Growth persistence and profile robustness of high-growth firms

Abstract

This paper analyses the persistence of high business growth and the robustness of the profile characteristics of high-growth firms (HGFs). By having company information for all firms that are active in Flanders (i.e., the northern part of Belgium) for a ten-year period (i.e., from 2000 to 2009), different subsets of HGFs were identified for different time periods. Several questions arise, such as whether the firms that were qualified as an HGF in a certain period were able to maintain the high growth rates for multiple (consecutive) periods and whether the profile characteristics of the HGF-subsets are stable over time. It appeared that the majority of the firms that were identified as an HGF in the period 2000-2009 were 'one-shot HGFs'. Notwithstanding the rapidly changing composition of the subsets, the profile features of the HGFs in the subset remained relatively constant over time.

Full text

I
nt. J. Entrepreneurial Venturing, Vol. 6, No. 4, 2014 299
Copyright © 2014 Inderscience Enterprises Ltd.
Growth persistence and profile robustness of
high-growth firms
Yannick Dillen*
Department Management, Faculty of Applied Economics,
University of Antwerp,
Prinsstraat 13, 2000 Antwerp, Belgium
and
Chamber of Commerce and Industry Antwerp,
Markgravestraat 12, 2000 Antwerp, Belgium
E-mail: yannick.dillen@voka.be
*Corresponding author
Eddy Laveren and Rudy Martens
Department Management, Faculty of Applied Economics,
University of Antwerp,
Prinsstraat 13, 2000 Antwerp, Belgium
E-mail: eddy.laveren@ua.ac.be
E-mail: rudy.martens@ua.ac.be
Sven De Vocht and Eric Van Imschoot
Chamber of Commerce and Industry Antwerp,
Markgravestraat 12, 2000 Antwerp, Belgium
E-mail: sven.devocht@voka.be
E-mail: eric.vanimschoot@voka.be
Abstract: This paper analyses the persistence of high business growth and the
robustness of the profile characteristics of high-growth firms (HGFs). By
having company information for all firms that are active in Flanders (i.e., the
northern part of Belgium) for a ten-year period (i.e., from 2000 to 2009),
different subsets of HGFs were identified for different time periods. Several
questions arise, such as whether the firms that were qualified as an HGF in a
certain period were able to maintain the high growth rates for multiple
(consecutive) periods and whether the profile characteristics of the
HGF-subsets are stable over time. It appeared that the majority of the firms that
were identified as an HGF in the period 2000–2009 were ‘one-shot HGFs’.
Notwithstanding the rapidly changing composition of the subsets, the profile
features of the HGFs in the subset remained relatively constant over time.
Keywords: growth; business growth; firm growth; rapid-growth firms;
high-growth firms; HGFs; growth persistence.
Reference to this paper should be made as follows: Dillen, Y., Laveren, E.,
Martens, R., De Vocht, S. and Van Imschoot, E. (2014) ‘Growth persistence
and profile robustness of high-growth firms’, Int. J. Entrepreneurial Venturing,
Vol. 6, No. 4, pp.299–330.

300 Y. Dillen et al.
Biographical notes: Yannick Dillen is a PhD student at the Faculty of Applied
Economics (Department Management) at the University of Antwerp, Belgium.
His research focuses on business growth and high-growth firms. He is affiliated
to the Chamber of Commerce and Industry Antwerp where he works on project
on high-growth firms.
Eddy Laveren is a Professor of Finance and Entrepreneurship in the
Department of Accounting and Finance at the Faculty of Applied Economics at
the University of Antwerp, Belgium. His research interests include financial
management and growth management of small firms and family businesses.
Rudy Martens is the Dean of the Faculty of Applied Economics at the
University of Antwerp, Belgium. His teaching is situated in the field of
strategic and general management. His research focuses mainly on strategy
processes, knowledge management and management of SMEs.
Sven De Vocht is Director of Entrepreneurship at the Chamber of Commerce
and Industry Antwerp.
Eric Van Imschoot is Administrative Director at the Chamber of Commerce
and Industry Antwerp.
1 Introduction
In the last decade, there has been an increasing awareness of the importance of
high-growth firms (HGFs) for the employment growth and dynamics in a region or
economy (European Commission, 2011; Harms et al., 2010). This awareness has –
together with the raised availability of high-quality data on business growth – produced a
number of excellent studies in the research field of HGFs. The studies of Davidsson and
Henrekson (2002) and Barringer et al. (2005) gave interesting insights into the attributes
and determinants that can be associated with rapid firm growth, whereas the study of
Delmar et al. (2003) found the growth pattern of HGFs during the high-growth period to
be extremely heterogeneous. Recently, some studies have pointed to the episodic nature
of high business growth. Being a HGF appears to be frequently a temporary phenomenon
(Garnsey et al., 2006). Mason and Brown (2010) stated in a research report that HGFs
indeed appear to have difficulty in sustaining their frenetic pace of growth, whereas Acs
et al. (2008) found that the most typical pattern for former HGFs is to return to the
average industry growth rates. Also Parker et al. (2010) concluded that firms may show
fast growth for short periods of time, but only very few are able to maintain this growth
pace into the medium or long term. The approach of Parker et al. (2010) was novel in the
way that they identified a group of HGFs in one period and tracked their subsequent
performance over time.
We try to get some more profound insights into the phenomenon of growth
persistence which can be defined in this context as the capability of firms to retain strong
positive growth rates for a long period of time. Concretely, strong growth rates are
measured in two ways: the change in the number of employees and the change in the
yearly value added. Our focus will be on the growth persistence of HGFs with an
approach that differs from the one that has been used in other studies. By having

Growth persistence and profile robustness of high-
g
rowth firms 301
company information over a ten-year period, different subsets of HGFs were identified
for different time periods. The question arises
1 whether the firms that were qualified as a HGF in a certain period were able to
maintain the high growth rates for multiple periods
2 whether the profile characteristics (e.g., firm size, firm age, industry, total assets,
solvency, …) of such a HGF-subset are consistent over time.
Hence, we not only try to gain a deeper insight into the growth persistence of HGFs but
also to learn more about the stability or ‘robustness’ of the profile characteristics of the
different subsets of HGFs. Stated differently, we will test if the profile features of HGFs
remain stable over different time periods and different groups of HGFs. Up to now,
studies identified a group of HGFs, analysed their profile characteristics and compared
them to the profile characteristics of the remainder of the business population for the
same period. In our approach, the profile characteristics of HGFs can be compared over
time and between different HGF-subsets.
Only a very limited number of HGFs could hold their strong growth rates for a long
time period. This finding has implications from a managerial and academic perspective
but especially from a policy perspective. Initiatives that have been undertaken in recent
years by government and other institutions to support HGFs should be aware of the fact
that the clear majority of the firms that can be qualified as a HGF at a certain moment in
time will only be a HGF for a relatively short period. Measures that are taken to support
these firms should be seen in the light of the short-lived character of high firm growth.
These initiatives should primarily be directed to the firms with the potential to become a
HGF in a later phase. The challenge then is to identify the firms that have the potential of
becoming a HGF in the future. Notwithstanding the high in-and outflow of firms in the
HGF-subsets, we found that the profile features of these different subsets remain
relatively constant over time which can be a reassurance for policy makers and
government institutions who have launched initiatives towards HGFs as the initiatives
may benefit from the finding that the subsets consist of more or less the same type of
firms. Supportive measures should then not ever be adjusted to the changing outlook of
the HGF-subset.
This paper proceeds as follows. First, some theoretical perspectives on the length of
the high-growth period will be introduced. Second, the database and the research method
are described. Third, we analyse the persistence of the high-growth phenomenon and the
results are discussed. Fourth, the different profile characteristics of the identified HGFs
are shown and compared, including a comparison between some features of the so-called
‘one-shot HGFs’ and ‘persistent HGFs’. Finally, some concluding comments are offered.
2 Theoretical perspectives on the persistence of high business growth
Different theories and theoretical concepts exist that can be linked to the persistence of
high business growth. In this section, we will discuss the life cycle theory of the firm, the
concepts of minimum efficient size (MES), sustained competitive advantage, Gibrat’s
law and the growth theory of Penrose.

