CESIS Electronic Working Paper Series

Paper No. 270

Born global firms in knowledge intensive business services

(KIBS) – what do we know of their performance?

Torbjörn Halldin

March 2010

The Royal Institute of technology Centre of Excellence for Science and Innovation Studies (CESIS)

http://www.cesis.se

2

Born global firms in knowledge intensive business services

(KIBS) – what do we know of their performance?

Torbjörn Halldin

Division of Economics, the Royal Institute of Technology

Abstract

This paper studies the performance of KIBS firms that are aimed for global markets

from inception. Despite the increasing importance of KIBS, no previous study has

investigated born global firms in this sector of the economy. Three definitions are used

to categorize firms as born global. Both OLS and a nearest neighbor matching approach

are implemented to find evidence of higher growth in employment and sales per

employee five years after firm foundation. Finally, the findings are robust to a finer

array of born global definitions and time horizons of firm performance.

Keywords: Born global firms, firm performance, knowledge intensive business sector

JEL classification: L25, L26, M13, M21, F14

3

1 Introduction

In recent years, many economies have experienced an enormous increase in the amount of various

specialized business services. These services provide important inputs to firms in all sectors of the

economy. Many of these business services are knowledge intensive, including for instance IT- and

management consultancies, R&D and engineering activities and legal and accountancy services. The large

knowledge component of such sectors in terms of a highly educated labor force and high R&D

expenditures makes their role in technology transfer and knowledge creation crucial to the growth of the

economy (Miles, 2005). Firms characterized as knowledge intensive business services (KIBS) have

become a focus of increasing attention both among scholars and policy makers. The contribution to the

economy of KIBS lies not only in its share of GDP. Rather, the new knowledge created from interactions

with clients and the application of such knowledge in production, particularly within the manufacturing

industry, have possibly an even larger impact on modern economies (Tomlinson, 2000).

Globalization has played a major role on the business climate. Many national borders are fading away in

order to facilitate trade and exchange of knowledge. Having markets opening up for international

competition makes it imperative for regions to provide the appropriate preconditions for new firms to

survive and become competitive globally. For small open countries like Sweden, with the ambition to

remain and develop a large knowledge content of their economies, it is especially important to enhance the

possibilities for KIBS to grow and become productive on the global arena. Due to trends in globalization,

the internationalization pattern of many new firms is changing. In previous studies, the internationalization

process of firms has been seen as a linear and time-prolonged event based on the stage theories set up by

Johanson and Vahlne (1977; 1990; 2006) and Vernon (1966; 1971; 1979). Such a longitudinal

perspective has been applied on how KIBS gradually acquire key competencies necessary for expansion

and success in foreign markets (Roberts, 2000). However, the last two decades have witnessed about

several examples of firms that from inception perceive the world as one market (Chetty and Campbell-

Hunt, 2004). These firms, typically knowledge intensive, are globally oriented from start and could be

defined as firms that ‘‘from inception, seek to derive significant competitive advantage from the use of

resources and the sales of outputs in multiple countries’’ (Oviatt and McDougall, 1994:p. 49).

The empirical literature on born global firms has focused on the manufacturing sector. Most studies are

qualitative by nature and none addresses the KIBS separately, see Rialp-Criado et al. (2005) for a review

on born global firms. Much of the services KIBS firms provide are aimed for the manufacturing industry.

The past decades have shown a rapid expansion of international activities among manufacturing firms.

4

Whether there has been a similar growth of KIBS firms on an international basis is less studied. A separate

treatment of KIBS firms is further motivated by the sector’s importance in generating new firms and

employment. In fact, KIBS firms are even creating more new firms and employment than the

manufacturing industry as a whole.1 Hence, in studies of entrepreneurship, KIBS has an important role to

play. The present paper will try to empirically investigate the performance of born global KIBS firms and

how these firms compare to other newly founded firms within the same knowledge intensive industry.

The paper is organized as follows. Section 2 reviews the literature on KIBS as a distinct research topic.

Briefly, it also goes through the original stage theories of firm internationalization and the early literature

on born global firms. It also delineates some factors facilitating the creation of born global firms. Finally it

develops the specific research hypotheses addressed in the paper. In section 3 the data selection is

described whereas section 4 shows descriptive statistics. Section 5 goes through the methodology used in

the paper and in section 6 the empirical results of the paper are stated. Finally, the findings are

summarized in section 7 with some concluding remarks.

2 Literature review and hypotheses

KIBS

There are many definitions of KIBS used in the literature.2 Miles et al. (1995) defined KIBS as “services

that involved economic activities which are intended to result in the creation, accumulation or

dissemination of knowledge”. They characterize KIBS as firms that

rely heavily upon professional knowledge

either are themselves primary sources of information and knowledge (reports, training

consultancy etc.)

or use their knowledge to produce intermediary services for their clients' production processes (e.g.

communication and computer services)

are of competitive importance and supplied primarily to business.

1 See the descriptive statistics in chapter 4 for a more detailed description of the importance of Swedish KIBS for

entrepreneurship. 2 A recent review on the role of KIBS, primarily in innovation, and various other definitions is Muller and Doloreux (2009).

5

Den Hertog (2000) narrows down the definition of KIBS as “private companies or organizations that rely

heavily on professional knowledge, i.e., knowledge or expertise related to a specific (technical) discipline

or (technical) functional-domain to supply intermediate products and services that are knowledge based”.

These definitions boil down to the acronym KIBS which is constructed of two parts. “Knowledge

intensive firms” are firms that, as opposed to the labeling of labor-intensive or capital-intensive operations,

undertake complex operations of a knowledge-intensive nature with human capital as the dominant factor

(Alvesson, 1995). “Business services” are such services demanded by firms and public organizations not

produced for private consumption (Strambach, 2001).

Hence, there are some variants on how to define KIBS. However, most scholars agree upon the branches

and firms comprised by KIBS. Usually, a nomenclature following NACE3 is used, where KIBS include

computer and related activities, research and development, and other business services.4

The majority of studies focusing on KIBS are concerned with innovation patterns. If, how and why KIBS

innovate has been studied in numerous papers, see e.g. Camacho & Rodriguez (2005), Evangelista (2000),

Cainelli et al. (2006) or Tether (2003). Quantitative research comparing KIBS to manufacturing firms in

terms of innovation processes has been performed in Sirilli & Evangelista (1998), Freel (2006) and Wong

& He (2005) and cross-country analyses related to KIBS is found in Tether (2003) and Miozzo &

Grimshaw (2005). Cainelli et al. (2004) relate innovation among KIBS firms to economic performance.

Internationalization of firms and the literature on born global firms

The international orientation of KIBS is, however, less investigated. This is especially true when it comes

to those KIBS firms that can be categorized as born global. The theory on how firms internationalize is

very much based on the manufacturing industry. Traditionally, a sequential approach to the

internationalization process has been the starting point when investigating firms’ internationalization

patterns. In such incremental stage theories, firms start selling their products to their home market and

thereafter sequentially enter other markets. Two main models on stage theories of internationalization can

be identified: Product life cycle theory by Vernon (1966, 1971, 1979) and the Uppsala internationalization

model (Johanson and Vahlne, 1977, 1990, 2006; Johanson and Wiedersheim-Paul, 1975).

Vernon (1966; 1971) states that the internationalization process follows the product life cycle. At early

stages products are produced within the home country and exported to other countries. Thereafter, as the

3 NACE is a European classification of economic activities. 4 See table 1 below for the classification used in this paper.

6

product standardization improves the flexibility of having production at home is no longer as important.

Hence, production plants are set up in other countries in order to serve local markets. Eventually, due to

cost savings, such offshoring reaches less-developed countries as production locations. In later work,

Vernon (1979) identified many firms launching new products in several markets at once. Hence, the

globalization effect of decreasing distances, both logistically and culturally, had made his product life

cycle theory inapplicable to some kinds of firms.

Similarly to the Vernon model, the internationalization model developed by the Uppsala scholars

Johanson and Vahlne (1977, 1990, 2006) states that the “enterprise gradually increases its international

involvement” (Johanson and Vahlne ,1990, p.11). They complement the notion of physical distance with

psychic distance. Hereby they acknowledge that differences in languages, cultures, political system etc. do

matter for the decision of firms to internationalize. As foreign market knowledge increases, a firm’s

psychic distance to the foreign market shrinks. Consequently, it can expand its presence in foreign

countries even further. In this regard, the increasing trends towards a more globalized world are better

reflected by this stage theory approach compared to the product life cycle theory.

