GEM High Growth Report 07b

8/3/2019 GEM High Growth Report 07b

1/50

G L OBAL ENTREP RENEURSH I P MONI TOR

2007 Global Report on High-Growth Entrepreneurship

Erkko Autio


2/50

G E re reerh M r

2007 G Rer

Hgh-Grh Erereerh

Erkko Autio

Founding and Sponsoring Institutions

Babson College, Babson Park, MA, US

London Business School, London, UK

Although GEM data were used in the preparation of this report, their interpretation and use are the sole responsibility of the author.

2007 by Erkko Autio, Babson College, London Business School, and Global Entrpreneurship Research Consortium (GERA)


3/50

Frerd

Mazars is delighted to sponsor this report into high-growth entrepreneurship.

Having worked with entrepreneurs around the globe or many years, we understand the issues they

ace. We recognise the importance o entrepreneurship in all its orms, and the importance o high-growthentrepreneurship in particular, to a countrys economy through wealth and job creation.

We urge governments, policy makers and the business community to create the appropriate economic conditions

which will encourage and support an entrepreneurial environment and to develop and promote specic

initiatives aimed at increasing levels o high-growth entrepreneurial activity.

Alistair Fraser

Global Head o the International Customer Lineor Owner-Managed Businesses at Mazars


4/50

Tle Cntents

List Tles 4

List Figres 4

Exectie Smmr 5

Methds 6

Adult-population survey 6

National expert interviews 6

Standardized cross-national data 6

Terminlg 8

Intrdctin 9

Contributions of Entrepreneurial Growth Expectations to Job Creation 9

Summary of Job Contribution Analysis 12

Comparison of Global Regions: Prevalence and Anatomy of High-Expectation Entrepreneurship 12

Adult-population prevalence of High-Expectation Entrepreneurship in World Regions 12

Anatomy of Entrepreneurship in World Regions 14

Summary of the Comparison of World Regions 16

Comparison of Individual Countries: Population Prevalence and

Anatomy of High-Expectation Entrepreneurship 17

Nascent and New Entrepreneur Data 17

Comparison with Established High-Growth Entrepreneurship 19

Summary of Country Comparison 28

Prfle High-Expecttin Entrepreners 9

Summary 33

High-Expecttin Entreprenership nd Ntinl Entrepreneril Frmerk Cnditins 4

Summary 37

Cnclsins nd Discssin 8

Reerences 4

GEM Ntinl Tems 006 4

GEM Spnsrs 47

Tle Cntents


5/50

4

List Tles

Table 1 GEM Method: Sources of Data 7

Table 2 Expected Job Creation by Growth Expectation 11

Table 3 Adult-Population Prevalence Rate of High-Expectation (Nascent and New) and High-Growth

(Established) Entrepreneurs in GEM 2000 2006 Countries 22

Table 4 Relative Prevalence of High-Expectation Nascent and New (20 or More Expected Jobs) and

High-Growth Established (20 or More Current Employees) Entrepreneurs 24

Table 5 Correlations Between Prevalence Rates of High-Expectation (Nascent and New Entrepreneurs,

20 or More Expected Jobs) and High-Growth (Established Entrepreneurs, 20 or More Current

Employees) Entrepreneurship 25

Table 6 Demographic Characteristics of Nascent, New, and Established Entrepreneurs in

High-Income Countries 31

Table 7 Demographic Characteristics of Nascent, New, and Established Entrepreneurs in

Middle- and Low-Income Countries 32Table 8 Bivariate Correlations Between National Entrepreneurial Framework Conditions and

Entrerepeneurship 36

Table 9 Trade-Offs Between SME and High-Growth Entrepreneurship Policies 39

List Figres

Figure 1 Nascent and New Entrepreneurs Categorized According to Growth Expectation

(Global Data, GEM 2000-2006 Countries, Adult-population prevalence Rate) 10

Figure 2 Adult-Population Prevalence of Low-Expectation Entrepreneurial Activity in World Regions

(Percentage of Adult-Age Population) 13

Figure 3 Adult-Population Prevalence of High-Expectation Entrepreneurial Activity in World Regions(Prevalence as Percentage of Adult-Age Population) 14

Figure 4 Anatomy of Entrepreneurship: Relative Prevalence of Low-Expectation Entrepreneurship in

World Regions (Prevalence as Percentage of all Nascent and New Entrepreneurs) 15

Figure 5 Anatomy of Entrepreneurship: Relative Prevalence of High-Expectation Entrepreneurship in

World Regions (Percentage of all Nascent and New Entrepreneurs) 16

Figure 6 Adult-Population Prevalence of High-Expectation Entrepreneurship

(% of 18-64 Year Olds Involved in Nascent or New Firms Expecting 20 or More Jobs) 18

Figure 7 Anatomy of Entrepreneurship: Relative Prevalence of High-Expectation Entrepreneurship

(Nascent and New, 20 or More Expected Jobs) 19

Figure 8 Adult-Population Prevalence of High-Growth Established Entrepreneurs

(Firm Over 42 Months Old, 20 or More Current Jobs) 20

Figure 9 Anatomy of Entrepreneurship: Relative Prevalence of High-Growth Established Entrepreneurs

(Firm Over 42 Months Old, 20 or More Current Jobs) 21

Figure 10 Adult-Population Prevalence of High-Expectation Entrepreneurial Activity

(Nascent and New, 20 or More Expected Jobs) 26

Figure 11 Entrepreneurial Anatomy: Relative Prevalence of High-Expectation Activity

(Nascent and New Entrepreneurs, 20 or More Expected Jobs, as % of All Start-Up Activity) 27

Tle Cntents


6/50

High-expectation and high-growth entrepreneursrepresent only a small percentage o all

entrepreneurial activity.1 Even though 12.3% o theadult-age population in countries that participated

in the Global Entrepreneurship Monitor studybetween 2000 and 2006 are active in emerging andnew entrepreneurial businesses, only 6.5% o new

entrepreneurs (owner-managers o entrepreneurialrms less than 42 months old) expected to have 20

or more jobs in ve years time.

Even though high-expectation entrepreneurship

is rare, its contribution to expected job creationis important. Nascent and new entrepreneurs

expecting to create more than 100 jobs in veyears represent only 1.7% o all nascent and new

entrepreneurs, yet they expect to create nearly 50%

o all expected jobs. Almost 90% o all expected new

jobs are oreseen by less than one-quarter o nascentand new entrepreneurs.

The general patterns o entrepreneurial growth

expectations vary according to a countrysincome level. Even though high-income countries

(generally speaking) have lower rates o overallentrepreneurial activity than do low-income

countries, they generally have higher rates o high-

growth and high-expectation entrepreneurship.