302 Y. Dillen et al.
2.1 Life cycle theory of the firm
Following the life cycle theory of a firm, business growth can be considered as a natural
process where every company will go through (Churchill and Lewis, 1983; Greiner,
1972; Kanzanjian, 1988). A firm will grow via sequential steps, from birth to maturity,
where each phase can be characterised by priorities, configuration issues and strategies.
In the phase model of Greiner (1972) a company goes through various stages of evolution
and revolution where each growth phase is associated with series of internal crises. After
a crisis is solved a period of growth can follow as a result of which the company enters
the second phase of the life cycle. By each time surviving the crisis that is unique to a
specific stage, a firm can evolve to the last phase of the life cycle. Phelps et al. (2007)
identified different states that can be related to managerial problems. To encounter a
period of growth, a firm must successfully resolve the challenges that are associated by
each phase or state. In this logic, a period of growth in general and a period of high
growth in particular is a temporary phenomenon in the life of a firm which may occur
several times in the life of a firm.
2.2 Minimum efficient size
As to Lipczynski et al. (2005), the long-term survival of small firms depends on their
ability to achieve the so-called minimum efficient scale or MES. Almus (2002) defined
the MES as a certain threshold that newly founded businesses should reach if they want
to survive in their sector. Small firms that are not able to achieve the MES tend to exit. It
can be expected therefore that faster-than-average growth may occur among a cohort of
surviving firms when these firms are striving to reach the MES (Lipczynski et al., 2005).
As this MES varies strongly per sector, firms that are active in sectors characterised by a
high MES often will have a greater tendency to grow fast (Almus, 2002). Consequently,
high-firm growth can be linked to the fact that firms need to grow rapidly in order to
achieve the MES. Once this threshold has been attained, however, growth rates are
expected to become more moderate. Again, following this logic, a high-growth period
can be seen as a temporary phase in the life of a firm.
2.3 Sustained competitive advantage
Also the resource-based view can give some explanations to the persistence of high
business growth. As to Barney (1991), a firm is said to have a competitive advantage if a
value creating strategy is implemented which is not at the same time implemented by a
current or potential competitor. A sustained competitive advantage arises then when these
competitors are not able to duplicate the benefits of the value creating strategy.
Concretely, a firm may obtain a sustained competitive advantage by “implementing
strategies that exploit their internal strengths, through responding to environmental
opportunities, while neutralizing external threats and avoiding internal weaknesses”
[Barney, (1991), p.99]. To have the potential to build up a competitive advantage, a firm
resource has to be valuable1 and rare among competitors. Apart from being valuable and
rare, firm resources have to be imperfectly imitable and not easily substitutable for
strategically equivalent resources to have the potential for contributing to a sustained
competitive advantage (Barney, 1991). HGFs could relate the extreme growth rates to a
certain competitive advantage that enables this growth. The distinction between a

Growth persistence and profile robustness of high-
g
rowth firms 303
competitive advantage and a sustained competitive advantage can then be linked to the
growth persistence of HGFs. Firms that can maintain the HGF-status for several
consecutive periods are more likely to have a sustained2 competitive advantage that will
not be competed away by possible duplication efforts of competing firms. The fact that
achieving a sustained competitive advantage is much more difficult for a firm than
achieving a competitive advantage – as the resources that are at the basis of the sustained
competitive advantage have to be apart from valuable and rare also imperfectly imitable
and not easily substitutable (cf., Barney, 1991) – could imply that extending the
high-growth period for a longer time is a difficult task for a large part of the HGFs.
Consequently, the amount of ‘one-shot HGFs’, i.e., the firms that are only a HGF for a
short period (cf., supra), would be much higher than the amount of ‘persistent HGFs’,
i.e., the firms that retain high growth rates for a long period of time, as moving from a
competitive advantage to a sustained competitive advantage may be only reserved for the
happy few.
2.4 Gibrat’s Law
Robert Gibrat tried to explain the differences in growth rates between firms already back
in 1931. He stated in his famous Gibrat’s Law that firm growth rates are independent
random variables. This implies that firms that grow faster in one period will not grow
faster (or slower) than other firms in a later time period (Gibrat, 1931). However, Sutton
(1997) conducted a review of Gibrat’s Law and concluded that growth rates were serially
correlated: firms that grew faster in one period were more likely than others to have an
above-average growth rate in subsequent periods. A study of Coad (2009) found, on the
other hand, that the serial correlation of growth rates is more likely to be negative for
firms that have experienced extreme growth rates. Even more than 80 years after the
emergence of Gibrat’s Law, evidence is mixed regarding the serial correlation of firm
growth. A large proportion of so-called ‘one-shot HGFs’ (cf., infra) would be consistent
with Gibrat’s Law. If there are, on the other hand, many firms that can hold their
HGF-status for multiple periods, this would be contrary to the propositions of Gibrat.
2.5 Growth theory of Penrose
Also the growth theory of Penrose (1959) may give important insights into the
persistence of business growth. In her view, the bundle of resources a firm possesses, will
limit the opportunities a firm can seize (Kor and Mahoney, 2004). More specifically,
management availability is considered as the most constraining element to firm growth.
Penrose (1959) stated that managers function as a catalyst in the conversion of the firm’s
resources into capabilities. However, the pool of managerial resources is limited and not
easily transferrable from one firm to another as the experience of the management with
the firm-specific resources produces knowledge that is unique to the firm. Hence, the
experience-based knowledge is proprietary as it cannot be easily purchased on the market
and transferred to new managers (Kor and Mahoney, 2004). The availability and quality
of the management is consequently a bottleneck for the growth rates that a firm can
achieve in the long run. Attracting new managers may be necessary for a HGF to
overcome certain growing pains, but the effectiveness of the appointment of these new
managers may be limited as their knowledge is not fully applicable to the new

304 Y. Dillen et al.
firm-specific setting. This may impede HGFs to keep on growing fast for a longer time
period.
As was stated in the introduction, high firm growth is often characterised by its
episodic nature. The five concepts and theories that are mentioned in this theoretical
section all give explanations for the potential difficulties that firm face to retain high
growth rates for a long time period. In the empirical analysis that we have conducted, it is
analysed to what extent HGFs are capable of sustaining their high growth for multiple
periods, by making a distinction between ‘one-shot HGFs’ and ‘persistent HGFs’. The
way in which the analysis is carried out is new and can give more profound insights into
the described phenomenon of temporary high growth.
3 Method
3.1 Data
The Central Balance Sheet Office of the National Bank of Belgium provided an extensive
database that consists of company information for all firms that were active in Flanders
(i.e., the northern part of Belgium) in the period 2000–2009. Hence, financial statement
information of an entire decade is available for about 200.000 firms. This makes a
longitudinal approach possible. The database contains a wide variety of items such as
firm age, sector, year of foundation as well as a number of balance sheet items. In our
study we selected only those firms for which all the necessary data are available. In
accordance with an increasing number of studies in the domain, the OECD-definition of a
HGF will be used as a guideline (e.g., BERR, 2008; Mason and Brown, 2010;
Anyadike-Danes et al., 2009; Mason et al., 2009; Teruel and de Wit, 2011). According to
this definition, a HGF is a firm:
“that experiences annualized growth rates in employees or turnover greater than
20 percent per annum over a three year period with a minimum of 10
employees at the beginning of the study period.” [Eurostat – OECD,
(2007), p.61]
In order to avoid small businesses from becoming too dominant in the group of HGFs, a
minimum employee size is imposed of ten employees in the first year of the analysed
period. The average annual growth percentage is set on 20% and has to be reviewed
along the entire period of time. This implies that it is not necessary for the business to
grow more than 20% in each of the three years. Practically, a firm should have a growth
rate of at least 72.8%3 during the entire period in order to be classified as a HGF. The
OECD-definition normally leaves the choice between two growth measures, i.e.,
employment and turnover. However, in our analysis, turnover will be replaced by value-
added as a growth indicator. First, there is a pragmatic reason for this as the majority of
the Belgian SMEs are not obliged to publish turnover figures, whereas the publication of
value-added figures is mandatory for all Belgian firms. However, another and perhaps
even more important reason for the choice for value-added as growth measure is the fact
that value-added figures also have an important social value as the sum of all value-added
figures is a building block of the domestic product of a region. Hence, the two growth
measures that are used have both – directly or indirectly – an impact on the prosperity of
a region as strong employment growth gives rise to job creation and strong value-added
growth enables GDP growth.