Neither the Vernon nor the Uppsala models are, however, fully able to explain the internationalization of

small firms in today’s global markets, see the criticism in e.g. Andersson & Wictor (2003) or Chetty &

Campbell-Hunt (2004). A new view on how firms internationalize, focusing on the evolution of firms born

for global markets, have been developed in order to explain the internationalization process of firms that

start their international activities from birth entering different countries simultaneously. These firms have

been labeled born global firms. This concept was first used in a McKinsey study of manufacturing

exporters in Australia (McKinsey & Co., 1993), which showed a number of small and medium-sized firms

that right from start competed against established players on the global arena. Hence, the existence of such

firms contradicted the stage theories in Vernon (1966, 1971) and Johanson and Vahlne (1977, 1990). The

McKinsey study defined a born global firm as “one which views the world as its marketplace from the

outset; it does not see foreign markets as useful adjuncts to the domestic market”. A similar definition is

found in Oviatt and McDougall (1994). They define born globals as firms that ‘‘from inception, seek to

derive significant competitive advantage from the use of resources and the sales of outputs in multiple

countries’’. There has been a number of follow-up studies on the McKinsey report with the phenomenon

labeled differently: born globals (e.g. Knight and Cavusgil, 1996), global start-ups (Oviatt and McCougall,

1994), international new ventures (McDougall et al. 1994) and instant exporters (McAuley, 1999).

7

There are a number of factors influencing the creation of born global firms. Some have been present since

the original stage theories were constructed, whereas some are influenced by the change in business

climate caused by globalization. Knight and Cavusgil (1996) present several trends that have given rise to

the emergence of born global firms:

The pressure to specialize in order to be competitive has created an increasing amount of niche

markets. In order to be successful in niche markets firms have to increase their customer base by

going global.

Advances in technology regarding production and transportation.

Advances in communication technology.

Advantages of small firms in terms of quicker response time, higher flexibility, adaptability etc.

Globalization itself in terms of knowledge, decreased trade barriers and facilitating institutions.

Entrepreneurs nowadays have more international experience and foreign market knowledge.

Trends towards global networks which are facilitated by advances in information technology.

Compared to the time when the incremental stage models of internationalization were constructed, the

mentioned trends create a breeding ground for setting up born global firms. However, despite these

improved preconditions, it is but a few of all new firms that become born global. There are a number of

characteristics that distinguish these firms. Freeman et al. (2006) think of a number of key variables that

can be positively associated with rapid internationalization.

A too small domestic market.

Commitment and belief by senior management to the idea of internationalization.

Personal networks.

Unique technology as source of competitive advantage.

Growth through partnership and alliances.

Hence, some firms are better suited for a born global approach than others. Adding to these determinants,

one must not forget the perhaps most important reason behind becoming a born global firm, which is to

lock-in new customers and swiftly exploit proprietary knowledge as the main source of competitive

advantage (Bell et al., 2003). This is particularly important in sectors with rapid technological change.

8

Hypotheses

The purpose of this paper is to investigate differences in performance characteristics between born global

KIBS firms and other newly founded KIBS firms. In line with the performance estimations on born global

manufacturing firms in Halldin (2011), the paper tries to provide answers to similar research questions on

born global KIBS firms. The paper will try to investigate the following hypotheses:

Hypothesis 1: KIBS firms that take an active part on global export markets from

the outset are growing faster in terms of employment and sales than other newly

founded KIBS firms.

Hypothesis 2: Born global KIBS firms become more productive than other KIBS

firms.

Hypothesis 3: The sunk costs associated with rapid export market entrance lead

to lower short-run profitability among born global KIBS firms than other KIBS

firms.

These hypotheses are in large motivated by the huge literature on performance of exporters vis-à-vis non-

exporters (see e.g. Clerides et al. 1998, Bernard and Jensen 1999 or Bernard et al. 2007). Such studies

mainly address the issue of an export-productivity link. A “learning-by-exporting” effect is found in

Castellani (2002), Castellani and Zanfei (2003), Criscuolo et al. (2004), Hansson and Lundin (2004),

Greenaway and Kneller (2007), Andersson and Lööf (2009). These studies find evidence that exporting

firms acquire knowledge and technology that can enable firm growth.

The first hypothesis deals with such a potential growth differential between born global KIBS firms and

other KIBS firms with more incremental attitudes towards internationalization. It is predicted that born

global firms grow faster, both in terms of employment and sales. Because of the higher competitive

pressure on world markets, firms that early on are active on export markets should be forced to achieve

higher productivity. According to this reasoning, the answer to the second hypothesis should be

affirmative. However, a higher productivity among born global firms might also reflect a self-selection of

already more productive firms into becoming exporters as is predicted by the Melitz (2003) model where

only the most productive firms choose to enter export markets. When it comes to profitability, we know

that entering export markets is associated with sometimes large sunk cost, see e.g. Braunerhjelm (1999).

9

Incurring such sunk costs at early stages in a firm’s global expansion should be reflected in lower short-

run profitability, which is what the last hypothesis predicts.

3 Data

The paper uses a dataset including business statistics, international trade data on exports in manufacturing

goods, and data on how firms are founded, i.e. firms resulting from spinoffs and mergers or truly new

firms without such links to existing firms. The data is provided by Statistics Sweden. Since trade statistics

is limited to exports in manufacturing goods, much of the exporting activities of firms, especially within

the service sector, are not included. Despite this fact, many KIBS firms tend to export considerable

amounts of manufacturing goods. Hereby, the different classifications of born global firms outlined below

are believed to capture the outwards orientation of KIBS firms to a satisfactory degree.5

KIBS is defined according to the NACE classifications, where KIBS include computer and related

activities (NACE 72), research and development (NACE 73), and other business services (NACE 74). For

a more detailed overview of the included sectors, see Table 1.

5 In the best of worlds, access to export statistics on service goods would of course be of interest, both when studying the

manufacturing industry as in Halldin (2011) and the different strands of the service sector industry, KIBS for instance, as is

performed in this study. Hence, the object of study in this paper is rather KIBS firms with high export intensities in manufacturing

goods.

10

Table 1 Description of the industries included in KIBS

Many spinoffs and merged firms, which are categorized as new in the database, have different

characteristics compared to new firms in total.6 The possibility to identify new firms resulting from

spinoffs and mergers makes us able to sort out truly new firms, which is what we want to include in our

sample. The year of firm birth is defined as the year when the firm first appears in the business statistics.

There are many empirical definitions on born global firms in the literature7. In line with Halldin (2011) the

following three definitions of born global firms are chosen:

6 Take firms size for instance, which is considerably larger for merged firms than for other new firms. 7 See Bals et al. (2008) for other empirical definitions of born global firms. A common definition of born global firms (used e.g.

by Moen & Servais, 2002 and Oviatt & McDougall, 1997) is one that views born globals “as firms less than 20 years old that

internationalized on average within three years of founding and generate at least 25 percent of total sales from abroad” (Knight et

al., 2004, p. 649). This definition is also used on Swedish data in Nordman & Melén, 2008 and Melén & Nordman, 2009.

However, the time period of study here does not allow for an investigation of firms as old as 20 years. Therefore, the three

different definitions are used instead.

NACE code Description

72 Computer and related activities

721 Hardware consultancy

722 Software consultancy and supply

723 Data processing

724 Database activities

725 Maintenance and repair of office, accounting and computing machinery

726 Other computer-related activities

73 Research and development

7310 Research and experimental development in natural sciences and engineering

7320 Research and experimental development in social sciences and humanities

74 Other business activities

741

7411 Legal activities

7412 Accounting, book-keeping and auditing activities; tax consultancy

7413 Market research and public opinion polling

7414 Business and management consultancy activities

742 Architectural and engineering activities and related technical consultancy

743 Technical testing and analysis

744 Advertising

7484 Other business activities n.e.c.

Legal, accounting, book-keeping and auditing activities; tax consultancy; market

research and public opinion polling; business and management consultancy;

11

1. Stringent definition: New firms with at least 25 percent of sales in exports within two years from

inception (abbreviation used below: BGF 2:25)

2. Modest definition: New firms with at least 10 percent of sales in exports within five years from

inception (abbreviation used below: BGF 5:10)

3. Alternative definition: New firms with at least an average of 25 percent of sales in exports for

three consecutive years no later than year two, three and four after firm foundation (abbreviation

used below: BGF 3ma:25)

Using more than one definition is argued to better capture the born global phenomenon first defined in

McKinsey & Co. (1993) or Oviatt and McDougall (1994). In order to better capture persistence in export

behavior, the first two definitions are supplemented with an alternative definition. Occasional exporters

that happen to fulfill the requirements of the definitions are not included. Many of such firms exiting

export markets during the time period of study have low sales and cannot be perceived as born global in

the sense of Oviatt and McCougall (1994).