O all world regions, entrepreneurial activity inArica and South America, while high in terms

o adult-age population prevalence,2 is the most

heavily tilted toward low-expectation activity. Inthe richest world regions, including North America,

highly developed Asia, the European Union, andOceania, the anatomy o entrepreneurial activity3

is tilted toward high-expectation entrepreneurship,especially in the category o 20 or more

expected jobs.

There are signicant dierences between individual

countries in terms o both the adult populationand the relative prevalence o high-expectation

entrepreneurial activity. Among high-income

countries, the dierence between the United Statesand Greeces adult-population prevalence rate o

high-expectation entrepreneurship is teenold.

The worlds two largest emerging economies, Chinaand India, exhibit signicantly dierent levels o

high-expectation and high-growth entrepreneurship.

The dierence between China and India isover sixold.

Education and household income, as well asentrepreneurial activities and attitudes, appear

important or high-expectation and high-growthentrepreneurship. High-expectation and high-

growth entrepreneurs are better educated thanother entrepreneurs and the general population.High-expectation and high-growth entrepreneurs

are signicantly more likely to have graduateexperience than low-growth entrepreneurs and the

general population.

High-expectation and high-growth entrepreneurs

are likely to be wealthier than other entrepreneursand the general population in high-, middle-, and

low-income countries.


are generally overrepresented in the manuacturing

and transportation, communication, and utilitiessectors, but underrepresented in agriculture andconsumer services.

The adult population and relative prevalence ohigh-expectation and high-growth entrepreneurship

are, in general, positively associated with thequality o national entrepreneurial policy conditions.

The pattern o positive associations is stronger or

the relative prevalence o high-growth expectations(i.e., percentage o entrepreneurs who expect rapid

growth).

The adult-population prevalence o overall (low-

expectation) entrepreneurship is either nega-tively or neutrally associated with the quality o

national entrepreneurial policy conditions, possiblydue to the high prevalence o low-expectation

entrepreneurship in low-income countries.

Both high- and low-expectation entrepreneurship

are, in general, positively associated with nationalcultural and societal ramework conditions that

aect entrepreneurship. Thus, dierent acets oentrepreneurship react dierently to national policy

and culturalsocietal environments.

1 High-expectation entrepreneurs are nascent and new entrepreneurs who expect more than 20 employees in 5 years time.High-growth entrepreneurs are established entrepreneurs who currently have 20 or more employees.

2 Adult-population prevalence reers to the percentage o adult-age population (1864 years) who are active in entrepreneurship.3 The anatomy o entrepreneurial activity reers to the percentage o the overall population o entrepreneurs who are growth-oriented.

Exectie Smmr


7/50

6

Methds

Since its inception in 1999, the GlobalEntrepreneurship Monitors (GEM) major activity

has been the creation o a large data set andthe construction o harmonized measures o

entrepreneurial activity. GEM collects three typeso data: adult-population surveys, national expertinterviews, and standardized cross-national data.

Adult-Population Survey

Representative samples o randomly selected

adults, ranging in size rom 1,000 to almost 42,000individuals, are surveyed each year in each GEM

country in order to provide a harmonized measure

o the prevalence o entrepreneurial activity. Theannual surveys generally take place between May

and August and are based on three main elements:the sample o respondents, the interview schedule

used to collect the data, and the creation o measuresestimating entrepreneurship at the national level. Theinterview schedule consists o a set o core questions

used to derive entrepreneurial activity rates andadditional questions concerning the attributes and

characteristics o the respondents. The interviewschedule is approved by GEM national teams as

a collective decision in an annual meeting held

in January each year. Both survey and collectionprocedures are revised annually.

All countries in GEM conduct a national adult-

population survey. All countries use region strati-

cation, except or very small countries like Iceland.

Most countries conduct telephone surveys. In somemiddle-income countries where phone penetrationrates are low, interviews are conducted ace to ace

using random door-to-door procedures that also result

in a representative national sample.

While the survey vendors in each country areamong the best available, virtually every data set

provided by every vendor requires some adjustmentsand corrections. Once all data sets are checked

and harmonized, the les are consolidated into a

single data le, each respondent having a uniqueidentication number. The GEM coordination

team then processes the data set to identiy peopleconsidered as entrepreneurially active and to compute

other variables related to entrepreneurial activity.

National Expert Interviews

Each GEM national team conducts up to 50 ace-

to-ace interviews with experts in their respectivecountries. The interviews are intended to assess a

number o entrepreneurial ramework conditions.Experts are selected on the basis o reputation and

experience. In the interviews, experts express their

views on national strengths and weaknesses as acontext or entrepreneurship and indicate what policy

or program changes they believe would enhancethe level o entrepreneurship in their country. The

national experts also complete a standardizedquestionnaire so that GEM can obtain a quantitativemeasure o their opinions concerning their country

as a suitable context or entrepreneurial activity. Thequestionnaire consists o sets o ve to six related

items grouped on the basis o countries and individualcharacteristics relevant or entrepreneurship. These

data are not used in this report but were analyzed in

previous reports.

Standardized Cross-National Data

Standardized cross-national data are obtained rominternational data sources such as the World Bank,

the International Monetary Fund, and the UnitedNations. These data serve in establishing the link

between national levels o entrepreneurial activity and

macroeconomic conditions, as well as the impact o thestate o national conditions required or establishing

this link. While virtually all o the sources o thesecross-national harmonized data are ree, it takes some

eort to annually update, organize, and describe this

material to provide useul consolidated data sets orthe analysis. GEMs data sources are summarized in

Table 1.


8/50

DaTa SouRCE DESCRIPTIoN

GEM Adult-Population Survey Telephone and interview survey conducted by a polling organization in each GEM country, o a minimum o 2,000randomly selected respondents. The data is harmonized to be representative o the adult-age (1864 years old)population o the country.

National Expert Interviews Combined mail questionnaire and interview survey o at least 36 national experts in each GEM countryknowledgeable o national ramework conditions or entrepreneurial activity. The survey questionnairecollects data on fnance, policy, government programs, education and training, technology transer, physicaland business service inrastructure, market openness, social and cultural norms, IPR protection, emaleentrepreneurship, and policy support or high-growth entrepreneurial frms.

Standardized Cross-National Data Compilation o data rom third sources that describe general national conditions: national economy,demographics, society, inrastructure, and institutions. Data are compiled rom publicly available sources suchas the United Nations, the World Bank, the Organization or Economic Co-operation and Development, anddedicated international surveys.

Tle . GEM Methd: Srces Dt

Typically, GEM annual and special reports are based on a single year o data. Because the ocus o the

present report is on a small subset o the overall entrepreneurial phenomenon, a combined data set coveringyears 20002006 is used. Thus, the ndings o the present report are based on a rather sizeable data set

that contains 678,714 interviews o adult-age individuals in 53 countries during years 20002006.

Meth


9/50

8

Terminlg

Entrepreneurship GEM defnes entrepreneurship as any attempt by individuals to start a new frm, including any attempt or sel-employment.