Growth persistence and profile robustness of high-
g
rowth firms 305
Hence, we consider a HGF as a firm:
“that realizes annualized growth rates in employees or value-added figures that
are greater than 20 percent per annum over a three year period with a minimum
of 10 employees at the beginning of the study period.”
As company information is available for a ten-year period (i.e., from 2000 to 2009),
seven overlapping three-year periods can be identified (cf., Figure 1). Furthermore, two
types of HGFs will be analysed for each period, namely ‘employment HGFs’ and
‘value-added HGFs’. ‘Employment HGFs’ are the firms that have realised an annual
employee growth of at least 20% for a three-year period, whereas ‘value-added HGFs’
are the firms that have realised an annual growth in their value-added of at least 20% for
a three-year period. A total of 14 HGF-subsets were created (i.e., seven subsets of
‘employment HGFs’ and seven subsets of ‘value-added HGFs’). A schematic overview is
presented in Figure 1.
Figure 1 Schematic overview of the adopted approach
3.2 Research method
Two topics will be analysed (cf., Figure 1). First, the overlap between the different
HGF-subsets is assessed, which will provide insight into the phenomenon of the growth
persistence of HGFs. Second, some profile characteristics – such as the firm size, firm
age and industry – will be analysed for the 14 HGF-subsets and a comparison will be
made between them. Hence, it is tested if these characteristics are stable over time. The
potential differences between ‘employment HGFs’ and ‘value-added HGFs’ will be
discussed as well.
Table 1 gives an overview of the absolute and relative number of HGFs in the 14
identified subsets. Starting from the Flemish business population and after filtering for
five excluded NACE-sections4 and the minimum employee size of ten employees in the
first year of the analysed period, an absolute number of firms ranging from 11,504 (in the
period 2000–2003) to 13,295 (in the period 2006–2009) is obtained. After applying the
OECD-definition for the seven overlapping three-year periods, a relative number of

306 Y. Dillen et al.
‘employment HGFs’ ranging from 3.00% (for the period 2000–2003) to 4.39% (for the
period 2005–2008) was found. With respect to the ‘value-added HGFs’, a minimum
percentage of 7.08% (for the period 2000–2003) and a maximum percentage of 9.75%
(for the period 2005–2008) was encountered. Hence, it appears that the number of
‘value-added HGFs’ is approximately twice as high as the number of ‘employment
HGFs’ and that the number of HGFs has steadily risen from the first until the sixth
analysed period with a relapse in the last three-year period. A possible explanation for the
lower number in the last period is the fact that this period was more affected by the
financial and economic crisis. The overlap between the two HGF-categories is
represented in the last column of Table 1. A relatively high number of firms can be
catalogued both as an ‘employment HGF’ and ‘value-added HGF’ in the same period.
For instance, 370 of the 1,008 ‘value-added HGFs’ (i.e., 36.70%) for the period
2006–2009 could also be qualified as an ‘employment HGF’ in this period.
Table 1 Composition of the 14 HGF-subsets
Period
Employment HGFs
(seven subsets)
Value-added HGFs
(seven subsets) Absolute number
of firms that are
in both
HGF-subsets
Absolute
number
Relative*
number
Absolute
number
Relative*
number
2000–2003 345 3.00% 814 7.08% 231
2001–2004 381 3.16% 917 7.60% 259
2002–2005 407 3.28% 944 7.61% 281
2003–2006 461 3.67% 1,137 9.05% 314
2004–2007 551 4.25% 1,240 9.56% 406
2005–2008 583 4.39% 1,296 9.75% 433
2006–2009 489 3.68% 1,008 7.58% 370
Note: *Which is the absolute number relative to the number of Flemish firms with at least
ten employees in the first year of the period.
Testing the statistical significance of the differences between the average firm age, firm
size, total assets and solvency of the different subsets has been done by applying the
student’s t-test which is identical to the one-way ANOVA done on data with two
categories. The average values of each three-year period are compared to each other.
P-values lower than 0.05 indicate a significant difference between the average firm age,
firm size, total assets or solvency of two different HGF-subsets. The tables with the
detailed calculations of the p-values can be found in Appendix 3.
4 Assessing the overlap between the subsets: how persistent is high
business growth?
Based on the earlier mentioned theoretical considerations, one would expect to encounter
a limited number of HGFs that are able to realise high growth rates for a long period of
time. The identification of the seven (overlapping) HGF-subsets offers the opportunity to
learn more about this persistence of high business growth. Table 2 gives an overview of
the number of firms that could be qualified as a HGF in one or more three-year periods. It

Growth persistence and profile robustness of high-
g
rowth firms 307
appears that 1 908 different firms could be qualified as an ‘employment HGF’ in one or
more of the seven periods and that 4,319 different firms were identified as a ‘value-added
HGF’ in one or more of the seven periods. With respect to the ‘employment HGFs’, 1
083 of the 1 908 firms (i.e., 56.8%) were only a HGF in one period and are qualified in
this paper as so-called ‘one-shot HGFs’. A one-shot HGF is consequently a firm that is
only an ‘employment HGF’ or ‘value-added HGF’ in one of the seven overlapping
periods. Practically the same percentage of ‘one-shot HGFs’ is found for the total group
of ‘value-added HGFs’. Hence, it appears that more than half of the firms that were
qualified as a HGF between 2000 and 2009 could only retain this status for one
three-year period5. Furthermore, it becomes clear that a very limited number of firms
were able to be a HGF in four or more periods. The firms that could be identified as a
HGF in four or more periods out of the seven overlapping periods are defined in this
paper as ‘persistent HGFs’. Being a ‘persistent HGF’ appears to be achieved by only
90 ‘employment HGFs’ (i.e., 4.72% of the total number of ‘employment HGFs’) and by
254 ‘value-added HGFs’ (i.e., 10.37%). Hence, it also appears that it is easier – or better
stated less difficult – for the ‘value-added HGFs’ to become a ‘persistent HGF’. Finally,
it is noteworthy to mention that only one ‘employment HGF’ and five ‘value-added
HGFs’ were present in each of the seven overlapping HGF-subsets.
Table 2 Number of firms that are qualified as a HGF for one or more three-year periods
Number of high-growth periods
Number of firms that are x time(s) a HGF
Employment HGFs Value-added HGFs
One-shot
HGFs
1 1,083 2,449
2 461 1,061
3 274 555
Persistent
HGFS
4 68 174
5 15 61
6 6 14
7 1 5
Total 1,908 4,319
Table 3 gives an overview of the number of firms that have been qualified as an
‘employment HGF’ or ‘value-added HGF’ in the first analysed period (i.e., the numbers
before the first arrow) and the number of firms of this initial group that could retain this
status in the next overlapping three-year period(s). So concretely, 134 firms of the
345 that were identified as an ‘employment HGF’ in the period 2000–2003 were still an
‘employment HGF’ in the period 2001–2004 (i.e., 38.84%). Fifty four of the initial
345 firms could retain the HGF-status for three consecutive overlapping periods (i.e.,
15.65%). The same tendency can be found for ‘value-added HGFs’ with an initial group
of 814 firms of which 326 firms were still a ‘value-added HGF’ in the period 2001–2004
(i.e., 40.05%). 140 of the initial 814 ‘value-added HGFs’ retained their HGF-status for
three consecutive periods (i.e., 16.65%).

