
0.6842 ***
(11.118)
0.6916 ***
-0.0741
Ln (R&D)t
Ln (Coop ent)t
Note 1: * p< 0.10, ** p< 0.05, *** p< 0.01.
374
0.8270
R2adjusted
Observation
0.8423
R2
55.038 ***
Yes
Fixed effect
F-value
17.896 ***
Chi2 stat for Hausman Test
Ln (Total ent)t
Ln (Pub ent)t
(-0.102)
(10.564)
(3.677)
(11.760)
(14.961)
T-value
380
50.132 ***
0.8105
0.8270
Yes
19.322 ***
-0.0129
0.6485 ***
(5.0431)
0.6874 ***
Ln (L)t
Ln (Offic ent)t
0.4840 ***
(12.662)
0.5009 ***
Ln(K)t
(-1.207)
15.2389 ***
(15.999)
15.4663 ***
Constant
Coefficient
T-value
Coefficient
Model (6)
Variable
Model (5)
(-2.440)
(12.988)
(6.750)
(10.215)
(15.699)
T-value
267
53.634 ***
0.8636
0.88
Yes
45.259 ***
-0.126 ***
0.8548 ***
1.0268 ***
0.5444 ***
18.325 ***
Coefficient
Model (7)
(2.434)
(10.563)
(0.902)
(11.570)
(6.001)
T-value
380
51.169 ***
0.8137
0.8299
Yes
15.9993 ***
0.9271 ***
0.6778 ***
0.4153
0.4744 ***
11.2866 ***
Coefficient
Model (8)
Table (7). Estimation results of empirical model (dependent variable: Gross Domestic Production(without oil)) (continued)
17
0.4363 ***
-0.3182 ***
(14.28)
(1.818)
0.6141 ***
0.1995 *
Ln(K)t
Ln (L)t
No
0.5778
173.8967 ***
No
0.4788
0.4760
174.103 ***
382
Fixed effect
R2
R2adjusted
F-value
Observation
377
126.632 ***
0.5720
0.5765
No
5.0491
(1.653)
(9.726)
(-1.141)
(9.757)
(17.722)
T-value
Note 1: * p< 0.10, ** p< 0.05, *** p< 0.01. Note 2: Model (1), (2) and (3) estimated by Random effect method.
380
0.5811
4.6752
3.9284
0.668 ***
(9.750)
0.1003 *
-0.1607
0.423 ***
10.120 ***
Coefficient
Model (3)
(-2.778)
(10.109)
(13.127)
T-value
Chi2 stat for Hausman Test
Ln (Gov ent)t
Ln (Pri ent)t
0.6729 ***
10.8826 ***
(8.866)
6.8657 ***
Constant
Ln (R&D)t
Coefficient
T-value
Coefficient
Model (2)
Variable
Model (1)
Table (8). Estimation results of empirical model (dependent variable: Industrial value added)
(6.279)
(9.267)
(2.639)
(7.846)
(12.649)
T-value
284
82.569 ***
0.9048
0.9159
Yes
17.576 ***
0.309 ***
0.664 ***
0.364 ***
0.369 ***
12.355 ***
Coefficient
Model (4)
18
0.6755 ***
(10.611)
0.6979 ***
0.1860 ***
Ln (R&D)t
Ln (Coop ent)t
0.6055
R2adjusted
380
130.1808 ***
0.5768
(-1.616)
(10.948)
(4.7469)
(8.870)
(11.212)
T-value
267
67.8481 ***
0.8894
0.9027
Yes
18.6244 ***
-0.0936
0.8060 ***
0.8077 ***
0.4414 ***
14.641 ***
Coefficient
Model (7)
Note 1: * p< 0.10, ** p< 0.05, *** p< 0.01. Note 2: Model (5), (6) and (8) estimated by Random effect method.
374
0.6097
R2
Observation
No
No
Fixed effect
0.5813
4.6966
144.128 ***
(9.737)
(-0.3910)
4.9103
F-value
(10.093)
(11.266)
T-value
(-1.5707)
Chi2 stat for Hausman Test
Ln (Total ent)t
Ln (Pub ent)t
-0.0528
-0.2702
(-2.471)
-0.2895 **
Ln (L)t
Ln (Offic ent)t
0.4365 ***
(11.627)
0.4580 ***
Ln(K)t
(2.9919)
10.7166 ***
(12.799)
10.7702 ***
Constant
Coefficient
T-value
Coefficient
Model (6)
Variable
Model (5)
(2.355)
(9.602)
(1.850)
(10.152)
(4.467)
380
133.32 ***
0.5827
0.5871
No
6.0226
0.9039 **
0.6611 ***
0.6696 *
0.4359 ***
7.4295 ***
T-value
Model (8)
Coefficient
Table (9). Estimation results of empirical model (dependent variable: Industrial value added) (continued)
REFERENCES
Acs, Z., Braunerhjelm, P., Audretsch, D., and Carlson, B. (2009). The Knowledge
Spillover Theory of Entrepreneurship. Small Business Economics, 2, 15-30.