Nascent entrepreneur An adult-age individual (1864 years old) who: has, during the past 12 months, taken tangible action to start a new business would personally own all or part o the new frm would actively participate in the day-to-day management o the new frm has not yet paid salaries or anyone or more than three months

New entrepreneur An adult-age individual who: is currently actively managing a new frm personally owns all or part o the new frm the frm in question is not more than 42 months old

Established entrepreneur An adult-age individual who: is currently actively managing a frm personally owns all or part o the frm the frm in question is over 42 months old

Start-up attempt Nascent or new entrepreneurial frm, as defned above.

Total early-stage entrepreneurialactivity (TEA)

Total early stage entrepreneurial activity reers to the total rate o early-stage entrepreneurial activity among the adultpopulation aged 1864 years. In some instances, this rate is less than the combined percentages or nascent and new frmentrepreneurs. This is because, in circumstances where respondents qualiy as both a nascent and a new frm entrepreneur,they are counted only once.

High-expectation entrepreneur A nascent or new entrepreneur who expects to employ at least 20 employees within fve years time.

High-growth entrepreneur Established entrepreneur who currently employs 20 or more employees.

Adult-population prevalence rate ohigh-expectation entrepreneurs

The percentage o all adult-age individuals in a given country who qualiy as either nascent or new high-expectationentrepreneurs.

Relative prevalence rate o high-expectation entrepreneurs

The percentage o start-up attempts (either nascent or new entrepreneurs) who qualiy as high-expectation entrepreneurs.

Anatomy o entrepreneurship Relative prevalence rate o either high-expectation or high-growth entrepreneurs.


10/50

Intrdctin

Since its inception in 1999, the Global

Entrepreneurship Monitor (GEM) ResearchConsortium has uncovered complex and non-trivial

relationships between entrepreneurship and economic

growth (Wennekers et al., 2005). Entrepreneurialprocesses are undoubtedly linked with macroeconomic

conditions, but detailed relationships may vary(e.g., as a unction o economic development). Both

the GEM data and that rom other sources point

to one important conclusion, however: Not allentrepreneurial activity similarly contributes to

economic growth. Specically, the importance o high-growth entrepreneurial activity or job creation is

increasingly emphasized (Birch et al., 1997; Delmar etal., 2003; Storey, 1994). All entrepreneurial activity isimportant, but high-growth entrepreneurial activity is

particularly so.

Because high-growth entrepreneurial activity israre, most population-level studies on high-growth

entrepreneurship employ post-hoc design and are

overwhelmingly limited to single-country data.International comparative studies on high-growth

entrepreneurship are virtually nonexistent, leavingresearchers and policy makers in the dark as they

seek to understand the high-growth entrepreneurship

phenomenon.

Thanks to its comparative, multiyear approach,GEM is uniquely positioned to address this gap.

GEM measures the growth expectations o both

nascent, new, and established entrepreneurs and

entrepreneurial rms. GEMs is the only data setthat allows international comparisons o the nascententrepreneurial process. The accumulation o data

over several years has created data sets largeenough to allow the kind o ne-grained analysis

necessary to isolate the small number o high-growth

entrepreneurial attempts rom the overall data onnascent and new entrepreneurs.

This report is the second in a series o GEM reviews

o high-expectation entrepreneurial activity. This

report expands the global survey o high-expectationentrepreneurship initiated in 2005 (Autio, 2005). In

this report, we expand on the number o countries

reported and analyze the prevalence o high-expectation entrepreneurial activity in dierent world

regions.4 We also expand the study o the anatomy

o high-expectation entrepreneurial activity in

dierent countries and world regions, as well as on thecharacteristics o the individuals who report high-

expectation entrepreneurial activity.

ContRibutions of

EntREpREnEuRial GRowtH

ExpECtations to Job CREation

Start-up attempts can be categorized accordingto their growth ambition. GEM asks all identied

entrepreneurial attempts how many employeesthey expect to have within ve years time. Figure 1

shows nascent and new entrepreneurs in the GEM

20002006 data, categorized according to expectedjob creation.5 The gure shows the population-level

prevalence (as percentage o adult-age populationrom 18 to 64 years old) o all start-up attempts, both

nascent and new entrepreneurs, at dierent levels ogrowth expectation.

In the GEM 20002006 data set, the adult-populationprevalence o any kind o entrepreneurial activity was

12.3%.6 It is interesting to note that nearly one-halo all start-up attempts do not expect to create any

jobs within ve years. Approximately one-hal o all

entrepreneurial activity thus represents part-timeentrepreneurial activity that may complement income

rom regular employment.

Some 6.3% o the adult-age population in theparticipating GEM countries was involved in nascent

or new start-up attempts that envisioned employing

at least one person within ve years. This percentagealls rapidly as a unction o growth expectation. Some

5.5% o the adult-age population was involved withrms that expected two or more employees within

ve years. Only 2.9% expected to employ at least ve

employees. Start-up attempts expecting to employ 10or more employees represented only 1.7% o the adult-

age population. Only 0.9% o the adult-age populationwas involved in start-up attempts expecting 20

or more jobs. This percentage halved or the 50+category and again or the 100+ category.

4 Because isolating high-expectation entrepreneurs requires large data sets, it is necessary to combine several years o country data in order tomeaningully isolate high-expectation entrepreneurs. With every new round o data collection, more countries meet the minimum data thresholdrequired or meaningul analysis.

5 Methodological note: All start-up attempts or which the data concerning expected jobs in ve years was missing were set as expecting zerojobs in ve years. Figure 1 thus represents a conservative estimate.

6 Methodological note: Weighted according to population and sample size.


11/50

0

Expectations o high growth are rare among nascent

and new entrepreneurs. Only some 7% o all start-upattempts expected to create 20 or more jobs. As many

as 70% did not expect any job creation at all. And only

some 3% o all start-up attempts expected 50 or morejobs.

The scarcity o high-growth expectations among

nascent and new entrepreneurs does not correspond

to their expected contribution to job creation at thecohort level. Overall, the identied nascent and new

entrepreneurs expected to employ some 640,000employees in ve years time (size o base sample:

678,714 adult-population interviews). While this

gure is undoubtedly overoptimistic, the contributionso dierent expectation categories are nevertheless

revealing. These are summarized in Table 2.

In Table 2, we can see that rms expecting to create20 or more jobs, while only representing 7.4% o all

nascent and new rms, expected to create some 73%

o all new jobs created by the cohort. The remaining92.6% o all early-stage entrepreneurs contributed

only an additional 27.4% to the cohort job creationtotal. This statistic was quite similar or both nascent

and new rms, improving our condence in the overall

distribution.

It is remarkable how concentrated the job creation

potential is. Even though early-stage entrepreneursexpecting more than 100 jobs represented less than

2% (1.7% o all) o the cohort, they expected to create

nearly hal o the total jobs within the cohort. Start-up attempts expecting 50 or more jobs (3.4% o all)

represented nearly 60% o total expected jobs. Thus,the high end o the distribution is quite strongly

overemphasized in terms o job creation potential by

nascent and new rms.