308 Y. Dillen et al.
Table 3 Number of consecutive ‘employment HGFs’ and ‘value-added HGFs’
2000–2003 2001–2004 2002–2005 2003–2006 2004–2007 2005–2008 2006–2009
Employment HGFs 345 → 134 → 54 → 11 → 3 → 2 → 1
Value-added HGFs 814 → 326 → 140 → 53 → 25 → 12 → 5

Growth persistence and profile robustness of high-
g
rowth firms 309
Table 4 Number of ‘one-shot HGFs’ in the subsets of ‘employment HGFs’
2000–2003 2001–2004 2002–2005 2003–2006 2004–2007 2005–2008 2006–2009
Number of ‘employment HGFs’ 345 381 407 461 551 583 489
Number of ‘one-shot HGFs’ 176 115 103 111 143 177 258
51.01% 30.18% 25.31% 24.08% 25.95% 30.36% 52.76%

310 Y. Dillen et al.
Table 5 Number of ‘one-shot HGFs’ in the subsets of ‘value-added HGFs’
2000–2003 2001–2004 2002–2005 2003–2006 2004–2007 2005–2008 2006–2009
Number of ‘value-added HGFs’ 814 917 944 1,137 1,240 1,296 1,008
Number of ‘one-shot HGFs’ 377 265 264 314 364 417 448
46.31% 28.90% 27.97% 27.62% 29.35% 32.18% 44.44%

Growth persistence and profile robustness of high-
g
rowth firms 311
Table 4 and Table 5 represent the number of firms that are qualified as an ‘employment
HGF’ (Table 4) or a ‘value-added HGF’ (Table 5) in only one of the seven periods. For
instance, 177 of the 583 ‘employment HGFs’ for the period 2005–2008 could only be
qualified as a HGF in this period (i.e., 30.36%). The differences in the percentages of
‘one-shot HGFs’ are though remarkable. For the period 2006–2009 nearly 53% of the
‘employment HGFs’ were ‘one-shot HGFs’, whereas only 24.08% ‘one-shot HGFs’ were
reported for the subset of the period 2003–2006.
It appeared already that a large part of the HGFs were not able to retain their strong
growth rates for a longer period. In this context, it is interesting to consider the growth
rates that are realised by the firms that were a HGF in a certain period and who were no
longer a HGF one period later. Table 6 shows the number of this so-called ‘drop-out’
firms that have been qualified as an ‘employment HGF’ or ‘value-added HGF’ in the
period 2000–2003 and were no longer an ‘employment HGF’ or ‘value-added HGF’ in
the period 2001–2004. For the ‘employment HGFs’, it appeared that approximately a
quarter of the ‘drop-out’ firms could recapture the HGF-status in one of the subsequent
periods. With respect to the ‘value-added HGFs’, 183 of the 488 firms (i.e., 37,5%) that
were identified as a ‘value-added HGF’ in the period 2000–2003 and that were no longer
a ‘value-added HGF’ in the period 2001–2004 came back as a HGF in a later three-year
period. The percentage of firms that could return as a HGF after a drop-out is
consequently higher for the ‘value-added HGFs’ (i.e., 38% of the ‘value-added HGFs’
compared to 26% of the ‘employment HGFs’). Though it holds for both HGF-categories
that the clear majority of the ‘drop-outs’ could not regain their HGF-status in later years.
Nevertheless, it appears that the drop-out firms still realise – on average – growth rates
around 30% in the next overlapping three-year period. For instance, the 211 ‘employment
HGFs’ for the period 2000–2003 that were no longer an ‘employment HGF’ in the period
2001–2004 realised on average a growth rate of 28,2% in the period 2001–2004. Only 32
of these 211 firms had a negative growth rate in this subsequent three-year period.
Table 6 HGFs that return after a ‘drop-out’
Period
Employment HGFs Value-added HGFs
211 drop-out firms 488 drop-out firms
Number of firms that could return as a HGF after the drop-out
2002–2005 11 31
2003–2006 7 29
2004–2007 12 41
2005–2008 13 51
2006–2009 12 31
Total 55 / 211 (26.07%) 183 / 488 (37.5%)
5 Robustness of the profile characteristics of the different HGF-subsets
As was stated in the introduction, this paper not only tries to learn more about the
persistence of high firm growth but also wants to gain insight into the so-called
‘robustness’ of the profile characteristics of the different HGF-subsets. In recent years,
several studies have been conducted in which the profile features of HGFs are compared

312 Y. Dillen et al.
to the profile features of non-HGFs (e.g., Henrekson and Johansson, 2010; Davidsson and
Delmar, 2006). Given the existence of these studies, we will not make a comparison
between the profile characteristics of HGFs and non-HGFs as in these studies, it was
already found that HGFs are clearly younger and smaller than non-HGFs and that HGFs
are overrepresented in knowledge intensive service industries. In our study, however, we
will adopt a different approach in which we aim to make a comparison between the
profile characteristics of the different HGF-subsets. This kind of analysis, which has not
been conducted until now, can test if the profile features of HGFs are consistentfor the
different HGF-subsets and if these features are stable over time. The following section
will analyse the average firm age, the average firm size, the average amount of total
assets, the average solvency and the geographical and sectoral distribution of the different
HGF-subsets. With respect to outliers, observations that deviated from the average by
more than five times the standard deviation were omitted from the calculations.
5.1 Firm age
Table 7 gives an overview of the average age of the firms in the different HGF-subsets.
No significant differences were found between the average age of the firms in the seven
subsets of ‘employment HGFs’ as the average ages ranged only between 16.1 years (i.e.,
the 2000–2003 subset) and 17.8 years (i.e., the 2005–2008 and 2006–2009 subset). With
respect to the ‘value-added HGFs’ – who are on average older than the ‘employment
HGFs’ –, it appears that the average age of the firms in the two first subsets is
significantly lower than the averages of the other subset.6 As from the third analysed
period, however, the average age stabilises around 20 years. Hence, one can conclude
that the average firm age of the different subsets is quite stable over time.
Table 7 Average age of the firms in the HGF-subsets
Period Average firm age (year*)
Employment HGFs Value-added HGFs
2000–2003 16.1 17.5
2001–2004 17.5 18.5
2002–2005 17.1 20.0
2003–2006 16.6 20.4
2004–2007 17.1 20.8
2005–2008 17.8 20.2
2006–2009 17.8 19.5
Notes: *Firm age calculated on the basis of the last year of each three-year period;
p-values were calculated by using the Student’s t-test at a level of significance of
5%.
5.2 Firm size
Table 8 shows that also the average size of the different HGF-subsets – measured by the
number of employees – remains quite stable over time as no significant differences
between the averages were encountered. The average size of the ‘employment
HGF’-subsets varied between 82.5 and 102.5 whereas the average size of the