Audretsch, D.B., and Fritsch, M. (1996). Creative destruction: turbulence and Economic
growth, in Ernst Helmstad terand Mark Perlman(eds.), Behavioral Norms,
Technological Progress, and Economic Dynamics: Studies in Schumpeterian
Economics,pp.137–150.AnnArbor:University of Michigan Press.
Audretsch, D.B., and Fritsch, M.(2002). Growth Regimes over Time and Space.
Regional Studies, 36, 113–124.
Audretsch, D.B., and Keilbach, M. (2004a). Entrepreneurship Capital and Economic
Performance. Regional Science, 38(8), 949-959.
Audretsch, D.B., and Keilbach, M. (2004b). Entrepreneurship and regional Growth: An
Evolutionary Interpretation. Journal of Evolutionary Economics, 14, 605-616.
Audretsch, D.B., and Keilbach, M. (2005). Entrepreneurship Capital and regional
Growth. The Annals of Regional Science, 39, 457-469.
Audretsch, D.B., Keilbach, M.C., and Lehmann, E.E. (2006). Entrepreneurship and
Economic Growth. Oxford University Press.
Audretsch, D.B., Bönte, W., and Keilbach, M. (2008). Entrepreneurship Capital and its
Impact on Diffusion and Economic Performance. Journal of Business Venturing, 23,
687-698.
Baptista, R., Escaria, V., and Madruga, P. (2008). Entrepreneurship, Regional
Development and Job Creation: The Case of Portugal. Small Business Economics,
30, 49-58.
Baumol, M. (1968). Entrepreneurship in Economic Growth. American Economic
Review, 58, 64-71.
Callejon, M., and Segarra, A. (1999). Business Dynamics and Efficiency in Industries
and Regions: The Case of Spain. Small Business Economics, 13, 253-271.
Chun-Chien, K., and Chai-Hai, Y. (2008). Knowledge Capital and Spillover on
Regional Economic Growth: Evidence from China. China Economic Review, 19,
594-604.
Cohen, W., and Levinthal, D. (1989). Innovation and Learning: The Two Faces of
R&D. Economic Journal, 99, 569–596.
Cravo, T.A., Becker, B., and Gourlay, A. (2014). Regional Growth and SMEs in Brazil:
A Spatial Panel Approach. Regional Studies, 49, 1995-2016.
19
Dias, J., and McDermott, J. (2006). Institutions, Education, and Development: The Role
of Entrepreneurs. Journal of Development Economics, 80, 299-328.
Feldman, M. P., and Audretsch, D. B. (1999). Innovation in Cities: Science Based
Diversity, Specialization and Localized Competition. European Economic Review,
43, 409–429.
Fischer, M.M., and Nijkamp, P. (1988). The Role of Small Firms for Regional
Revitalization. Annals of Regional Science, 22 (1), 28-42.
Foelster, S. (2000). Do Entrepreneurship Create Jobs?. Small Business Economics, 14,
137-148.
Fritsch, M., and Mueller, P. (2004). Effects of New Business Formation on Regional
Development over Time. Regional Science, 38(8), 961-975.
Gleaser, E., Kallal, H., Scheinkman, J., and Shleifer, A. (1992). Growth of cities.
Journal of Political Economy, 100, 1126–1152.
Grossman, G. M., and Helpman, E. (1991). Innovation and Growth in the Global
Economy. MIT Press, Cambridge MA.
Hannan, M. T., and Freeman, J. (1989). Organizational Ecology. Harvard University
Press, Cambridge, MA.
Hart, M., and Hanvey, E. (1995). Job Generation and New Small Firms: Some Evidence
from the Late 1980s. Small Business Economics, 7, 97-109.
Holtz-Eakin, D., and Kao, Ch. (2003). Entrepreneurship and Economic Growth: The
Proof is in Productivity. Center for Policy Research Working Paper, Maxwell
School, Syracuse University.
Jacobs, J.(1969). The Economy of Cities. Vintage, New York.
Lucas, R. E. (1988). On the mechanics of economic development. Journal of Monetary
Economics, 22, 3–39.
Mueller, P. (2007). Exploiting Entrepreneurial Opportunities: The Impact of
Entrepreneurship on Growth. Small Business Economics, 28, 355-362.