The distributions observed coincide well withpublished studies o realized growth potential among

entrepreneurial rms, as well as with studies o rm

size distributions (Cabral et al., 2003; Lotti et al.,2001). The observed disproportionate contribution by

high-expectation start-ups to expected job creation isconsistent with ndings reported in studies o realized

growth (Audretsch, 2002; Birch et al., 1997; Davis etal., 1996; Delmar et al., 2003; Storey, 1994). Also, the

distributions observed or nascent and new rms are

very similar to one another, suggesting that the shapeo the distribution does not change when nascent

entrepreneurs enter the entrepreneurial process.

Figre . Nscent nd Ne Entrepreners Ctegrized accrding t Grth Expecttin(Gll Dt, GEM 000006 Cntries, adlt-Ppltin Prelence Rte)

13%

12%

11%

10%

9%

8%

7%

6%

5%

4%

3%

2%

1%

0%

0+ jobs 1+ jobs 2+ jobs 5+ jobs 10+ jobs 20+ jobs 50+ jobs 100+ jobs


12/50

Tle . Expected J Cretin Grth Expecttin

NEw ENTREPRENEuRSToTaL ExPECTEDJobS FoR CaTEGoRy NuMbER oF CaSES

% oF CoHoRTIN CaTEGoRy

INCREMENTaL

CoNTRIbuTIoN ToCoHoRT ToTaL JobS, %

% oF CoHoRTToTaL JobS

0 or more jobs 281,068 46,149 100.0% 0.0% 100.0%

1 or more jobs 281,068 32,304 70.0% 4.9% 100.0%

2 or more jobs 267,214 18,450 40.0% 8.5% 95.1%

5 or more jobs 243,282 9,217 20.0% 7.8% 86.6%

10 or more jobs 221,243 5,353 11.6% 9.6% 78.7%

20 or more jobs 194,240 2,987 6.5% 15.9% 69.1%

50 or more jobs 149,534 1,101 2.4% 10.4% 53.2%

100 or more jobs 120,303 573 1.2% 42.8% 42.8%

NaSCENT ENTREPRENEuRS

0 or more jobs 380,710 41,529 100.0% 0.0% 100.0%

1 or more jobs 380,710 31,146 70.0% 2.9% 100.0%

2 or more jobs 369,625 20,061 48.3% 6.2% 97.1%

5 or more jobs 345,883 10,999 26.5% 7.0% 90.9%

10 or more jobs 319,075 6,460 15.6% 8.6% 83.8%

20 or more jobs 286,414 3,527 8.5% 10.7% 75.2%

50 or more jobs 245,743 1,894 4.6% 14.8% 64.5%

100 or more jobs 189,354 948 2.3% 49.7% 49.7%

CoHoRT ToTaL

0 or more jobs 661,778 87,677 100.0% 0.0% 100.0%

1 or more jobs 661,778 63,450 72.4% 3.8% 100.0%

2 or more jobs 636,839 38,511 43.9% 7.2% 96.2%

5 or more jobs 589,165 20,216 23.1% 7.4% 89.0%

10 or more jobs 540,319 11,814 13.5% 9.0% 81.6%

20 or more jobs 480,654 6,514 7.4% 12.9% 72.6%

50 or more jobs 395,278 2,995 3.4% 12.9% 59.7%

100 or more jobs 309,657 1,522 1.7% 46.8% 46.8%


13/50

7 Methodoligcal Note: GEMs coverage o Arican countries is narrow. Uganda and South Arica are the only Arican countries included in the GEMdata set.

Summary of Job Contribution Analysis

Expected job contributions are unevenlydistributed across populations o nascent and new

entrepreneurs. While some 6.3% o the adult-agepopulation in the GEM countries were involved innascent and new rms expecting any jobs, only 0.9%

o the adult-age population expected to create 20 ormore jobs in ve years time.

Even though high-expectation entrepreneurship

is rare, its contribution to expected job creation

is important. Nascent and new entrepreneursexpecting to create more than 100 jobs in ve

years represent only 1.7% o all nascent and newentrepreneurs, yet they expect to create nearly 50%

o all expected jobs within the cohort. Almost 90% o

all expected new jobs are oreseen by less than one

quarter o nascent and new entrepreneurs.

The above observation is consistent with received

studies o realized growth in rm populations.

The distributions are similar or nascent and newentrepreneurs, when analyzed separately.

I nearly 50% o all new jobs are expected

by 1.7% o all new entrepreneurs, this hasimportant implications or the design o SME and

entrepreneurship policies. These implications will be

examined later in this report.

CoMpaRison of Global REGions:pREvalEnCE and anatoMy of HiGH-

ExpECtation EntREpREnEuRsHip

Are there dierences between world regions in terms

o entrepreneurial growth ambition? Two questions

can be asked here. First, does the adult-populationprevalence rate o high-expectation entrepreneurs

vary according to global region? And second, does theanatomy o entrepreneurship vary in terms o the

relative prevalence o high-expectation entrepreneurs?

The rst question measures the incidence o high-

expectation entrepreneurs relative to the entire adult-age population. The second question measures theincidence o high-expectation entrepreneurs relative to

the population o nascent and new entrepreneurs.

Adult-Population Prevalence of High-

Expectation Entrepreneurship in World Regions

Figure 2 reports the incidence o low expectation

entrepreneurial activity across world regions. Thepopulation-level prevalence o nascent and new

entrepreneurs expecting zero, one, or ve or moreemployees in ve years is reported. As is well known

rom previous GEM global reports, the incidence o

low-expectation entrepreneurial activity is highest inregions with low GDP per capita. The incidence o any

orm o entrepreneurial activity is highest in SouthAmerica, with a population-level entrepreneurial

activity rate o some 15.5%. That is, a total o 15.5%o adult-age population (1864 years old) in South

American GEM countries are either nascent or new

entrepreneurs. High early-stage entrepreneurialactivity rates are also reported or Arica and

developing Asia.

For the highly developed economies, the highest rates

o entrepreneurial activity are reported or Oceania(Australia and New Zealand), ollowed by North

America. The lowest rates o overall entrepreneurialactivity (combines nascent and new entrepreneurs, all

levels o growth expectation) are reported or highly

developed Asia and the European Union (includingNorway, Switzerland, and Iceland).

The prevalence patterns change or higher levels

o growth expectation. While South America andArica7 remain on top or the category o one or more

expected jobs, developing Asia (including India andChina) alls behind Oceania and North America inthis category. For the category o ve or more expected

jobs, Arica also alls behind Oceania and NorthAmerica, and North America draws to level with

South America. Thus, when moving toward higher

levels o growth aspiration, the prevalence rates oregions with low levels o GDP per capita tend to

drop aster than in regions with high levels o GDPper capita. Highly developed Asia and the European

Union are exceptions to this pattern, however, as

these two regions signal consistently low levels oentrepreneurial activity regardless o the level o

growth expectation.