Growth persistence and profile robustness of high-
g
rowth firms 313
‘value-added HGF’-subsets ranged from 64.9 to 73.8. Therefore, ‘value-added HGFs’
appear to be larger than ‘employment HGFs’.
Table 8 Average size of the firms in the HGF-subsets
Period Average firm size (number of employees)
Employment HGFs Value-added HGFs
2000–2003 83.4 69.3
2001–2004 85.0 73.8
2002–2005 99.3 64.9
2003–2006 97.4 69.9
2004–2007 101.3 65.1
2005–2008 102.5 71.1
2006–2009 82.5 71.6
5.3 Total assets
Table 9 shows that the ‘employment HGF’-subsets for the periods 2004–2007 and
2005–2008 have a significantly higher average total amount of assets than most other
subsets.7 The same holds for the ‘value-added HGF’-subsets for the periods 2001–2004
and 2002–2005. The subset of ‘value-added HGFs’ for the period 2006–2009 has,
however, a significantly lower average.8 Apart from these ‘outlying’ averages, the
average total assets of the other subsets varies roughly around 20 million EUR, both for
‘employment HGFs’ and ‘value-added HGFs’. Consequently, even though there are
respectively two and three aberrant results, it still appears that the fluctuations remain
fairly limited for the other HGF-subsets.
Table 9 Average total assets of the firms in the HGF-subsets
Period Total assets (average; inEUR)
Employment HGFs Value-added HGFs
2000–2003 16,257,270 19,989,299
2001–2004 19,170,046 33,063,890
2002–2005 18,824,550 27,627,605
2003–2006 23,078,790 21,952,835
2004–2007 26,063,515 20,128,904
2005–2008 32,703,325 17,077,250
2006–2009 20,237,284 12,622,710
5.4 Solvency
The solvency of the firms in the ‘employment HGF’-subsets remain quite stable over the
analysed decade as only one significant difference between the seven averages was
encountered (cf., Table 10).9 The average debt ratios vary from 71.0% to 75.2%. With
respect to the ‘value-added HGFs’, the average debt ratios of the first subsets are
significantly higher than the averages of the latest subsets. The average solvency of the

314 Y. Dillen et al.
2003–2006 subset (i.e., 77.1%) is also significantly higher than all other averages of the
‘value-added HGF’-subsets. No clear differences between the two HGF-categories can
be encountered.10
Table 10 Average solvency of the firms in the HGF-subsets
Period Solvency (liabilities/total assets)
Employment HGFs Value-added HGFs
2000–2003 75.2% 72.7%
2001–2004 73.1% 73.0%
2002–2005 74.4% 71.3%
2003–2006 72.6% 77.1%
2004–2007 71.7% 68.9%
2005–2008 73.6% 69.7%
2006–2009 71.0% 69.2%
5.5 Sectoral distribution
A similar picture can be found for the sectoral distribution over the seven periods. Tables
B and C – who are included in Appendix – show the distribution of the HGFs over the
15 included NACE-sections. The differences in the relative shares of each section over
the seven periods remain fairly limited. The strongest fluctuation can be found in NACE-
Section G (i.e., wholesale and retail trade) for the ‘employment HGF’-subsets with a
rather small difference of 6.39 percentage points between the maximum and minimum
share.
6 ‘One-shot HGFs’ versus ‘persistent HGFs’
Table 11 compares some profile characteristics of the so-called ‘one-shot HGFs’ and
‘persistent HGFs’. It has to be noted, though, that this comparison is made between
subsets of which the size is very diverse. The group of 90 persistent ‘employment HGFs’
is compared to a group of 1,083 one-shot ‘employment HGFs’, whereas the 254 persistent
‘value-added HGFs’ are compared to 2,449 one-shot ‘value-added HGFs’. It appears that
the persistent HGFs – with an average age of approximately 25 years – are clearly older
than the one-shot HGFs. It could be expected that the persistent HGFs are larger than the
one-shot HGFs given their longer period of high-growth. The magnitude of the difference
in size is however remarkable. The 90 firms that were qualified as persistent ‘employment
HGFs’ had an average workforce of 233.2 employees in 2009 whereas the one-shot
‘employment HGFs’ had an average number of 68.3 employees in the last year of their
high-growth period. A similar picture can be found for the ‘value-added HGFs’ where
the persistent HGFs had an average number of 161.7 employees in 2009 compared to an
average workforce of 55,2 for the one-shot HGFs. Also with respect to the total amount
of assets, large differences can be found between both categories of HGFs. It appears as
well that the differences between persistent and one-shot HGFs are greater for
‘employment HGFs’ than for ‘value-added HGFs’.

Growth persistence and profile robustness of high-
g
rowth firms 315
Table 11 One-shot HGFs versus persistent HGFs
One-shot ‘employment HGFs’
(1,083 firms)
Persistent ‘employment HGFs’
(90 firms)
Average firm age 15.7 years 24.6 years
Average firm size 68.3 employees 233.2 employees
Average total assets 13,437,600 EUR 101,865,798 EUR
Average solvency 74.61% 72.32%
One-shot ‘value-added HGFs’
(2,449 firms)
Persistent ‘value-added HGFs’
(254 firms)
Average firm age 19.8 years 25.0 years
Average firm size 55.2 employees 161.7 employees
Average total assets 15,342,165 EUR 54,988,313 EUR
Average solvency 72.32% 71.30%
7 Conclusions and implications
7.1 Persistence of high business growth
It appeared that more than half of the firms that were identified as a HGF in the period
2000–2009 could only maintain this status for one of the seven analysed three-year
periods, and this both for ‘employment HGFs’ and ‘value-added HGFs’. Furthermore, it
became clear that only a very limited number of firms were able to be qualified as a HGF
in at least four of the seven periods. There is – in other words – a large difference
between the number of ‘one-shot HGFs’ and the number of ‘persistent HGFs’. This
phenomenon of temporary high-growth can be linked to the five concepts and theories
that were mentioned earlier on.
1 With respect to the MES that a firm may want to reach in order to survive in their
sector, temporary high-growth can be explained by the fact that these firms will try
to attain this size as fast as possible and once the size is reached their growth
intentions will fade away.
2 The life cycle theory of the firm can explain temporary high-growth as a firm grows,
according to this theory, in sequential steps from birth to maturity. A relatively short
period of high-growth may occur in one of the earlier phases as growth is considered
as a natural process where every company will go through.
3 The difference that is made in the resource-based view between a competitive
advantage and a sustained competitive advantage can also give an explanation to the
encountered small number of ‘persistent HGFs’ as possessing a sustained
competitive advantage is considered to be ‘reserved’ for a very limited number of
firms, namely the firms that own resources that are apart from valuable and rare also
imperfectly imitable and not easily substitutable.