One advantage of the compiled dataset is that it allows detailed monitoring of labor flows. Hereby, it is

possible to exclude firms that have merged or spun off part of their business during the time period of

study. Such firms cannot be perceived as the same firm between two consecutive years.8 Furthermore,

new firms that are the result of a merger or a spinoff are not included since they cannot be categorized as

truly new firms. The data compiled by Statistics Sweden only include new firms where at least one person

has its main employment.

In total, the dataset stretches from 1997-2008. It is not, however, the whole period that can be utilized

when identifying new born global firms. For born global firms of the stringent definition, the nine-year

period 1998-2006 is used to calculate the number of new firms. Similarly, for born global firms of the

modest definition we use the six-year period 1998-2003 and for those firms that are categorized according

to the alternative definition, the seven-year period 1998-2004 is used.9 Since the intention of the paper is

to describe performance differences of born global firms vis-à-vis the total bulk of new firms,

performance five years after firm inception is chosen. Taking the time span of the studied period into

account, five years seems as good time period in order to include a satisfactory amount of born global

firms without measuring performance differences too close to firm birth. Hence, firms being born during

8 Replacing the staff, merging or spinning off part of the business would imply a somewhat new firm. Therefore, in order not to

compare apples and pears, these firms are removed. Hereby, only organic growth is allowed. 9 When investigating the activities of new firms the years after birth, the data is, hence, available to 2008.

12

the years 1998-2003 are investigated. Furthermore, it is only those firms surviving the first five years that

are included in the final sample.

When comparing the characteristics of born global firms to other firms, the fact that exporting firms are

found to have somewhat different characteristics compared to non-exporting firms must be taken into

consideration. New firms with subsequent presence on export markets, but not satisfying the conditions

necessary to be defined as born global, will therefore be chosen as control group instead of the whole

sample of new firms. Future exporters are defined as those new firms that start exporting during the time

window of study, i.e. 1998-2008.10

The resulting unbalanced dataset includes 584 firms of which 22, 25

and 55 firms can be categorized as born global firms according to the three chosen definitions.

4 Descriptive statistics

Figures 1 and 2 show how the number of new firms and employment in new firms has varied during the

studied time period.11

Clearly, the service sector has experienced a significant increase in the number of

new firms born each year whereas this increase for manufacturing firms is much more modest. KIBS firms

is somewhere in-between with an increase in the number of new firms by almost 60 percent from 1998-

2008. There has been a similar increase in employment by these new firms. The number of new KIBS

firms by far exceeds the number of new manufacturing firm over this decade. Compared to the attention

received by the manufacturing industry in the literature, there are much fewer studies on KIBS. In recent

years, the figures indicate that the generation of new firms and employment in these firms is

approximately three to four times larger within the KIBS industry than the manufacturing industry.

Therefore, when studying entrepreneurship, KIBS has an important role to play.

10 Hence, it is only a subset of all future exporters that are included in the control group since firms starting their export endeavor

later than 2008 cannot be detected in the data. 11 The figures are constructed based upon the sample of firms studied in this paper. Some discrepancies to official statistics might

be present due to the restriction of the data sample. For instance, firms that, during the studied time period, have undergone

significant restructuring in terms of M&As or other events causing major employment turbulence are not included in figures.

13

Figure 1 Creation of new firms divided by industry.

Figure 2 Employment in new firms divided by industry

Judging from the increased literature on born global firms, one would think that the total amount of these

types of firms have become richer. However, Table 2 does not confirm such a trend. In the table, the

number of new and born global firms founded over the 1998-2006 time period is shown. Not even a whole

percent of all new KIBS firms a given year can be categorized as born global. This share has not increased

during the time period of study. Thus, the trends towards globalization together with the other beneficial

0

5000

10000

15000

20000

25000

30000

1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008

KIBS firms Manuf firms Service firms

0

5000

10000

15000

20000

25000

30000

35000

40000

45000

1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008

KIBS empl Manuf empl Service empl

14

factors for starting a born global business mentioned above do not seem to have resulted in more born

global firms, at least not in Sweden.12

Table 2 New KIBS firms and born global KIBS firms 1998-2006

When it comes to the three subcategories of KIBS, i.e. NACE 72, 73 and 74, the distribution of born

global firms are unevenly spread. It is the research and development category, NACE 73, which has the

highest frequency of born global firms. Almost three percent of the new firms in this category are defined

as born global and as much as 24 percent of the firms beginning to export during the time period of study

can be considered as born global. These figures are to be compared to 0.25 and 8 percent for NACE 72

and 74. These two sectors have considerably lower amounts of born global firms.13

Table 3 Decomposition of born global firms into subcategories of KIBS. (Number of firms founded 1998-

2003.)

12 Remember that the absence of statistics on service exports might underestimate the number of born global firms. 13 When looking at those firms surviving the first five years, similar differences between the NACE categories appear.

New KIBS firms # Bgf 2:25 # Bgf 3ma:25 # Bgf 5:10 % Bgf 2:25 % Bgf 3ma:25 % Bgf 5:10

1998 4925 19 5 24 0,39% 0,10% 0,49%

1999 4353 14 6 25 0,28% 0,12% 0,51%

2000 5172 15 8 25 0,30% 0,16% 0,51%

2001 5000 14 3 24 0,28% 0,06% 0,49%

2002 5252 9 7 16 0,18% 0,14% 0,32%

2003 5001 12 9 21 0,24% 0,18% 0,43%

2004 5459 12 9 0,24% 0,18%

2005 6399 9 7 0,18% 0,14%

2006 6906 21 0,43%

Sector NACE 72 NACE 73 NACE 74

Bgf2:25 19 9 55

Bgf5:10 28 13 94

Bgf3ma:25 8 6 24

Bgf average of the three def 18 9 58

New future exporters 232 39 692

All new firms 6460 324 22919

Bgf as share of new future exporters 7,90% 23,93% 8,33%

Bgf as share of all new firms 0,28% 2,88% 0,25%

15

The international expansion of the manufacturing sector has during the past decades been enormous, both

in terms of export and FDI activities. KIBS firms are to a large extent providing services for the

manufacturing industry, e.g. R&D services, accountancy and management services, technical services of

advertising and market research. Hence, it is of interest to compare the international activities and

presence at export markets for the manufacturing sector and KIBS.14

In the first two rows of Table 4, we

can compare the target destinations of the initial export market entries for these two sectors. The table

show similar results except for the lower degree of KIBS firms entering the Nordic region as the first

export market target (49 percent compared to 71 percent for manufacturing firms). In this sense, KIBS

firms seem to be less bound to geographically close regions when embarking on international endeavors.

The remaining columns in Table 4, showing data on the three types of born global firms defined

previously, indicate a similar pattern where manufacturing firms are more likely than KIBS firms to

choose one of the nearby Nordic countries as starting point for export activities. The step-by-step

expansion to more geographically distant markets, predicted by the original stage theories of

internationalization developed by the Uppsala scholars Johanson and Vahlne (1977, 1990, 2006), is less

followed by born global KIBS firms, at least the first year of export market entry. Hence, if one can talk

about degrees of born globalness, KIBS firms seem to be more born global than manufacturing firms.

14 Due to threshold values for registration of exports to EU countries (Eliasson et al., 2011), some export data to this region is not

reported. Hence, there might be a moderate underestimation of the number of born global firms.

16

Table 4 A comparison of export market destinations between manufacturing firms and KIBS firms.

The variables used in the estimations performed below are described in Table 5. The main independent

variables interesting to this paper are the dummies for born global firms. The outcome variables are firm

size, sales per employee, profits over sales and value added per employee. These are measured five years

following firm birth. The estimations control for birth-year characteristics such as firm size, human capital,

equity ratio, dummies for Swedish firms with only national affiliates or with foreign affiliates.