14/50

Figre . adlt-Ppltin Prelence L-Expecttin Entrepreneril actiit in wrld Regins(Percentge adlt-age Ppltin)

13%

12%

11%

10%

9%

8%

7%

6%

5%

4%

3%

2%

1%

0%

14%

15%

16%

17%

0+ jobs 1+ jobs 5+ jobs

World 2.9%

World 6.3%

World 12.3%

South

Ameri

caAfr

ica

Asia

Deve

loping

Ocea

nia

North

Ameri

ca

Asia

Highly

Deve

loped

EU+

IS+NO

+SW Wo

rld

South

Ameri

caAfr

ica

Asia

Deve

loping

Ocea

nia

North

Ameri

ca

Asia

Highly

Deve

loped

EU+

IS+NO

+SW Wo

rld

South

Ameri

caAfr

ica

Asia

Deve

loping

Ocea

nia

North

Ameri

ca

Asia

Highly

Deve

loped

EU+

IS+NO

+SW Wo

rld

For even higher levels o growth aspiration, the

patterns are strengthened. North America signals

clearly the highest rate o high-expectationentrepreneurial activity, both or the categories o 20

and 50 or more expected jobs in ve years. For thecategory o 20 or more expected jobs, Oceania is at the

same level with South America and developing Asia.For the category o 50 or more expected jobs, South

America alls behind developing Asia. For Arica, the

rates o high-expectation entrepreneurial activityapproach those observed or highly developed Asia and

the European Union.

The level o growth aspiration thus appears to vary

signicantly according to economic context. Eventhough developing economies have high overall adult-

population rates o entrepreneurial activity, North

American and Oceanias entrepreneurial populations

may be o better quality, in terms o high-growth

potential. The dierences between North Americaand Oceania, on the one hand, and Arica and South

America, on the other, may be indicative o dierentsocietal opportunity structures that these regions

oer or their citizens. It is known that countrieswith low levels o GDP per capita tend to have higher

levels o necessity-driven entrepreneurial activity,

whereby individuals start new rms in the absenceo other viable sources o income. In the high-income

economies o North America and Oceania, it may bethat individuals with higher human and social capital

are attracted to entrepreneurship because o perceived

opportunities or wealth creation. The EuropeanUnion and highly developed Asia appear less dynamic

in this regard.


15/50

4

Figre . adlt-Ppltin Prelence High-Expecttin Entrepreneril actiit in wrld Regins(Prelence s Percentge adlt-age Ppltin)

2.0%

1.5%

1.0%

0.5%

0.0%

20+ jobs

World 0.9%

50+ jobs

South

Ameri

caAfr

ica

Asia

Deve

loping

Ocea

nia

North

Ameri

ca

Asia

Highly

Deve

loped

EU+

IS+NO

+SW Wo

rld

South

Ameri

caAfr

ica

Asia

Deve

loping

Ocea

nia

North

Ameri

ca

Asia

Highly

Deve

loped

EU+

IS+NO

+SW Wo

rld

World 0.4%

Anatomy of Entrepreneurship in World Regions

The observed patterns suggest that the anatomy oentrepreneurship varies across world regions. By

anatomy, we reer to the relative prevalence o high-

expectation entrepreneurial activityrelative to theregions population o nascent and new entrepreneurs.

Figure 4 shows the relative prevalence o low-expectation entrepreneurship in world regions. As can

be seen, overall, the relative prevalence o low-growthentrepreneurship is quite similar across regions, with

the exception o developing Asia, where a signicantly

greater number o nascent and new entrepreneurs

than in other regions expect no jobs. When we compare1+, 2+, and 5+ categories, we can observe how

Arica and South America drop in terms o relativeprevalence as growth expectations increase. For the

5+ category, North America and highly developedAsian countries have higher relative prevalence rates

than other world regions. In the European Union and

Oceania, the relative prevalence is at the same levelor each category.


16/50

Figre 4. antm Entreprenership: Reltie Prelence L-Expecttin Entreprenershipin wrld Regins (Prelence s Percentge ll Nscent nd Ne Entrepreners)

70%

60%

50%

40%

30%

20%

10%

0%

1+ jobs 2+ jobs 5+ jobs

Asia

Develop

ingAfrica

Asia

Highly

Develop

ed

EU+

IS+

NO+

SW

North

Ameri

ca

South

Ameri

ca+

Caribbe

an

Oceania

World

Asia

Develop

ingAfrica

Asia

Highly

Develop

ed

EU+

IS+

NO+

SW

North

Ameri

ca

South

Ameri

ca+

Caribbe

an

Oceania

World

Asia

Develop

ingAfrica

Asia

Highly

Develop

ed

EU+

IS+

NO+

SW

North

Ameri

ca

South

Ameri

ca+

Caribbe

an

Oceania

World

World 58.9%

World 43.9%

World 23.1%

For high-expectation entrepreneurial activity, theabove patterns are largely continued, as shown in

Figure 5. In North America, the relative prevalence ohigh-expectation entrepreneurship is clearly higher

than in other world regions. Highly developed Asia,

developing Asia, the European Union, and Oceaniaare roughly at the same level or both 20+ and 50+

categories. For Arica and South America, the relativeprevalence o high-expectation entrepreneurship is

lower than in other world regions.


17/50

6

Figre 5. antm Entreprenership: Reltie Prelence High-Expecttin Entreprenership in wrldRegins (Percentge ll Nscent nd Ne Entrepreners)

15%

14%

13%

12%

11%

10%

9%

8%

7%

6%

5%

4%

3%

2%

1%

0%

20+ jobs 50+ jobs

Asia

Develop

ingAfrica

Asia

Highly

Develop

ed

EU+

IS+

NO+

SW

North

Ameri

ca

South

Ameri

ca+

Caribbe

an

Oceania

World

Asia

Develop

ingAfrica

Asia

Highly

Develop

ed

EU+

IS+

NO+

SW

North

Ameri

ca

South

Ameri

ca+

Caribbe

an

Oceania

World

World 7.4%

World 3.4%

Summary of the Comparison of World Regions

The patterns o entrepreneurial growth expectationsvary across world regions. Regions with low levels

o GDP per capita exhibit high overall levels o

entrepreneurial activity, but some regions (NorthAmerica and Oceania in particular) with high levels

o GDP per capita exhibit higher levels o high-expectation entrepreneurial activity.

North America (the United States, Canada, and

Mexico) clearly exhibits the highest levels o high-

expectation entrepreneurship in categories o 20 and50 or more expected jobs within ve years.

Whereas South America exhibits the highest

overall levels o nascent and new entrepreneurialactivity, it is behind North America, developing

Asia, and Oceania in terms o high-expectation

entrepreneurial activity (50 or more expected jobs inve years).