316 Y. Dillen et al.
4 The low number of ‘persistent HGFs’ is in line with Gibrat’s Law as this law states
that firm growth rates are independent random variables. Following this logic, firms
that have grown fast in one period will not grow faster than other firms in the next
period.
5 In the growth theory of Edith Penrose, it is stated that the availability and quality of
the management can be considered as a bottleneck for the growth rates that a firm
can achieve in the long run. The difficulties that HGFs are having to sustain high
growth rates can consequently be linked to this theory: the pool of managerial
resources is limited which impedes HGFs from attracting experienced managers that
are needed to overcome certain growing pains and to keep the firm on the
high-growth track.
The life cycle theory of the firm, the MES, (sustained) competitive advantage, Gibrat’s
Law and the growth theory of Penrose all together, give all rise to the conclusion that
high firm growth has an episodic nature. The striking difference between the encountered
amount of ‘one-shot HGFs’ and ‘persistent HGFs’ clearly supports this statement.
‘Persistent HGFs’ appeared to be clearly older and larger than the ‘one-shot HGFs’
which is not surprisingly given their long period of strong growth rates. It was also shown
that only 40% to 50% of the firms that were a HGF in period x were still a HGF in period
x + 1. If three consecutive periods are considered (i.e., the periods x; x + 1 and x + 2),
this percentage reduces even to approximately 20%. Moreover, the majority of the
so-called ‘drop-out’ firms could not return as a HGF in a later period. Realising high
growth rates for more than three consecutive years or returning as a HGF after a drop-out
are consequently difficult tasks for most of the Flemish HGFs.
As mentioned, the above findings confirm the earlier made statement that only a very
limited number of firms are able to maintain high growth rates for a long period of time.
This may have implications from a policy, academic and managerial perspective. In
recent years, many regions have launched governmental initiatives that aim to stimulate
the HGFs that are present in their region. In establishing these initiatives, it has to be kept
in mind that the majority of the firms that can be qualified as a HGF at a certain moment
in time will only be a HGF for a relatively short period. Supportive measures towards
these firms should be seen in the light of the short-lived character of high business
growth. It can be suggested that the initiatives should primarily be directed to firms that
have the potential to become a HGF instead of the firms that have already – according to
the OECD definition – attained the status of a HGF. The latter would carry the risk that
the support only comes about at a moment when a firm’s high growth period has already
passed. From an academic perspective, the presence in the HGF-subset of ‘one-shot
HGFs’ and ‘persistent HGFs’ may have some implications. Given the profile differences
that exist between these two types of HGFs, it could be useful to separate between both
categories when the aim is to analyse the determinants of the high growth rates.
Moreover, when conducting case studies in HGFs, focusing on the ‘persistent HGFs’
could give new insights as these type of firms have proven to be capable of sustaining
high growth for a long period. Given their strong track record of high growth, these firms
could be more suitable for case analyses that have the goal to study the determinants of
high business growth. From a managerial point of view, HGFs are often considered to be
a sort of best practice. Managers of non-HGFs could look upon these firms as an example
for their own business. However, these managers have to realise that a large part of the
HGFs are ‘one-shot HGFs’ whose strong growth rates could possibly be related to an

Growth persistence and profile robustness of high-
g
rowth firms 317
accidental or unforeseen growth boost in a certain year. Again, the group of ‘persistent
HGFs’ may serve better as a best practice for these managers.
7.2 Profile robustness of HGFs
Different profile characteristics were analysed over time for the 14 HGF-subsets. As
presented earlier, a large part of the firms in these subsets could be qualified as ‘one-shot
HGFs’. The subsets consist consequently of many different firms. Hereby, it could be
expected that the profile characteristics are subject to the rapidly changing composition of
the subsets and that there would consequently be large differences between the averages
of the different subsets. It appeared, however, that very few significant differences could
be encountered with respect to the average firm age, firm size, total assets and solvency
of the HGF-subsets. Also the sectoral distribution of the HGFs over the seven
overlapping periods remained fairly stable. So, the profile features of the firms in the
HGF-subsets remain relatively constant over time, notwithstanding the high in- and
out-flow of firms in these subsets. This message could be a reassurance for the policy
makers and government institutions who have launched initiatives towards HGFs. These
initiatives may benefit from the finding that the subsets consist of more or less the same
type of firms as supportive measures should then not ever be adjusted to the changing
outlook of the HGF-subset.
8 Limitations and future research
As mentioned before, the majority of the Flemish firms do not have the obligation to
publish turnover figures. Because of this limited availability, value-added was chosen as
a growth measure instead of turnover. This could possibly hamper the comparability of
the findings as the applied OECD definition prefers normally employment and turnover
as growth indicators.
The distinction that was made between ‘persistent HGFs’ and ‘one-shot HGFs’ could
offer new opportunities for future research. What are the exact factors that prevent the
‘one-shot HGFs’ from retaining their high growth rates for a longer time period? Is it
merely a matter of a declining demand for their service or product, or are there other
internal and external factors that play a role in explaining the phenomenon of temporary
high growth. Learning more about this topic is important as it might give new ‘one-shot
HGFs’ and potential HGFs new insights into how the growth pains that are associated
with high growth can be overcome.
Moreover, it could be instructive to analyse the sectoral distribution of the two HGF-
types. Are there any industries that are characterised by a relatively higher amount of
‘persistent’ or ‘one-shot HGFs’? And what are the possible explanations for this
presence? Second, the determinants of the high growth rates that are realised by the
‘persistent HGFs’ on the one hand and the ‘one-shot HGFs’ on the other hand can be
analysed in future research as this paper only considered some profile characteristics of
this two types of HGFs. Which factors determine the high growth of these firms? Why
are many HGFs not able to retain their high growth for more than one period? And which
factors can explain the persistent high growth of the ‘persistent HGFs’? Also with respect
to the ‘drop-out’ firms, more research can be performed in the future. Why are certain
firms able to return as a HGF after such a drop-out and others not?

318 Y. Dillen et al.
It appears that many questions still need to be addressed when it comes to the issue of
the persistence of high business growth. In any case, it has become clear in this article
that future research on HGFs should ideally always take the distinction between
‘persistent’ and ‘one-shot HGFs’ into account as they represent clearly two different
types of HGFs.
References
Acs, Z., Parsons, W. and Spencer, T. (2008) ‘High impact firms: Gazelles revisited’, Corporate
Research Board. US Small Business Administration, June.
Almus, M. (2002) ‘What characterizes a fast-growing firm?’, Applied Economics, Vol. 34, No. 2,
pp.1497–1508.
Anyadike-Danes, M., Bonner, K., Hart, M. and Mason, C. (2009) ‘Measuring business growth:
high growth firms and their contribution to employment in the UK’, October, Research Report
NESTA.
Barney, J. (1991) ‘Firm resources and sustained competitive advantage’, Journal of Management,
Vol. 17, No. 1, pp.99–120.
Barringer, B., Jones, F. and Neubaum, D. (2005) ‘A quantitative content analysis of the
characteristics of rapid-growth firms and their founders’, Journal of Business Venturing,
Vol. 20, No. 5, pp.663–687.
BERR (2008) High Growth Firms in the UK: Lessons from an Analysis of Comparative UK
Performance, November, BERR Economics Paper no. 3, London, UK.
Churchill, N. and Lewis, V. (1983) ‘The five stages of small business growth’, Harvard Business
Review, May–June, Vol. 61, No. 3, pp.30–50.
Coad, A. (2009) The Growth of Firms: A Survey of Theories and Empirical Evidence, Edward
Elgar, Cheltenham.
Davidsson, P. and Delmar, F. (2006) ‘High-growth firms and their contribution to employment: the
case of Sweden’, in Davidsson, P., Delmar, F. and Wiklund, J. (Eds.): Entrepreneurship and
the Growth of Firm, pp.156–178, Edward Elgar, Cheltenham, UK and Northampton, MA.
Davidsson, P. and Henrekson, M. (2002) ‘Determinants of the prevalence of start-ups and
high-growth firms’, Small Business Economics, Vol. 19, No. 2, pp.81–104.
Delmar, F., Davidsson, P. and Gartner, W. (2003) ‘Arriving at the high-growth firm’, Journal of
Business Venturing, Vol. 18, No. 2, pp.189–216.
European Commission (2011) Policies in Support of High-Growth Innovative SMEs, June, INNO
Grips Policy Brief by Empirica Communication and Technology Research, principal author:
Stefan Lilischkis, Bonn, Germany.
Eurostat – OECD (2007) Manual on Business Demography Statistics, December, European
Communities/OECD, Paris, France.
Garnsey, E., Stam, E. and Hefferman, P. (2006) ‘New firm growth: exploring processes and paths’,
Industry & Innovation, Vol. 13, No. 1, pp.1–20.
Gibrat, R. (1931) Les inéqalitéséconomique, Recueil Sirey, Paris.
Greiner, L.E. (1972) ‘Evolution and revolution as organizations grow’, Harvard Business Review,
May–June, Vol. 50, No. 4, pp.37–46.
Harms, R., Reschke, C.H., Kraus, S. and Fink, M. (2010) ‘Antecedents to innovation and growth:
analyzing the impact of entrepreneurial orientation and management by objectives’,
International Journal of Technology Management, Vol. 52, Nos. 1/2, pp.135–152.
Henrekson, M. and Johansson, D. (2010) ‘Gazelles as job creators: a survey and interpretation of
the evidence’, Small Business Economics, Vol. 35, No. 2, pp.227–244.
Kanzanjian, R. (1988) ‘Relation of dominant problems to stages of growth in technology-based
new ventures’, Academy of Management Journal, Vol. 31, No. 2, pp.257–279.