Furthermore, the birth-year values of the performance measures of firm size, sales per employee, profits

over sales and value added per employee are also controlled for. In the estimations and the descriptive

statistics, all quantitative variables are winsorized at one percent. Hereby, winsorizing removes the

extreme outliers and assigns them the values given by the 1st and 99th percentile values respectively.15

15 The estimations are also run without removal of outliers producing no major differences to the results.

A. Export market descriptives for new manufacturing firms the year of export market entry 1998-2003

number of firms percentage number of firms percentage number of firms percentage number of firms percentage

Baltpol 58 15% 25 19% 20 26% 32 15%

Nordic 273 71% 82 62% 52 68% 152 70%

G8 123 32% 63 48% 41 53% 82 38%

EU 34 9% 21 16% 15 19% 26 12%

Others 106 28% 49 37% 30 39% 68 31%

mean median mean median mean median mean median

Exports 1261277 71550 3253226 475221 5080480 583502 2193651 222388

Export destinations 2,05 1 3,00 1 3,66 2 2,49 1

Export destinations (total) 5,83 2 10,26 4 14,30 6 8,28 3

B. Export market descriptives for new KIBS firms the year of export market entry 1998-2003

number of firms percentage number of firms percentage number of firms percentage number of firms percentage

Baltpol 35 12% 15 19% 6 17% 19 16%

Nordic 148 49% 36 44% 17 49% 65 54%

G8 93 31% 32 40% 19 54% 43 36%

EU 22 7% 16 20% 10 29% 17 14%

Others 78 26% 30 37% 12 34% 35 29%

mean median mean median mean median mean median

Exports 858280 24000 3056935 348066 5306844 348066 2089899 169222

Export destinations 1,55 1 2,41 1 3,17 2 2,12 1

Export destinations (total) 4,07 1 6,43 2 12,80 10 6,64 2

Note 1: Baltpol stands for Poland and the Baltic states; Nordic is Norway, Denmark, Finland and Iceland; G8 is USA, Canada, Great Britain, Germany, France,

Italy, Japan and Russia; EU is the 27 members of the EU except those included in G8; Others are the countries not listed above.Note 2: Export destinations is the number of export destination countries the year of the firm's first export market entry; Export destination (total) is the

number of export market destination countries for the firm during 1998-2008

All firms (in total 385) BGF 2:25 (in total 132) BGF 3ma:25 (in total 77) BGF 5:10 (in total 216)

All firms (in total 301) BGF 2:25 (in total 81) BGF 3ma:25 (in total 35) BGF 5:10 (in total 121)

17

Table 5 Definition of independent and dependent variables

In order to complete the description of the estimated variables, Table 6 reports summary statistics for

dependent and independent variables. In the estimations, the sample will contain firms exporting at least

one year during the time window of study, i.e. 1998-2008. Therefore, alongside the full sample of new

firms, the table will also report figures for future exporters. The last three columns show statistics for the

chosen definitions of born global firms.

Table 6 Summary statistics

Comparing born global firms to the overall sample of firms surviving the first five years we see that they

the year of foundation distinguish themselves by higher sales and higher labor productivity. The

differences in other control variables are not as clear. The statistics on the dependent variables five years

after the firms are being founded show that born global firms are larger, have higher sales and are more

productive. However, compared to the year of foundation, born global firms have decreased their average

sales and labor productivity.

Independent variables

Bgf One of the three definitions described above

Human Share of employees with post-secondary education

Eq.ratio Equity over total assets

Size Employment

Sw.aff Swedish group with Swedish daughters

For.aff Swedish group with foreign daughers

Sales Sales per empl

Profits Profit over sales

Lp Value added per empl

Dependent variables (five years after firm birth)

Size Employment

Sales Sales per empl

Profits Profit over sales

Lp Value added per empl

Independent variables Mean SD Mean SD Mean SD Mean SD Mean SD

Size 1,30 1,30 2,23 3,51 1,48 1,16 2,32 3,48 1,98 2,50

Sales 734935 2069633 1516416 6171028 8180016 22800000 7517064 21900000 4176633 14700000

Profits 0,32 3,36 0,05 1,05 0,07 0,30 -0,98 3,62 -0,48 2,59

Lp 376250 701128 474318 1931587 1961910 7109194 1779097 6833989 1022847 4533071

Human 0,58 0,48 0,51 0,45 0,54 0,47 0,58 0,46 0,56 0,46

Eq,ratio 0,23 5,52 0,30 0,97 0,28 0,21 0,25 0,25 0,31 0,27

Sw,aff 0,02 0,13 0,03 0,16 0,00 0,00 0,08 0,28 0,09 0,29

For,aff 0,00 0,04 0,00 0,06 0,00 0,00 0,00 0,00 0,00 0,00

Dependent variables Mean SD Mean SD Mean SD Mean SD Mean SD

Size 1,71 6,34 4,93 23,93 5,35 7,36 6,64 7,97 7,42 10,23

Sales 891460 1686312 1711025 2930317 4120756 5807030 3955229 5601573 3269293 4987204

Profits -70,30 5019,64 -0,70 13,26 -0,22 1,08 -0,34 1,19 -0,43 2,10

Lp 475827 461426 508121 532352 553407 492632 542270 496326 582525 528951

Full sample Future exporters BGF 2:25 BGF 3ma:25 BGF 5:10

(9186-9844 obs) (580-598 obs) (22-23 obs) (25 obs) (53-57 obs)

18

5 Methodology

5.1 OLS

In order to investigate the performance of born global firms compared to other new firms we want to

regress performance five years after firm birth (measured by firm size, sales per employee, profits over

sales and value added per employee) on the born global firm dummies and a number of controls

delineating firm characteristics at birth.16

The reduced form specifications will be estimated using OLS

with the four specifications looking as follows:17

A Sizeit+5 = a0 + a1 Bgfi + a2 Sizeit + a3 Eq.ratioit + a4 Humanit + a5 Sw.affit + a6 For.affit +

industry dummies + time dummies + vit

B Salesit+5 = b0 + b1 Bgfi + b2 Sizeit + b3 Eq.ratioit + b4 Humanit + b5 Sw.affit + b6 For.affit + b7 Salesit +

industry dummies + time dummies + vit

C Profitsit+5 = c0 + c1 Bgfi + c2 Sizeit + c3 Eq.ratioit + c4 Humanit + c5 Sw.affit + c6 For.affit + c7 Profitsit +

industry dummies + time dummies + vit

D Lpit+5 = d0 + d1 Bgfi + d2 Sizeit + d3 Eq.ratioit + d4 Humanit + d5 Sw.affit + d6 For.affit + d7 Lpit +

industry dummies + time dummies + vit

The subscript i indexes firms and t time. All the quantitative performance variables used in the estimations

are in logarithms. Profits is profits over sales and Lp is labor productivity. Sales represent sales per

employee and Size is number of employees. Human is measured by the ratio of employees with a post-

secondary education to the total number of employees. This variable serves as proxy for the human capital

of the firm. To capture differences in financing, Eq.ratio represents the ratio between equity and balance

sheet total. The three dummy variables for born global firms, here abbreviated by Bgf, are defined as

explained above and the dummies Sw_aff and For_aff represent firms with Swedish and foreign affiliates

respectively. Industry class dummies are included to control for fixed effects across industry classes.18

Time dummies control for business cycle effects.

16 Firm birth is here defined as the first year a firm shows up in the business statistics.

17 See Appendix A for a correlation table. The relatively low correlations do not indicate severe problems with multicollinearity. 18 The industry classes are NACE 72-74.

19

5.2 Nearest neighbor matching

The simple OLS estimations above are not able to control for the possible self-selection of future high-

performers into becoming born global firms. In the absence of a counterfactual for these firms it cannot be

excluded that a born global firm should have performed differently than other firms even without the rapid

entrance into export markets. To circumvent this potential problem, this paper implements a matching

procedure based on Abadie et al. (2004) and Abadie & Imbens (2002) called nearest neighbor matching.