O low-GDP regions, Arica and South Americaexhibit low levels o high-expectation activity,

whereas developing Asia (including China, India,and Chinese Taipei) exhibits more vibrant activity

even at high levels o growth expectation.

As a region, the European Union (plus Iceland,

Norway, and Switzerland) and highly developed

Asia (including Japan and South Korea) exhibit

the lowest rates o adult-population prevalence

o entrepreneurial activity at all levels o growthexpectation. (See Figures 2 and 3.)

Also, the anatomy o entrepreneurial activity

(relative prevalence o dierent levels o growthexpectations) varies signicantly across world

regions. Again, regions with low levels o GDP per

capita exhibit higher relative prevalence rates o lowgrowth expectations, whereas richer regions exhibit

higher relative prevalence rates o high growthexpectations.

O all world regions, entrepreneurial activity inArica and South America, while high in terms

o adult-age population prevalence, is the mostheavily tilted toward low-expectation activity. In

the richest world regions, including North America,highly developed Asia, the European Union, and

Oceania, the anatomy o entrepreneurial activity is

tilted toward high-expectation entrepreneurship,especially in the category o 20 or more expected

jobs.

North America has the highest adult-age population

prevalence o high-expectation entrepreneurship,and in addition, its anatomy o entrepreneurial

activity is the most heavily tilted toward high-expectation entrepreneurship o any world region.


18/50

CoMpaRison of individual

CountRiEs: population

pREvalEnCE and anatoMy of HiGH-

ExpECtation EntREpREnEuRsHip

Nascent and New Entrepreneur Data

The accumulating research suggests that the nature

o entrepreneurship varies according to economiccontext (Minniti et al., 2006; Wennekers et al.,

1999). In low-income countries (countries with a

low GDP per capita), a relatively greater portion oentrepreneurial activity is initiated because o the

scarcity o alternative sources o income. In high-income countries, necessity-driven entrepreneurship

is virtually nonexistent, and new entrepreneurialinitiatives are typically launched by individualswho already have a job. The dierence between

income substitution and income generation mayalso aect the level o growth aspiration in new

entrepreneurial initiatives. One may speculate that

i an entrepreneurial initiative is undertaken in theabsence o alternative sources o income, the required

growth potential may be lower than in a situation inwhich the ocal entrepreneur already has a source o

steady income. Because o these reasons, the ollowinganalysis groups GEM countries according to their

wealth, as measured by GDP per capita.

Figure 6 presents the adult-population prevalenceo high-expectation entrepreneurial activity8 (bothnascent and new entrepreneurs) in the GEM

countries, grouped on the basis o GDP per capita.

Three country groups are ormed: those with percapita GDP o less than USD 10,000 in 2006; those

with per capita GDP rom 10,000 to 20,000 USD;and those with per capita GDP greater than 20,000

USD in 2006.9 The vertical bars indicate the 95%

condence interval. I vertical bars overlap betweentwo countries, the dierence between those countries

is not considered statistically signicant.

Figure 6 is broadly consistent with the notion thatthe adult-population prevalence o high-expectation

entrepreneurial activity varies according to economic

context. In the group o high-income countries,our countries exhibit clearly higher levels o high-

expectation entrepreneurial activity than other high-income countries: the United States, New Zealand,

Iceland, and Canada. In these countries, over 1% o

the adult-age population is involved with nascentand new ventures that expect 20 or more jobs in ve

years time. In the United Kingdom, Switzerland,Germany, Norway, and Denmark, the adult-population

prevalence rate o high-expectation activity is between

0.5 and 0.8%. O the high-income countries, the

lowest levels o high-expectation entrepreneurshipare observed or Greece, Japan, Spain, Belgium,

France, Finland, and Italy, where the adult-populationprevalence o this activity is clearly less than 0.5%.

The dierences in prevalence rates are considerable,ranging rom the US mean o 1.5% to Greeces mean o

approximately 0.1%.

Looking at high-income countries, it is interesting

to observe how the level o high-expectation entre-preneurship may vary even among broadly similar

economies. O large EU economies, the United

Kingdom and Germany exhibit clearly higher levelso high-expectation entrepreneurship than do France

and Italy. O the Benelux countries, clearly higherlevel high-expectation entrepreneurship is observed

or the Netherlands, as compared to Belgium. O theScandinavian countries, the level o high-expectationentrepreneurship in Iceland is our times as high as

that o Finland.

The comparison between the Scandinavian countriesshows that even countries with quite similar scal and

social security regimes and labor-market institutions

can have dramatically dierent population prevalencerates o high-expectation entrepreneurship. The

Scandinavian countries are known or their well-developed social security systems, high levels o

taxation, and corporatist and regulated labor market

institutions. However, the industry structures between

the Scandinavian countries vary, with Finlandand Swedens industrial bases being dominatedby large-scale orest and engineering industries,

and Denmark, Norway, and Icelands industrial

bases being dominated by small- and medium-sizedindustrial activity. Thus, the dierences between

Scandinavian countries may refect the infuence oindustry structural conditions on high-expectation

entrepreneurship.

O the middle- and low-income economies, China

clearly stands out as a hotbed o high-expectationentrepreneurship.10 The indicated level o Chinas

adult-population prevalence o high-expectation

entrepreneurship is the highest o any GEM country,even though the dierence to the United States, New

Zealand, and Iceland is not statistically signicant.Most other middle- and low-income economies in the

sample exhibit lower adult-population prevalencerates o high-expectation entrepreneurship than

most high-income countries. It is interesting to note

that Indias adult-population prevalence o high-expectation entrepreneurship is only one-th that o

Chinas.

8 Percentage o 18 to 64 year olds involved with either nascent or new entrepreneurial rms who expect at least 20 jobs in ve years time.9 Some GEM countries are excluded rom this graph because o small sample size.10Shenzen is excluded rom Chinas data because o its anomalous nature.


19/50

8

Figre 6. adlt-Ppltin Prelence High-Expecttin Entreprenership(Percent 8 t 64 yer olds Inled in Nscent r Ne Firms Expecting 0 r Mre Js)

2.0%

1.5%

1.0%

0.5%

0.0%

World 0.66%

New

Zealand

UnitedStates

Iceland

Canada

Australia

Singapore

Ireland

Israel

UnitedKingdom

Switzerland

Germany

Norway

Denmark

Netherlands

Slovenia

Sweden

Italy

Finalnd

France

Belgium

Spain

Japan

Greece

Argentina

Russia

Poland

Croatia

Mexico

SouthAfrica

Hungary

China

Brazil

Jamaica

India

World

The analysis o the anatomy o entrepreneurial

activity (or the relative prevalence o high-expectation

entrepreneurs among the population o nascentand new entrepreneurs) reveals slightly dierent

pattern (Figure 7). The countries with arguablythe healthiest entrepreneurial anatomies, in

this comparison, are Singapore, Israel, and China.However, the margins o error or these countries

are large. In Singapore, over 15% o nascent and new

entrepreneurs aspire or rapid growth. Thus, in spiteo its low overall rate o entrepreneurial activity, the

contribution o entrepreneurs to the Singaporeaneconomy may be quite signicant. Greece and Spain

stand out as countries where only a small percentage

(less than 5%) o nascent and new entrepreneursanticipates rapid growth. Also France, Japan, Belgium,

Finland, Italy, and Norway exhibit low levels oentrepreneurial growth ambition, with less than 10%

o all start-up attempts expecting rapid growth.