Growth persistence and profile robustness of high-
g
rowth firms 319
Kor, Y. and Mahoney, J.T. (2004) ‘Edith Penrose’s (1959) contributions to the resource-based view
of strategic management’, Journal of Management Studies, Vol. 41, No. 1, pp.183–191.
Lipczynski, J., Wilson, J. and Goddard, J. (2005) Industrial Organization: Competition, Strategy
and Policy, Pearson Education, Essex.
Mason, C. and Brown, R. (2010) ‘High growth firms in Scotland’, October, Research report
Scottish Enterprise – University of Strathclyde Business School, Glasgow, UK.
Mason, G., Bishop, K. and Robinson, C. (2009) ‘Business growth and innovation: the wider impact
of rapidly growing firms in UK City-Regions’, October, NESTA Research Report BGI/36,
London, UK.
Parker, S., Storey, D. and Van Witteloostuijn, A. (2010) ‘What happens to gazelles? The
importance of dynamic management strategy’, Small Business Economics, Vol. 35, No. 2,
pp.203–222.
Penrose, E. (1959) The Theory of the Growth of the Firm, Oxford University Press Inc., New York.
Phelps, R., Adams, R. and Bessant, J. (2007) ‘Life cycle of growing organizations: a review with
implications for knowledge and learning’, International Journal of Management Reviews,
Vol. 9, No. 1, pp.1–30.
Sutton, J. (1997) ‘Gibrat’s legacy’, Journal of Economics Literature, Vol. 35, No. 1, pp.40–59.
Teruel, M. and de Wit, G. (2011) ‘Determinants of high-growth firms: why have some countries
more high-growth firms than others?’, January, EIM Research Report, Zoetermeer,
The Netherlands.
Notes
1 A resource is valuable if it has the potential to exploit opportunities and/or to neutralize threats
(Barney, 1991).
2 Barney (1991) points out that a sustained competitive advantage does not imply that it will
‘last forever’. It only suggests that other firms will not compete it away through duplication
efforts. Empirically, however, sustained competitive advantages last on average a long period
of time (Barney, 1991).
3 When a business gets below the minimum threshold of ten employees in year x-1 of x-2 it can
still be qualified as an HGF if the total growth between year x-3 and year x exceeds 72.8%.
4 A list with the excluded and included NACE-sections can be found in Appendix
(cf., Table A1).
5 Tables 4 and 5 give a more detailed picture of the presence of ‘one-shot HGFs’
6 ‘Value-added HGFs’: significant differences between the average age of the
2000–2003-subset and the average firm age of the subsets of the periods 2002–2005,
2003–2006, 2004–2007, 2005–2008 and 2006–2009. Also significant differences between the
average age of the 2001–2004 subset and the average age of the subsets of the periods
2002–2005, 2003–2006, 2004–2007 and 2005–2008.
7 ‘Employment HGFs’: two subsets (i.e., the subsets of the periods 2004–2007 and 2005–2008)
have a clearly higher average amount of assets than the other subsets.
8 ‘Value-added HGFs’: two subsets (i.e., the subsets of the periods 2004–2007 and 2005–2008)
have a clearly higher average amount of assets than the other subsets whereas the subset of the
period 2006–2009 has a clearly lower average.
9 ‘Employment HGFs’: only one significant difference could be found (i.e., between the average
solvency of the firms in the 2000–2003 subset and the firms in the 2006–2009 subset).
10 ‘Value-added HGFs’: the average solvency of the firms in the 2003–2006 subset is
significantly higher than all other averages. The average solvencies of the three latest subsets
(i.e., 2004–2007; 2005–2008 and 2006–2009) are significantly lower than the averages of the
two first HGF-subsets (i.e., 2000–2003 and 2001–2004).

320 Y. Dillen et al.
Appendix 1
Table A1 NACE-BEL 2008 sections
Excluded sections NACE-BEL 2008
Section K Financial and insurance activities
Section O Public administration and defence; compulsatory social security
Section P Education
Section Q Human health and social work activities
Section T Activities of households as employers
Section U Activities of extraterritorial organisations and bodies
Included sections NACE-BEL 2008
Section A Agriculture, forestry and fishing
Section B Mining and quarrying
Section C Manufacturing
Section D Electricity, gas, steam and air conditioning supply
Section E Water supply, sewerage, waste management and remediation activities
Section F Construction
Section G Wholesale and retail trade; repair of motor vehicle and motorcycles
Section H Transporting and storage
Section I Accommodation and food service activities
Section J Information and communication
Section L Real estate activities
Section M Professional, scientific and technical activities
Section N Administrative and support service activities
Section R Arts, entertainment and recreation
Section S Other service activities

Growth persistence and profile robustness of high-
g
rowth firms 321
Appendix 2
Table A2 Sectoral distribution of the ‘employment HGFs’ over the NACE-sections for the seven
overlapping periods
NACE-section 2000–2003 2001–2004 2002–2005 2003–2006 2004–2007 2005–2008 2006–2009
Difference between highest
and lowest percentage
A 1.21 0.55 0.25* 0.44 0.93 1.37** 1.02 1.12
B 0.60** 0.00* 0.00 0.00 0.00 0.00 0.00 0.60
C 15.71 17.53 16.54 17.52 19.22** 16.3 15.54* 3.68
D 0.00* 0.27 0.5** 0.44 0.37 0.00 0.20 0.50
E 0.60 1.92** 1.50 1.33 0.75 0.34* 0.41 1.58
F 10.57* 12.60 12.03 11.31 14.74** 13.38 12.88 4.17
G 25.98** 23.84 23.06 23.06 19.59* 21.27 21.68 6.39
H 16.31 16.16 16.54** 12.86 16.42 14.07 12.68* 3.86
I 3.32 3.84 3.26 4.21** 3.17 4.12 3.07* 1.14
J 5.14* 5.48 7.52 7.54 6.16 8.06 9.20** 4.06
L 0.00* 0.00 0.00 0.22 0.37 0.69** 0.61 0.69
M 8.76 7.12 7.02* 10.42 8.58 10.29 10.84** 3.82
N 8.46 8.49 9.77 7.32* 8.77 8.4 11.04** 3.72
R 2.11** 1.37 1.25 1.11 0.37 0.34 0.00* 2.11
S 1.21 0.82 0.75 2.22** 0.56* 1.37 0.82 1.66
Notes: *Minimu
m
-percenta
g
e for the section; **maximu
m
-percenta
g
e for the section