In the spirit of Abadie et al. (2004), the notation is as follows:

Let the performance outcome be denoted by Yi,

where (0) 0

( )(1) 1

i i

i i i

i i

Y if BgfY Y Bgf

Y if Bgf

The treatment group is in this case born global firms. In case there would have been access to the

counterfactual, it would have sufficed to calculate Yi(1)-Yi(0) for an individual firm to estimate the

performance differential. However, without such complete information, a similar “twin firm” serves as

proxy for the counterfactual.

Let characteristics used to identify similar firms among born global and other new firms be denoted by X,

where X is a vector of covariates.

There are two regularity conditions:

For all x in the support of X

i) Bgf is independent of (Y(0), Y(1)) conditional on X=x

ii) c < Pr(Bgf = 1|X = x) < 1-c, for some c > 0

For similar firms, i) implies that the choice of becoming a born global firm is purely random, i.e.

assignment to the group of born global firms is independent of the outcomes, conditional on the covariates.

Part ii) is an identification assumption stating that, given a certain covariate pattern, there has to be a

probability to find a similar firm in the opposite group of firms for a match to be possible.

The conditional independence assumption i) requires detailed data on firm characteristics. The covariates

used here in the matching procedure are firm size (employment), total sales, profits, value added, equity

20

ratio, the ratio of employees with post-secondary education, whether the firm has Swedish or foreign

affiliates, industry class and year. The matching is based on covariates the year of firm birth.

6 Results

Table 7 presents the results from the OLS regressions. The coefficients on the born global dummies show

that born global KIBS firms perform better when performance is measured by Size or Sales. These

significant results show that growth in employment among born global KIBS firms are about 50 percent

larger and growth in company sales about 100 percent larger the first five years subsequent to firm

foundation than the equivalent growth for non born global KIBS firms. When it comes to performance in

the productivity and labor productivity estimations, the regressions show more ambiguous results. We

only see significant results for one of the three born global definitions, negative coefficients when

performance is measured by productivity and positive when measured by labor productivity. Hence, these

results should be interpreted more cautiously.19 20

The point estimates for the control variables show that high performers the year of firm birth also perform

superior five years later. Except for profitability, larger initial size influences performance positively.

Large shares of equity in firms’ balance sheets (Eq.ratio) and high ratios of employees with post-

secondary education (Human) could be expected to positively influence performance. Such an effect can

only be found for Eq.ratio in the employment estimation whereas no such predicted effect can be found

for Human. Since but a few of the firms in the sample have affiliates, the point estimates on Sw.aff and

For.aff should not be put too much emphasis on.

19 One should keep in mind that for new firms, profits is probably not the major objective in the short run. Instead, survival and

employment and sales growth is a more important concern for young firms. The ambiguous results in the last two estimations are

therefore not surprising where profits is an important ingredient in the performance measurement. 20 For robustness purposes, Appendix B presents the coefficients of a number of alternative regressions where definitions of born

global firms and the time horizon of firm performance are allowed to vary. These robustness results strengthen the view of born

global firms performing superior in terms of size and sales. They also indicate that born global firms have a tendency to perform

better in terms of profitability and labor productivity when the time horizon is expanded to measure performance six and seven

years after inception. This tendency is very weak when the dependent variable is profitability but for labor productivity most of

the born global variables seem to have positive and significant coefficients.

21

Table 7 OLS results

Similarly to Halldin (2011), the OLS regressions of Table 7 are supplemented with a nearest neighbor

matching procedure which captures the potential self-selection problem of firms with somewhat different

characteristics being more prone to becoming born global firms.

Table 8 shows the results from the matching estimations. The following covariates measured at the year of

firm birth are used to match born global KIBS firms to other new KIBS firms with similar characteristics:

firm size (employment), total sales, profits, value added, equity ratio, the ratio of employees with post-

secondary education, whether the firm has Swedish or foreign affiliates, industry class and year. Both one

and four matches21

are used, with and without bias adjustment22

and the results are found in table 8.23

21 More matches take more of the available information into account when estimating, but more matches also tend to imply more

imprecise matches. A rather small number of matches should preferably be chosen according to Abadie & Imbens (2002).

Dep. var.

(1) (2) (3) (1) (2) (3) (1) (2) (3) (1) (2) (3)

Bgf2:25i 0.511*** 1.057*** -1.391 1.945***

[0.141] [0.297] [0.906] [0.621]

Bgf3ma:25i 0.454*** 0.797*** -2.213** 0.570

[0.137] [0.266] [0.987] [1.409]

Bgf5:10i 0.639*** 0.931*** -0.993 0.249

[0.116] [0.229] [0.640] [0.876]

Sizeit 0.850*** 0.843*** 0.851*** 0.482*** 0.464*** 0.480*** -0.995*** -0.987*** -0.983*** 0.770* 0.738* 0.741*

[0.046] [0.046] [0.045] [0.173] [0.172] [0.173] [0.353] [0.347] [0.351] [0.404] [0.402] [0.401]

Eq.ratioit 0.026*** 0.028*** 0.027*** 0.259 0.262 0.260 -0.080 -0.086 -0.080 0.230 0.231 0.229

[0.010] [0.010] [0.010] [0.243] [0.244] [0.244] [0.065] [0.063] [0.064] [0.231] [0.229] [0.228]

Humanit 0.070 0.068 0.059 0.055 0.054 0.040 0.487 0.489 0.501 0.370 0.373 0.372

[0.066] [0.066] [0.065] [0.257] [0.258] [0.257] [0.458] [0.456] [0.457] [0.532] [0.534] [0.533]

Sw.affit 0.236 0.179 0.081 0.743 0.631 0.514 -1.548 -1.249 -1.219 -2.803 -2.909 -2.909

[0.274] [0.278] [0.252] [0.562] [0.564] [0.552] [1.783] [1.771] [1.798] [3.100] [3.069] [3.099]

For.affit 2.393 2.389 2.447 0.484 0.470 0.542* -3.109 -3.166 -3.158 4.547*** 4.483*** 4.494***

[1.986] [1.986] [1.972] [0.298] [0.297] [0.305] [3.677] [3.667] [3.679] [1.618] [1.652] [1.657]

Salesit 0.159** 0.158** 0.161**

[0.075] [0.076] [0.075]

Profitsit 0.163*** 0.159*** 0.162***

[0.047] [0.047] [0.047]

Lpit 0.235*** 0.237*** 0.237***

[0.068] [0.069] [0.069]

Time dummies Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes

Industry dummies Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes

Observations 584 584 584 584 584 584 566 566 566 584 584 584

R-squared 0.42 0.42 0.45 0.10 0.10 0.11 0.09 0.09 0.09 0.14 0.13 0.13

Robust standard errors in brackets

* significant at 10%; ** significant at 5%; *** significant at 1%

Size, Sales, Profits and Lp are in logarithms

Sizeit+5 Salesit+5 Profitsit+5 Lpit+5

Born global firms are in (1) defined as firms with an exports to sales ratio of at least 25 percent during 1 year within 2 years of inception, in (2) as

at least an average of 25 percent 3 consecutive years during the first 4 years after foundation and in (3) as at least 25 percent during 1 year within 5

years of inception.

22

Except for a few exceptions, being a born global firm only has a positive influence when performance is

measured by employment or sales. The significant point estimates in the Size estimations vary from 0,35

to 0,88 and in the Sales estimations from 0,48 to 0,81. Hence, compared to the OLS estimates, the

importance of being a born global firm using a matching procedure is somewhat lower for performance in

Sales but about the same for performance in Size.

For the Profits and Lp estimations, some significant coefficients are found using some of the matching

estimators. Together with the OLS results, one might see a tendency for born global KIBS firms to

perform worse in terms of profitability and better in terms of labor productivity. However, as in the OLS

estimations, no conclusive effect on performance in profitability or labor productivity can be identified.

Table 8 Results from nearest neighbor matching estimations (1386 observations throughout)

22 Without exact matching in finite samples Abadie & Imbens (2002) show that the matching estimator will be biased. Abadie &

Imbens (2002) and Abadie et al. (2004) explain how to remove some of this bias using a bias-adjusted matching estimator. See

Rubin (1973) and Abadie and Imbens (2002) for formal derivations. 23 Since some firms end up in the born global categories based on their export share during the first year after foundation, i.e. the

same year as the covariates used in the matching procedure, one could potentially question what new results the methodology

brings to the analysis. Due to this fact, the matching regressions are also run on a subsample of firms assigned the status of born

global firms at the earliest one year after foundation. Hereby, the matching of similar “twin” firms precedes the assignment to the

treatment group. The results from matching on this subsample do not alter the fact that born global firms seem to perform better in

terms of Size and Sales.