O the medium- and low-income economies, Russias

entrepreneurs appear to be the most growth-oriented,

with over 10% o nascent and new entrepreneursanticipating rapid growth. India and Jamaicas

entrepreneurial activity is marked by low levelso growth expectation. In these countries, growth

ambitions are roughly at the same level with those inGreece.

Again, broadly speaking, the anatomy oentrepreneurship appears the most biased toward

high-expectation entrepreneurship in high-incomecountries, and middle- and low-income economies

appear to exhibit relatively lower levels o

entrepreneurial growth ambition. There are notableexceptions to this overall pattern, however, both

among high-income and middle- and low-incomeeconomies.


20/50

Figre 7. antm Entreprenership: Reltie Prelence High-Expecttin Entreprenership(Nscent nd Ne, 0 r Mre Expected Js)

25%

20%

15%

10%

5%

0%

World 8.6%

Singapore

Israel

Canada

UnitedStates

Slovenia

Germany

Sweden

Ireland

UnitedKingdom

Denmark

Iceland

Netherlands

Switzerland

New

Zealand

Australia

Norway

Italy

Finland

Belgium

Japan

France

Spain

Greece

Russia

Croatia

Argentina

Poland

SouthAfrica

Hungary

Mexico

China

Brazil

India

Jamaica

World

CoMpaRison witH EstablisHEd

HiGH-GRowtH EntREpREnEuRsHip

The GEM data ocus predominantly on the early parto the entrepreneurial process. By so doing, GEM

provides a unique cross-national look into start-ups

in gestation and gives an indication o where thepotential or entrepreneurial growth is likely to be

ound. Even though growth expectation does notnecessarily lead to growth, growth very rarely occurs

by accident, or without aspiration. This aspirational

aspect o entrepreneurship is uniquely captured inthe GEM data. However, not all growth aspirations

materialize as anticipated. Many expectations arebound to be unrealistic, and the degree o realism

(or overcondence) may vary among countries. In

addition, only one-third o nascent entrepreneurseventually started a rm (Reynolds, 2007). Thereore,

even though aspiration is an important preconditiono growth, it needs to be complemented by data on

realized growth to provide a more accurate picture o agiven countrys entrepreneurial growth potential.

In addition to growth expectations, GEM also collectsdata on realized growth by established entrepreneurs.

These data provide an alternative lens into the high-growth phenomenon, complementing the picture

provided by growth expectations.11 In this report,

all owner-managers o entrepreneurial rms thatcurrently have 20 or more employees are denoted as

high-growth established entrepreneurs.

Figure 8 shows the adult-population prevalence o

established entrepreneurs who currently employ20 or more employees. As in previous graphs, the

countries are grouped according to wealth. Broadlyspeaking, the adult-population prevalence o high-

growth established entrepreneurs appears to behigher in high-income countries than in middle-

and low-income countries. However, because o thescarcity o the phenomenon, the condence intervalsare quite large. O high-income countries, New

Zealand is ranked on top, but the dierence gainsstatistical signicance only in comparisons beginning

rom Norway. New Zealands rate o high-growth

entrepreneurship is signicantly higher than that oNorway, the United States, Germany, Sweden, Italy,

11 Note, however, that even this data has limitations. The data on realized growth reports only achieved scale o operation at the time ointerview, and not how quickly this size has been reached. Also, the minimum economic scale varies in dierent industry sectors. The termhigh-growth is nevertheless used to describe entrepreneurial rms with 20 or more current employees, because only a small minority o allentrepreneurial rms ever reach this scale. Because o the nature o the data, however, high-growth does not mean rapid growth, and itdoes not necessarily even mean current growth in this study.


21/50

0

and France. Interestingly, Iceland ranks higher than

other Scandinavian countries, with the exceptiono Denmark. Icelands adult-population prevalence

o high-growth entrepreneurs is three times that o

Finland.

O the low-income countries, China again stands out,and Indias perormance is much lower. Chinas adult-

population prevalence o high-growth entrepreneurs

is three to our times as high as that o India. AlsoJamaica, Mexico, Croatia, and South Arica stand out

as countries where established entrepreneurs seldomachieve rapid growth.

The data in Figure 8 provide a refection o the

established size distribution o entrepreneurial rmsin the GEM countries. It is important to observe

that industry structures dier rom one country to

another, and the minimum ecient scale o rms alsodiers across sectors. Whereas expectations provide

an idea o the uture-oriented growth aspirationso entrepreneurial rms, Figure 8 also refects (in

addition to past growth aspirations) the infuence o,

or example, industry structure, and possibly eventhe countrys institutional ramework on the size

distribution o established entrepreneurial rms.

Figre 8. adlt-Ppltin Prelence High-Grth Estlished Entrepreners(Firm oer 4 Mnths old, 0 r Mre Crrent Js)

1.0%

0.8%

0.6%

0.4%

0.2%

0.0%

World 0.38%

NewZealand

Iceland

Canada

Netherlands

Israel

Switzerland

Slovenia

Denmark

Australia

Japan

Ireland

Singapore

UnitedKingdom

Norway

UnitedStates

Germany

Sweden

Italy

Spain

Finalnd

Belgium

France

Poland

Hungary

Argentina

Croatia

SouthAfrica

Mexico

China

Brazil

India

Jamaica

World


22/50

The relative prevalence o high-growth established

entrepreneurs (or the anatomy o establishedentrepreneurship) is shown in Figure 9. In this

comparison, Singapore is ranked highest, although

the dierence gains statistical signicance onlyin comparison rom Germany onwards (with the

exception o France). Curiously, Finland stands outas a country where the relative prevalence o high-

expectation entrepreneurship is lower than it is in

other high-income GEM countries. Given Finlandslarge overall sample size, this nding appears quite

reliable and may be indicative o poor ability to grow

entrepreneurial rms. Among high-income countries,

low relative prevalence rates o high-growth entre-preneurship are also observed or Spain, Belgium,

Australia, and Sweden.

Among low-income countries, China again stands

out or its anatomy o entrepreneurship. However,even Chinas relative prevalence rate is signicantly

smaller than that o many high-income countries

starting rom Germany. Jamaica, India, and Brazilstand out or their low relative prevalence, together

with Mexico, and, in some comparisons, Finland.