322 Y. Dillen et al.
Table A3 Sectoral distribution of the ‘value-added HGFs’ over the NACE-sections for the
seven overlapping periods
NACE-section 2000–2003 2001–2004 2002–2005 2003–2006 2004–2007 2005–2008 2006–2009
Difference between highest
and lowest percentage
A 0.77 1.26 0.66 0.63* 1.57 1.17 1.68** 1.05
B 0.13** 0* 0 0 0 0 0 0.13
C 23.85** 20.66 22.55 21.62 22.52 20.75 19.9* 3.95
D 0.26** 0* 0 0.09 0 0.16 0.1 0.26
E 0.77 1.71 2.09** 1.8 1.16 0.7* 0.52 1.39
F 9.44* 12.21 11 13.78 15.56 15.82** 13.61 6.38
G 29.97 32.65** 32.23 30.45 29.14 28.03 27.85* 4.8
H 11.86** 10.62 9.02 8.47* 9.77 11.74 9.53 3.39
I 3.19** 2.74 2.42 2.16 1.32* 2.11 2.41 1.87
J 5.74 4.57* 5.5 6.13 4.8 5.17 6.28** 1.71
L 0.77** 0.23 0.55 0.18* 0.25 0.23 0.52 0.59
M 6.25* 6.51 7.48 6.85 6.79 6.97 9.21** 2.96
N 5.1 5.02 4.84* 6.04 5.38 6.11 7.02** 2.18
R 0.89 0.8 0.88 0.9** 0.75 0.39* 0.63 0.51
S 1.02 1.03** 0.77 0.9 0.99 0.63* 0.73 0.4
Notes: *Minimu
m
-
p
ercentage for the section; **maximu
m
-
p
ercentage for the section

Growth persistence and profile robustness of high-
g
rowth firms 323
Appendix 3
Table A4 Firm age (employment HGFs)
Period 2000–2003 2001–2004 2002–2005 2003–2006 2004–2007 2005–2008 2006–2009
2000–2003 / p = 0.105 p = 0.233 p = 0.595 p = 0.207 p = 0.081 p = 0,084
2001–2004 / p = 0.677 p = 0.266 p = 0.658 p = 0.742 p = 0.737
2002–2005 / p = 0.494 p = 0.998 p = 0.892 p = 0.887
2003–2006 / p = 0.464 p = 0.323 p = 0.309
2004–2007 / p = 0.337 p = 0.329
2005–2008 / p = 0.997
2006–2009 /
Note: P-values calculated by using the student’s t-test at a level of significance of 5%.

324 Y. Dillen et al.
Table A5 Firm age (value-added HGFs)
Period 2000–2003 2001–2004 2002–2005 2003–2006 2004–2007 2005–2008 2006–2009
2000–2003 / p = 0.088 p = 0.000* p = 0.000* p = 0.000* p = 0.000* p = 0,009*
2001–2004 / p = 0.016* p = 0.003* p = 0.000* p = 0.006* p = 0.062
2002–2005 / p = 0.622 p = 0.235 p = 0.924 p = 0.613
2003–2006 / p = 0.474 p = 0.834 p = 0.184
2004–2007 / p = 0.322 p = 0.021
2005–2008 / p = 0.212
2006–2009 /
Notes: P-values calculated by using the student’s t-test at a level of significance of 5%.
*Significant difference between the averages of the two involved periods.

Growth persistence and profile robustness of high-
g
rowth firms 325
Table A6 Firm size (employment HGFs)
Period 2000–2003 2001–2004 2002–2005 2003–2006 2004–2007 2005–2008 2006–2009
2000–2003 / p = 0.871 p = 0.271 p = 0.303 p = 0.172 p = 0.154 p = 0.911
2001–2004 / p = 0.346 p = 0.386 p = 0.236 p = 0.212 p = 0.769
2002–2005 / p = 0.911 p = 0.902 p = 0.842 p = 0.183
2003–2006 / p = 0.797 p = 0.736 p = 0.209
2004–2007 / p = 0.932 p = 0.101
2005–2008 / p = 0.088
2006–2009 /
Note: P-values calculated by using the student’s t-test at a level of significance of 5%.

326 Y. Dillen et al.
Table A7 Firm size (value-added HGFs)
Period 2000–2003 2001–2004 2002–2005 2003–2006 2004–2007 2005–2008 2006–2009
2000–2003 / p = 0.600 p = 0.599 p = 0.943 p = 0.557 p = 0.824 p = 0.738
2001–2004 / p = 0.311 p = 0.665 p = 0.248 p = 0.794 p = 0.387
2002–2005 / p = 0.570 p = 0.976 p = 0.440 p = 0.366
2003–2006 / p = 0.527 p = 0.888 p = 0.821
2004–2007 / p = 0.386 p = 0.304
2005–2008 / p = 0.926
2006–2009 /
Note: P-values calculated by using the student’s t-test at a level of significance of 5%.

Growth persistence and profile robustness of high-
g
rowth firms 327
Table A8 Total assets (employment HGFs)
Period 2000–2003 2001–2004 2002–2005 2003–2006 2004–2007 2005–2008 2006–2009
2000–2003 / p = 0.279 p = 0.282 p = 0.038* p = 0.012* p = 0.002* p = 0.128
2001–2004 / p = 0.838 p = 0.215 p = 0.035* p = 0.009* p = 0.642
2002–2005 / p = 0.133 p = 0.028* p = 0.007* p = 0.577
2003–2006 / p = 0.258 p = 0.041* p = 0.283
2004–2007 / p = 0.102 p = 0.127
2005–2008 / p = 0.010*
2006–2009 /
Notes: P-values calculated by using the student’s t-test at a level of significance of 5%.
*Significant difference between the averages of the two involved periods.

328 Y. Dillen et al.
Table A9 Total assets (value-added HGFs)
Period 2000–2003 2001–2004 2002–2005 2003–2006 2004–2007 2005–2008 2006–2009
2000–2003 / p = 0.044* p = 0.077 p = 0.528 p = 0.959 p = 0.263 p = 0.014*
2001–2004 / p = 0.411 p = 0.047* p = 0.013* p = 0.002* p = 0.000*
2002–2005 / p = 0.136 p = 0.030* p = 0.002* p = 0.000*
2003–2006 / p = 0.461 p = 0.040* p = 0.000*
2004–2007 / p = 0.137 p = 0.002*
2005–2008 / p = 0.048*
2006–2009 /
Notes: P-values calculated by using the student’s t-test at a level of significance of 5%.
*Significant difference between the averages of the two involved periods.

Growth persistence and profile robustness of high-
g
rowth firms 329
Table A10 Solvency (employment HGFs)
Period 2000–2003 2001–2004 2002–2005 2003–2006 2004–2007 2005–2008 2006–2009
2000–2003 / p = 0.243 p = 0.697 p = 0.191 p = 0.055 p = 0.404 p = 0.039*
2001–2004 / p = 0.448 p = 0.810 p = 0.424 p = 0.794 p = 0.289
2002–2005 / p = 0.352 p = 0.130 p = 0.651 p = 0.088
2003–2006 / p = 0.607 p = 0.626 p = 0.421
2004–2007 / p = 0.293 p = 0.716
2005–2008 / p = 0.188
2006–2009 /
Notes: P-values calculated by using the student’s t-test at a level of significance of 5%.
*Significant difference between the averages of the two involved periods.

330 Y. Dillen et al.
Table A11 Solvency (value-added HGFs)
Period 2000–2003 2001–2004 2002–2005 2003–2006 2004–2007 2005–2008 2006–2009
2000–2003 / p = 0.866 p = 0.435 p = 0.011* p = 0.013* p = 0.035* p = 0,027*
2001–2004 / p = 0.361 p = 0.021* p = 0.010* p = 0.028* p = 0.024*
2002–2005 / p = 0.007* p = 0.124 p = 0.280 p = 0.202
2003–2006 / p = 0.000* p = 0.000* p = 0.000*
2004–2007 / p = 0.509 p = 0.821
2005–2008 / p = 0.694
2006–2009 /
Notes: P-values calculated by using the student’s t-test at a level of significance of 5%.
*Significant difference between the averages of the two involved periods.