Dep.Var. Coefficient SE Coefficient SE Coefficient SE Number of matches Bias adjustment

Sizeit+5 0,43*** 0,16 0,44** 0,18 0,42*** 0,11 1 No

0,42*** 0,16 0,39** 0,17 0,35*** 0,10 4 No

0,88*** 0,16 0,68*** 0,18 0,66*** 0,11 1 Yes

0,85*** 0,16 0,69*** 0,17 0,69*** 0,10 4 Yes

Salesit+5 0,48* 0,25 0,79*** 0,14 0,48*** 0,17 1 No

0,44 0,51 0,80*** 0,20 0,55*** 0,15 4 No

0,35 0,25 0,81*** 0,14 0,54*** 0,17 1 Yes

0,51 0,51 0,80*** 0,20 0,57*** 0,15 4 Yes

Profitsit+5 -0,67 0,86 -0,86 1,02 -0,49 0,55 1 No

-0,79 0,86 -0,80 0,84 -0,58 0,48 4 No

-0,83 0,86 -1,04 1,02 -0,73 0,55 1 Yes

-1,01 0,86 -1,00 0,84 -0,87* 0,48 4 Yes

Lpit+5 1,14** 0,49 1,14* 0,61 -0,51 0,76 1 No

0,03 1,38 0,64 1,07 -0,69 0,64 4 No

0,78 0,49 1,21** 0,61 -0,25 0,76 1 Yes

-0,21 1,38 0,71 1,07 -0,45 0,64 4 Yes

Bgf2:25 Bgf3ma:25 Bgf5:10

23

7 Conclusions

This paper has studied a small amount of new firms that both belong to the knowledge intensive business

sector and can be categorized as born global. The focus of attention has been on whether these new firms

perform differently than the sector as a whole five years after foundation in terms of employment, sales

per employee, profits over sales and value added per employee. Most studies on born global firms have

been devoted to the manufacturing sector whereas this paper addresses born global firms in the context of

a knowledge intensive business sector. Compared to the manufacturing industry, this sector has grown

more over the past decade. It has become an important sector in many economies both in terms of

employment and knowledge generation. The paper’s quantitative approach with the total mass of a

country’s born global firms being investigated is novel to the literature on born global firms. In dataset of

new Swedish KIBS firms being founded between 1998-2003 and still existing five years later three

different definitions of born global firms are being employed, a stringent, a modest and an alternative one.

The initial OLS estimations are supplemented by a nearest neighbor matching approach in order to

circumvent the problem of not being able to observe the counterfactual for firms choosing to pursue a born

global strategy. By matching on a number of covariates a counterfactual is created based on firms

seemingly identical in the group of non born global firms. Evidence is found that born global KIBS firms

are performing superior when performance is measured by employment and sales per employee. However,

performance measured by profitability and labor productivity is not found to be greater for born global

KIBS firms. These results are confirmed by a sensitivity analysis with a richer set of born global

definitions and with varying time horizons of performance measurements.

The findings on born global KIBS firms coincide to a large extent with the findings on born global

manufacturing firms found in Halldin (2011). Both papers find that born global firms seem to prioritize

growth in employment and sales. Short-term profits is secondary to these firms. Similar findings of an

inverse relationship between growth in employment and sales and firm profitability have been found in

other studies, see for instance Markman & Gartner (2003) for a study on German firms. Since born global

firms have the highest concentration in research and development KIBS sectors where it often takes longer

to recoup investments in innovations, the low profitability five years after foundation is not surprising.

Future studies will contain longer time series and thereby it will be possible to investigate performance for

born global firms over longer time horizons.

24

References

Abadie, A., Drukker, D., Herr, J. L., & Imbens, G. W. (2004). Implementing matching estimators for

average treatment effects in stata. Stata Journal, 4(3), 290-311.

Abadie, A., & Imberns, G. (2002). Simple and bias-corrected matching estimators for average treatment

effects. NBER Technical Working Paper 283.

Andersson, M. and Lööf, H. (2009) Learning-by-Exporting Revisited: The Role of Intensity and

Persistence. Scandinavian journal of economics, 111: 893-916

Andersson, S., & Wictor, I. (2003). Innovative internationalisation in new firms: Born-globals—the

Swedish case. Journal of International Entrepreneurship, 1(3), 249–276.

Bals, L., Berry, H., & Hartmann, E. (2008). What is a ”Born global” firm? Working paper, Wharton

School, University of Pennsylvania, Philadelphia, United States.

Bell, J., McNaughton, R., Young, S., & Crick, D. (2003). Towards an integrative model of small firm

internationalization. Journal of International Entrepreneurship, 1, 339-362.

Bernard, A.B., Bradford Jensen, J. (1999). Exceptional exporter performance; cause, effect, or both?

Journal of International Economics, 47: 1-25.

Bernard, A.B., Bradford Jensen J., Redding, S.J., Schott, P.K. (2007). Firms in International Trade. NBER

WP series, 13054.

Braunerhjelm, P. (1999). Sunk Costs, Firm Size, and Internationalization. Weltwirtschaftliches Archiv,

Vol. 135, No. 4, 657–674.

Cainelli G, Evangelista R, Savona M. (2004). The impact of innovation on economic performance in

services. Service Industries Journal 24(1): 116–30.

Cainelli G, Evangelista R, Savona M. (2006) Innovation and economic performance in services: a firm-

level analysis. Cambridge Journal of Economics 30: 435–58.

25

Camacho JA, Rodriguez M. (2005) How innovative are services? An empirical analysis for Spain. Service

Industries Journal 25(2): 253–71.

Castellani, D. (2002). Export Behavior and Productivity Growth: evidence from Italian manufacturing

firms. Review of World Economics, 138: 605–628.

Castellani, D., Zanfei, A. (2003). Technology Gaps, Absorptive Capacity and the Impact of Inward

Investments on Productivity of European Firms. Economics of Innovation and New Technology, 12: 555-

576.

Chetty, S. and Campbell-Hunt, C. (2004) ‘A strategic approach to internationalization: a traditional vs a

‘‘Born-Global’’ approach’, Journal of International Marketing 12(1): 57–81.

Clerides, S., Lach, S., Tybout, J. (1998). Is Learning by Exporting Important: micro-dynamic evidence

from Colombia, Mexico and Morocco. Quarterly Journal of Economics, 113: 903-947.

Criscuolo, C., Haskel, J., Slaughter, M.J. (2004). Why are Some Firms more Innovative? – knowledge

inputs, knowledge stocks and the role of global engagement. Tuck School of Business, Mimeograph.

Eliasson, K, Hansson, P. and Lindvert, M. (2011). Do Firms Learn by Exporting or Learn to Export?

Evidence from Small and Medium-Sized Enterprises. Forthcoming in Small Business Economics.

Etemad, H., and Wright, R., (2003). Internationalization of SMEs: Toward a New Paradigm. Small

Business Economics 20: 1-4.

Evangelista R. (2000) Sectoral patterns of technological change in services. Economics of Innovation and

New Technology 9:183–221.

Freel M. (2006) Patterns of technological innovation in knowledge-intensive business services. Industry

and Innovation 13(3):335–58.

26

Freeman, S., Edwards, R., & Schroder, B. (2006), “How Smaller Born-Global Firms Use Networks and

Alliances to Overcome Constraints to Rapid Internationalization”, Journal of International Marketing, 14

(3), pp. 33-63.

Greenaway, D., Kneller, R. (2007). Industry Differences in the Effect of Export Market Entry: learning by

exporting? Review of World Economics, 143: 416-432.

Halldin (2011) Born global firms – do they perform differently? Working paper. Royal Institute of

Technology.

Hansson, P., Lundin, N. (2004). Export as an Indicator on or Promoter of Successful Swedish

Manufactruing Firms in the 1990s. Review of World Economics, 140: 415-445.

Johanson, J., & Vahlne, J.-E. (1977), “The internationalization process of the firm – A model of

knowledge development and increasing foreign market commitment”, Journal of International Business

Studies, 8 (1), pp. 23-32.

Johanson, J., & Vahlne, J.-E. (1990), “The mechanism of internazionalization”, International Marketing

Review, 7 (4), pp. 11-24.