Piergiovanni, R., Carree, M., and Santarelli, E. (2009). Creative Industries, New
Business Formation and Regional Economic Growth. Jena Economic Research
Papers.
Porter, M. (1990). The Comparative Advantage of Nations. Free Press, New York.
Romer, P. M. (1986). Increasing returns and long-run growth. Journal of Political
Economy, 94, 1002–1037.
20
Solow, R. (1956). A Contribution to the Theory of Economic Growth. Quarterly
Journal of Economics, 70, 65-94.
Van Stel, A., and Suddle, K. (2008). The Impact of New Firm Formation on Regional
Development in the Netherlands. Small Business Economics, 30, 31-47.
Variable
Y
Y per cap
IVA
K
L
R&D
Private Ent
Governmental Ent
Cooperative Ent
Official Ent
Public Ent
Total Ent
variables
Appendix 1. Variable definitions and sources
Variable Definition
Regional gross domestic production (without oil)
Regional gross domestic production per capita
Regional value added in industrial sectors
Stock of physical capital used in small manufacturing firms
Total number of employees who work in small manufacturing firms
Number of the two most qualified labor categories who work in small
manufacturing firms
Number of private small manufacturing firms
Number of governmental small manufacturing firms
Number of cooperative small manufacturing firms
Number of official small manufacturing firms
Number of public small manufacturing firms
Total number of manufacturing small firms
Data Source
Iranian regional accounts
Iranian regional accounts
Iranian regional accounts
Statistical center of Iran
Statistical center of Iran
Statistical center of Iran
Statistical center of Iran
Statistical center of Iran
Statistical center of Iran
Statistical center of Iran
Statistical center of Iran
Statistical center of Iran
Appendix 2. Panel unit root test (in level)
Individual intercept
Individual intercept and trend
Method
Statistics Prob*
Method
Statistics
Prob
Ln (GDP per
capita)
Levin, Lin &
Chu
Im,Pesaran&Shin
ADF-Fisher Chi2
PP-Fisher Chi2
Ln (GDP)
Levin, Lin &
Chu
PP-Fisher Chi2
-13.6343
0.0000
-9.3212
196.48
220.119
0.0000
0.0000
0.0000
-7.8821
0.0000
92.485
0.0045
-16.5990
0.0000
230.558
288.185
0.0000
0.0000
Ln (IVA)
Levin, Lin &
Chu
ADF-Fisher Chi2
PP-Fisher Chi2
-5.9765
0.0000
Ln (K)
Levin, Lin &
Chu
Im,Pesaran&Shin
ADF-Fisher Chi2
PP-Fisher Chi2
-1.9363
77.2385
87.1671
Ln (L)
Levin, Lin &
-7.7122
Levin, Lin &
Chu
Breitung T-test
Im,Pesaran&Shin
ADF-Fisher Chi2
Levin, Lin &
Chu
Breitung T-test
Levin, Lin &
Chu
ADF-Fisher Chi2
PP-Fisher Chi2
-7.7151
0.0000
-2.035
-2.3777
91.4214
0.0209
0.0087
0.0056
-8.3641
0.0000
-2.0054
0.0225
-4.3632
0.0000
75.6128
88.6992
0.0842
0.0094
Result
Stationary
Stationary
Stationary
-8.4546
0.0000
0.0264
0.0663
0.0125
Levin, Lin &
Chu
Im,Pesaran&Shin
ADF-Fisher Chi2
PP-Fisher Chi2
-4.2072
113.542
162.562
0.0000
0.0000
0.0000
Stationary
0.0000
Levin, Lin &
-11.9812
0.0000
Stationary
21
variables
Appendix 2.
Individual intercept
Method
Statistics
Chu
Im,Pesaran&Shin -4.0277
ADF-Fisher Chi2 109.357
111.131
PP-Fisher Chi2
Ln (R&D)
Levin, Lin &
Chu
Ln (Pri ent)
Levin, Lin &
Chu
Im,Pesaran&Shin
ADF-Fisher Chi2
PP-Fisher Chi2
Ln (Gov ent)
Levin, Lin &
Chu
Im,Pesaran&Shin
PP-Fisher Chi2
Ln (Coop
ent)
Levin, Lin &
Chu
Im,Pesaran&Shin
ADF-Fisher Chi2
Ln (Offic
ent)
Levin, Lin &
Chu
Im,Pesaran&Shin
ADF-Fisher Chi2
PP-Fisher Chi2
Ln (Pub ent)
Levin, Lin &
Chu
Im,Pesaran&Shin
ADF-Fisher Chi2
Levin, Lin &
Chu
Ln (Total
Im,Pesaran&Shin
ent)
ADF-Fisher Chi2
PP-Fisher Chi2
* Probabilities for Fisher tests are
assume asymptotic normality.