Johanson, J., Vahlne, J-E. (2006). “Commitment and Opportunity Development in the Internationalization

Process: A Note on the Uppsala Internationalization Process Model”, Management International Review,

46 (2), pp. 165-178.

Johanson, J, & Wiedersheim-Paul, F. (1975), “The internationalization of the firm: Four Swedish cases”,

Journal of Management Studies, 12 (3), pp. 305-322.

Knight, G. A., & Cavusgil, S. T. (1996). The born global firm: A challenge to traditional

internationalization theory. In S. T. Cavusgil, & T. K. Madsen, Export internationalizing research—

enrichment and challenges. Advances in International Marketing (Vol. 8) (pp. 11–26). New York, NY:

JAI Press.

Markman, G. & Gartner, W. (2002). Is extraordinary growth profitable? A study of Inc. 500 high-growth

companies. Entrepreneurship Theory and Practice, 27(1), 65-75.

27

McAuley, A. (1999). “Entrepreneurial Instant Exporters in the Scottish Arts and Crafts Sector,” Journal of

International Marketing 7(4), 67–82.

McDougall, P. P., Shane, S., & Oviatt, B. M. (1994). Explaining the formation of international new

ventures: The limits of theories from international business research. Journal of Business Venturing, 9(6),

469–487.

McKinsey & Co. (1993) Emerging Exporters. Australia's High Value-Added Manufacturing Exporters,

Melbourne: McKinsey & Company and the Australian Manufacturing Council.

Melén, S. & Rovira Nordman, E. (2009). The internationalisation modes of Born Globals: A longitudinal

study. European Management Journal, vol. 27, no. 4, pp. 243-254.

Melitz, M., (2003). The impact of trade on intra-industry reallocations and aggregate industry productivity.

Econometrica 71, 1695–1725.

Miles, I. (2005). Knowledge intensive business services: prospects and policies. Foresight – The Journal

of Future Studies, Strategic Thinking and Policy. 7(6):39–63.

Miles, I., Kastrinos, N., Flanagan, K., Bilderbeek, R. and Den Hertog, P. (1995). Knowledge-intensive

business services: users, carriers and sources of innovation. European Innovation Monitoring Systems.

EIMS Publication No 15. Innovation Programme, DG13, Luxembourg.

Miozzo M & Grimshaw D. (2005). Modularity and innovation in knowledge-intensive business services:

IT outsourcing in Germany and the UK. Research Policy 34(9):1419–39.

Moen, O. & Servais, P. (2002). Born Global or Gradual Global? Examining the Export Behavior of Small

and Medium-Sized Enterprises. Journal of International Marketing, 10, (3), 49-72.

Rovira Nordman, E. & Melén, S. (2008). The impact of different kinds of knowledge for the

internationalization process of Born Globals in the biotech business. Journal of World Business, vol. 43,

no. 2, pp. 171-185.

28

Oviatt, B. M.,& McDougall, P. P. (1994). Toward a theory of international new ventures. Journal of

International Business Studies, 25(1), 45–64

Oviatt, B. & McDougall, P. (1997). Challenges for Internationalization Process Theory: The Case of

International New Ventures, Management International Review, vol. 37, no. 2, pp. 85-99.

Oviatt, B. M., & McDougall, P. P. (2005). Defining international Entrepreneurship and Modeling the

Speed of Internationalisation. Entrepreneurship Theory & Practice, 29(5): 537-553.

Rialp, A., Rialp, J. and Knight, G.A. (2005). The phenomenon of early internationalizing firms: what do

we know after a decade (1993–2003) of scientific inquiry? International Business Review 14(2): 147–166.

Redding, S. (1999). Dynamic Comparative Advantage and the Welfare Effects of Trade, Oxford Economic

Papers, 51, 15–39.

Rubin, D. B. (1973). The Use of Matched Sampling and Regression Adjustment to Remove Bias in

Observational Studies. Biometrics, 29, 185-203.

Sirilli G, Evangelista R. (1998). Technological innovation in services and manufacturing: results from

Italian surveys. Research Policy 27(9):881–99.

Tether B.S. (2003). The sources and aims of innovation in services: variety between and within sectors.

Economics of Innovation and New Technology 12(6):481–505.

Vernon, R. (1966). “International investment and international trade in the product cycle”, Quarterly

Journal of Economics, 80 (2), pp. 190-207.

Vernon, R. (1971). Sovereignty at bay: the multinational spread of US enterprises, Basic Books, New

York.

Vernon, R. (1979). “The product life cycle hypothesis in a new international environment”, Oxford

Bulletin of Economics and Statistics, 41 (4), pp.255-67

29

Wong PK, He ZL. (2005). A comparative study of innovation behaviour in Singapore’s KIBS and

manufacturing firms. Service Industries Journal 25(1):23–42.

30

Appendix A: Correlation table

Variable Sizet+5 Salest+5 Profitst+5 Lpt+5 Bgf2:25 Bgf3ma:25 Bgf5:10 Sizet Salest Profitst Lpt Eq.ratiot Humant Sw.afft For.afft

Sizet+5 1.0000

Salest+5 0.0833 1.0000

Profitst+5 -0.185 -0.0239 1.0000

Lpt+5 0.0217 0.4439 0.3452 1.0000

Bgf2:25 0.0766 0.0685 -0.0448 0.0473 1.0000

Bgf3ma:25 0.1182 0.0658 -0.1019 -0.0233 0.8272 1.0000

Bgf5:10 0.1973 0.0876 -0.0703 -0.0438 0.6162 0.6435 1.0000

Sizet 0.5992 0.0616 -0.2168 0.0232 -0.0618 -0.0001 -0.0241 1.0000

Salest -0.1646 0.2572 0.1112 0.2517 0.0286 0.0601 -0.0418 -0.1708 1.0000

Profitst -0.258 0.0132 0.1914 0.0688 -0.0365 -0.0739 -0.0661 -0.2595 0.1067 1.0000

Lpt -0.154 0.2258 0.1253 0.2721 0.0371 0.0028 -0.0644 -0.1673 0.6047 0.4672 1.0000

Eq.ratiot 0.0327 0.0895 -0.0382 0.0389 -0.0042 -0.0112 0.0015 0.0131 -0.0051 0.0103 0.0337 1.0000

Humant 0.0777 0.0077 0.0199 0.0107 0.0169 0.0326 0.0378 0.0408 -0.0128 -0.0853 -0.0555 -0.0492 1.0000

Sw.afft 0.1088 0.0189 -0.0752 -0.092 -0.0332 0.0689 0.1226 0.1167 -0.1568 -0.0298 -0.1004 0.0449 0.0028 1.0000

For.afft 0.1425 0.0161 -0.028 0.0193 -0.0116 -0.0121 -0.0188 -0.0082 0.0199 -0.0388 -0.0879 -0.0032 0.0637 -0.0096 1.0000

31

Appendix B: Robustness regressions

Dep. Var.

(1) (2) (3) (4) (5) (1) (2) (3) (4) (5) (1) (2) (3) (4) (5) (1) (2) (3) (4) (5)

Bgf2:25i 0.243** 0.289** 0.511*** 0.415*** 0.341** 0.695 1.145*** 1.057*** 1.072*** 0.618*** -0.492 -1.052 -1.391 -0.587 -1.86 -0.773 0.145 1.945*** 1.018** 1.695*

[0.099] [0.125] [0.141] [0.135] [0.143] [0.49] [0.445] [0.297] [0.25] [0.16] [0.665] [0.738] [0.906] [1.033] [1.389] [1.189] [0.999] [0.621] [0.496] [0.88]

Bgf3:25i 0.161* 0.234** 0.491*** 0.408*** 0.394*** 0.684* 1.194*** 0.905*** 0.957*** 0.556*** -0.744 -1.109 -1.751* -0.741 -2.432* -1.279 -0.142 0.823 1.503*** 1.835**

[0.09] [0.113] [0.129] [0.131] [0.141] [0.41] [0.406] [0.270] [0.229] [0.165] [0.607] [0.706] [0.919] [0.94] [1.39] [1.116] [0.955] [1.280] [0.574] [0.848]

Bgf4:25i 0.2** 0.335*** 0.648*** 0.666*** 0.766*** 0.707** -0.051 0.832*** 0.75*** 0.581*** -0.609 -1.02 -2.030** -0.125 -2.009 -0.974 -1.337 0.434 0.519 0.984