Panel unit root test (in level)
Individual intercept and trend
Prob*
Method
Statistics
Prob
Chu
0.0000 Im,Pesaran&Shin -4.4034 0.0000
0.0001 ADF-Fisher Chi2 124.149 0.0000
0.0001
PP-Fisher Chi2
128.662 0.0000
-1.8045
0.0356
-7.8081
0.0000
-5.2417
129.282
126.752
0.0000
0.0000
0.0000
-3.0432
0.0012
-1.5752
75.5164
0.0576
0.0022
-7.7086
0.0000
-4.0200
109.294
0.0000
0.0001
-6.1178
0.0000
-4.052
111.406
132.72
Levin, Lin &
Chu
Im,Pesaran&Shin
ADF-Fisher Chi2
PP-Fisher Chi2
Levin, Lin &
Chu
Im,Pesaran&Shin
ADF-Fisher Chi2
PP-Fisher Chi2
Levin, Lin &
Chu
Breitung T-test
ADF-Fisher Chi2
Levin, Lin &
Chu
ADF-Fisher Chi2
PP-Fisher Chi2
-8.3642
0.0000
-2.1581
87.6476
116.261
0.0155
0.0115
0.0000
-18.9906
0.0000
-10.0757
182.830
211.222
0.0000
0.0000
0.0000
-10.4967
0.0000
-1.3304
72.9211
0.0917
0.0117
-8.0615
0.0000
78.3734
108.065
0.0558
0.0001
-8.5628
0.0000
0.0000
0.0001
0.0000
Levin, Lin &
Chu
Im,Pesaran&Shin
ADF-Fisher Chi2
PP-Fisher Chi2
-2.5758
98.8991
1333.96
0.0050
0.0018
0.0000
-8.0254
0.0000
Im,Pesaran&Shin
-1.4972
0.0672
-3.6884
92.0020
0.0001
0.0001
ADF-Fisher Chi2
PP-Fisher Chi2
83.5456
98.155
0.0006
0.0000
Result
Stationary
Stationary
Stationary
Stationary
Stationary
Stationary
Levin, Lin &
-7.2975 0.0000
Chu
-2.5886 0.0048 Im,Pesaran&Shin -2.7342 0.0031 Stationary
86.9917 0.0130 ADF-Fisher Chi2 99.9796 0.0009
108.923 0.0001
PP-Fisher Chi2
121.983 0.0000
computed using an asymptotic Chi-square distribution. All other tests
-5.3889
0.0000
22
Appendix 3. Ratio of small firms groups to total number of small firms (in 2013)
Province
(Pri ent/total
ent) × 100
East Aza
38.5
West Aza
58.4
Ardabil
32.8
Esfahan
46.5
Ilam
33.3
Bushehr
0
Tehran
42.7
Bakhtiari
47
S. Khorasan
27.2
Khorasan Ra.
35.5
N. Khorasan
79
Khuzestan
19.5
Zanjan
16.4
Semnan
39.8
Sistan
78.8
Fars
44.3
Ghazvin
29.8
Ghom
27
Kordestan
41.5
Kerman
44.5
Kermanshah
40
Kohkiloyeh
54.2
Golestan
19.6
Gilan
20
Lorestan
24
Mazandaran
24
Markazi
23.5
Hormozgan
45
Hamadan
21.5
Yazd
10.5
Source: Authors’ estimation
(Gov ent/total
ent) × 100
(Coop ent/total
ent) × 100
(Offic ent/total
ent) × 100
(Pub ent/total
ent) × 100
1.2
0.8
3.5
0.08
8.3
0
0.2
1
0
0.15
0
2.1
0
0.3
2.2
0.45
0
0.3
0.9
0.5
1
0
0.4
1
0
0.2
1.3
1
0
0
1.2
9.4
14
1.1
12.5
6.7
0.7
3
6.8
4
12.7
6.5
6.9
2.7
3.3
6
1
2.9
12.2
3.6
9.2
20.8
7.5
10
11
6.1
8.3
9
9
2.3
58.6
30.2
47.5
52
45.8
93.3
56.2
49
66
61.5
6.3
69.7
75
56.5
14.4
49.2
69.2
68.9
39.6
51.3
48.8
25
71
69
63
68.5
66.5
45
68.5
85.5
0.5
0.8
1.4
0.08
0
0
0.2
0
0
0.85
0
2.6
1.7
0.3
1.3
0
0
0.3
5.6
0.5
1
0
1.5
0
1
1.2
0.3
0
1
1.7
23