[0.088] [0.115] [0.141] [0.169] [0.19] [0.351] [0.748] [0.243] [0.226] [0.147] [0.576] [0.673] [0.869] [0.796] [1.279] [0.991] [1.03] [1.161] [0.895] [0.999]

Bgf5:25i 0.211** 0.34*** 0.669*** 0.68*** 0.781*** 0.794** 0.086 0.868*** 0.705*** 0.566*** -0.542 -0.94 -1.666* -0.108 -1.872 -1.175 -1.628 -0.033 0.499 0.985

[0.087] [0.113] [0.133] [0.163] [0.181] [0.349] [0.724] [0.259] [0.223] [0.148] [0.572] [0.664] [0.853] [0.761] [1.211] [0.986] [1.039] [1.173] [0.852] [0.942]

Bgf2ma:25i 0.168 0.245* 0.469*** 0.389*** 0.363*** 1.086*** 0.931** 0.958*** 0.883*** 0.633*** -0.721 -1.464* -2.206** -1.35 -0.42 -2.082 -0.374 0.262 0.932* 0.901

[0.105] [0.132] [0.151] [0.136] [0.137] [0.21] [0.454] [0.295] [0.263] [0.185] [0.717] [0.798] [0.976] [1.112] [1.221] [1.332] [1.102] [1.290] [0.535] [0.763]

Bgf3ma:25i 0.161* 0.211* 0.454*** 0.389*** 0.323** 0.988*** 1.033** 0.797*** 0.809*** 0.464** -0.93 -1.306 -2.213** -0.698 -1.404 -1.987 -0.541 0.570 1.261** 1.643*

[0.096] [0.117] [0.137] [0.135] [0.146] [0.216] [0.434] [0.266] [0.23] [0.182] [0.705] [0.795] [0.987] [1.001] [1.272] [1.251] [1.065] [1.409] [0.563] [0.86]

Bgf4ma:25i 0.152 0.196* 0.424*** 0.352*** 0.323** 0.951*** 1.004** 0.800*** 0.771*** 0.464** -0.892 -1.268 -2.010** -0.573 -1.404 -1.975 -0.537 0.550 1.222** 1.643*

[0.094] [0.115] [0.134] [0.133] [0.146] [0.215] [0.421] [0.258] [0.225] [0.182] [0.69] [0.774] [0.966] [0.956] [1.272] [1.224] [1.036] [1.357] [0.543] [0.86]

Bgf5ma:25i 0.144 0.18 0.391*** 0.308** 0.323** 0.959*** 0.959** 0.808*** 0.66*** 0.464** -0.797 -1.124 -1.846* -0.461 -1.404 -1.895 -0.537 0.569 1.086** 1.643*

[0.093] [0.113] [0.133] [0.133] [0.146] [0.211] [0.414] [0.251] [0.246] [0.182] [0.681] [0.764] [0.941] [0.91] [1.272] [1.2] [1.008] [1.305] [0.529] [0.86]

Bgf2ma:10i 0.221** 0.315*** 0.521*** 0.485*** 0.499*** 1.131*** 1.167*** 0.929*** 0.925*** 0.483*** -0.731 -1.044 -1.427* -0.694 -1.577 -0.556 0.036 0.997 1.475*** 1.578**

[0.086] [0.108] [0.117] [0.133] [0.151] [0.221] [0.374] [0.234] [0.208] [0.147] [0.594] [0.674] [0.806] [0.906] [1.212] [0.993] [0.875] [1.070] [0.505] [0.708]

Bgf3ma:10i 0.241*** 0.322*** 0.562*** 0.589*** 0.649*** 1.226*** 1.22*** 1.076*** 0.986*** 0.454*** -0.616 -0.901 -1.513** -0.344 -1.403 -0.143 0.424 1.339 1.843*** 0.155

[0.088] [0.11] [0.124] [0.152] [0.164] [0.24] [0.364] [0.274] [0.289] [0.169] [0.568] [0.641] [0.764] [0.868] [1.16] [0.914] [0.79] [0.974] [0.601] [1.306]

Bgf4ma:10i 0.225** 0.301*** 0.520*** 0.534*** 0.629*** 1.203*** 1.128*** 0.917*** 0.857*** 0.463*** -0.616 -1.05* -1.787** -0.289 -1.642 -0.002 -0.058 0.611 1.705*** 0.029

[0.088] [0.108] [0.126] [0.158] [0.171] [0.23] [0.348] [0.263] [0.263] [0.151] [0.559] [0.636] [0.753] [0.762] [1.099] [0.877] [0.814] [1.015] [0.578] [1.152]

Bgf5ma:10i 0.233*** 0.348*** 0.576*** 0.585*** 0.687*** 1.183*** 1.073*** 0.805*** 0.657*** 0.422** -0.285 -0.985 -1.971*** -0.663 -1.761* 0.276 -0.222 0.765 -0.002 -0.509

[0.09] [0.113] [0.131] [0.157] [0.17] [0.219] [0.336] [0.239] [0.252] [0.179] [0.534] [0.619] [0.718] [0.756] [1.045] [0.83] [0.805] [0.915] [1.229] [1.451]

Bgf2:10i 0.221*** 0.255** 0.457*** 0.476*** 0.428*** 0.756** 1.119*** 1.021*** 0.929*** 0.586*** -0.705 -1.055 -1.140 -0.685 -0.953 -0.07 0.624 1.796* 1.439*** 2.078**

[0.085] [0.104] [0.117] [0.133] [0.151] [0.37] [0.352] [0.218] [0.196] [0.16] [0.594] [0.681] [0.769] [0.877] [1.154] [0.893] [0.764] [0.949] [0.509] [0.83]

Bgf3:10i 0.195*** 0.243*** 0.433*** 0.437*** 0.464*** 0.866*** 1.233*** 1.017*** 0.914*** 0.566*** -0.622 -0.64 -1.060 -0.432 -0.77 -0.278 0.427 1.198 1.399*** 0.662

[0.075] [0.094] [0.111] [0.134] [0.151] [0.323] [0.34] [0.251] [0.262] [0.161] [0.51] [0.613] [0.696] [0.77] [1.049] [0.803] [0.706] [0.913] [0.519] [1.257]

Bgf4:10i 0.238*** 0.337*** 0.583*** 0.665*** 0.734*** 0.795*** 0.269 0.911*** 0.834*** 0.631*** -0.48 -0.808 -1.321** -0.07 -0.688 -0.071 -0.742 0.548 1.264* 0.204

[0.076] [0.1] [0.119] [0.144] [0.165] [0.287] [0.549] [0.228] [0.221] [0.164] [0.493] [0.559] [0.665] [0.69] [0.942] [0.735] [0.812] [0.881] [0.691] [1.057]

Bgf5:10i 0.271*** 0.365*** 0.639*** 0.718*** 0.793*** 0.855*** 0.379 0.931*** 0.786*** 0.633*** -0.411 -0.67 -0.993 -0.32 -0.97 -0.2 -0.917 0.249 0.462 0.126

[0.078] [0.098] [0.116] [0.142] [0.16] [0.283] [0.526] [0.229] [0.214] [0.17] [0.483] [0.545] [0.640] [0.696] [0.93] [0.724] [0.803] [0.876] [0.95] [0.997]

Observations 852 706 584 478 375 852 706 584 478 375 823 683 566 465 367 852 706 584 478 375

R-squared 0.45 0.42 0.42-0.44 0.43-0.44 0.46-0.50 0.11 0.17 0.10-0.11 0.09 0.06 0.09 0.09 0.09-0.10 0.07 0.04-0.05 0.20 0.21 0.13-0.14 0.12 0.24

Robust standard errors in brackets

* significant at 10%; ** significant at 5%; *** significant at 1%

(1)-(5) represent samples with firms surviving their first 3-7 years respectively. The dependent variables in (1)-(5) are accordingly Size, Sales, Profits and Lp 3-7 years after firm foundation, i.e. k=3,4,5,6,7.

The different definitions of born global firms are Bgfx:yi, where x is years after foundations and y is share of exports in sales. ma stands for moving average.

The same controls as in table 7 are used but their coefficients are obmitted here for illustrative purposes.

Sizeit+k Salesit+k Profitsit+k Lpit+k