Figre 9. antm Entreprenership: Reltie Prelence High-Grth Estlished Entrepreners(Firm oer 4 Mnths old, 0 r Mre Crrent Js)

18%

16%

14%

12%

10%

8%

6%

4%

2%

0%

World 6.0%

Singapore

Israel

Netherlands

Denmark

Slovenia

Iceland

Canada

Japan

Germany

UnitedKingdom

UnitedStates

France

New

Zealnad

Italy

Norway

Iralnd

Switzerland

Sweden

Australia

Belgium

Spain

Finalnd

Hungary

Poland

Croatia

SouthAfrica

Argentina

Mexico

China

Brazil

India

Jamaica

World


23/50

To assess the validity o growth expectations as a measure o entrepreneurial growth potential, GEMs growth

expectation data were compared against data on realized growth by established entrepreneurs in the GEMcountries. In addition to data on nascent and new entrepreneurs and their expected growth, GEM also records

data on established entrepreneurs and the realized growth o these. Table 3 shows the adult-population

prevalence rates o nascent and new high-expectation entrepreneurs (20 or more expected jobs), on the onehand, and established high-growth entrepreneurs (owner-managers who currently have 20 or more employees),

on the other. Both prevalence rates and rankings are shown.

Tle . adlt-Ppltin Prelence Rte High-Expecttin(Nscent nd Ne) nd High-Grth (Estlished) Entrepreners in GEM 000006 Cntries

aDuLT PoPuLaTIoN PREvaLENCE

NaSCENT aND NEw ENTREPRENEuRS ESTabLISHED ENTREPRENEuRS

CouNTRy PREvaLENCE RaNk PREvaLENCE RaNk

China (PRC) 1.68% 1 0.60% 3

USA 1.49% 2 0.40% 16New Zealand 1.41% 3 0.76% 1

Iceland 1.28% 4 0.73% 2

Canada 1.23% 5 0.57% 4

Argentina 1.07% 6 0.28% 21

Australia 1.04% 7 0.51% 10

Singapore 0.94% 8 0.44% 13

Israel 0.91% 9 0.54% 6

Ireland 0.91% 10 0.47% 12

United Kingdom 0.67% 11 0.42% 14

Switzerland 0.65% 12 0.54% 7

Brazil 0.65% 13 0.17% 26

Germany 0.64% 14 0.37% 18

Norway 0.63% 15 0.42% 15

Denmark 0.62% 16 0.51% 9

Poland 0.58% 17 0.40% 17

Netherlands 0.51% 18 0.56% 5

Slovenia 0.50% 19 0.52% 8

Croatia 0.47% 20 0.17% 27

Sweden 0.46% 21 0.34% 19

Mexico 0.35% 22 0.02% 32

Italy 0.35% 23 0.27% 22

South Arica 0.34% 24 0.07% 29

Hungary 0.30% 25 0.33% 20

Jamaica 0.30% 26 0.02% 31

Finland 0.29% 27 0.24% 24

France 0.25% 28 0.12% 28

India 0.24% 29 0.07% 30

Belgium 0.22% 30 0.19% 25

Spain 0.22% 31 0.27% 23

Japan 0.14% 32 0.48% 11


24/50

Table 3 shows that while some countries are more

optimistic than others, the correspondence betweenhigh expectations by nascent and new entrepreneurs

and high growth by established entrepreneurs is quite

good. The correlation between high expectations andhigh growth is high: 0.67 (statistically signicant at

the level o p < 0.001). Nevertheless, we can observesome important dierences between high expectations

and high growth. Even though the United States

ranks second in terms o high-expectations bynascent and new entrepreneurs, it only ranks

sixteenth in terms o realized growth by establishedentrepreneurs. This would seem to t well with the

general perception that US citizens tend to be quite

optimistic and sel-condent, perhaps more so thanin many other countries. China, on the other hand,

ranks highly in both measures, as do New Zealand,Iceland, and Canada. However, while Japan ranks last

(thirty-second) in terms o high expectations, it rankseleventh in terms o actual high-growth perormance.

This, again, would seem to be consistent with the

perception that the Japanese may be quite careul

and conservative in their general outlook. The lowestadult-population prevalence rates or established high-

growth entrepreneurs can be observed or Mexico,

Jamaica, India, and South Arica, none o which is ahigh-income country.

In Table 4, Singapore ranks on top among the 32

GEM countries in terms o relative prevalence o

high-expectation and high-growth entrepreneurs (i.e.,anatomy o entrepreneurship see Table 4 on page

24). The correlation between high-expectation andhigh-growth rates is statistically highly signicant:

0.75 (p < 0.001). In addition to Singapore, Israel

stands out or its high relative prevalence o high-expectation and high-growth entrepreneurs. The

lowest relative prevalence rates o high-growth andhigh-expectation entrepreneurs are reported or

Jamaica, India, Mexico, Spain, and Brazil, which, withthe exception o Spain, are all middle- to low-income

countries.


25/50

4

Tle 4. Reltie Prelence High-Expecttin Nscent nd Ne(0 r mre Expected Js) nd High-Grth Estlished (0 r Mre Crrent Emplees) Entrepreners

RELaTIvE PREvaLENCE (aNaToMy oF ENTREPRENEuRSHIP)

NaSCENT aND NEw ENTREPRENEuRS ESTabLISHED ENTREPRENEuRS

CouNTRy PREvaLENCE RaNk PREvaLENCE RaNk

Singapore 16.7% 1 12.2% 1

Israel 14.6% 2 11.2% 2

Canada 13.6% 3 9.0% 7

USA 13.0% 4 7.4% 13

China (PRC) 12.4% 5 4.7% 25

Slovenia 12.0% 6 9.6% 5

Germany 11.9% 7 8.5% 10

Sweden 11.6% 8 5.7% 21

Ireland 11.5% 9 6.3% 18United Kingdom 11.1% 10 7.8% 11

Denmark 11.1% 11 9.8% 4

Iceland 11.0% 12 9.2% 6

Netherlands 10.6% 13 10.6% 3

Switzerland 10.0% 14 6.2% 19

Croatia 9.6% 15 5.8% 20

New Zealand 9.2% 16 7.1% 15

Australia 8.9% 17 5.7% 22

Argentina 8.7% 18 3.6% 27

Poland 8.1% 19 7.5% 12

Norway 7.8% 20 6.6% 17

Italy 6.6% 21 6.7% 16

Finland 6.1% 22 2.9% 28

Belgium 6.1% 23 5.2% 24

Japan 6.0% 24 8.6% 8

South Arica 5.9% 25 5.7% 23

France 5.6% 26 7.4% 14

Hungary 5.5% 27 8.5% 9

Brazil 4.7% 28 1.8% 29

Spain 3.5% 29 4.1% 26

Mexico 3.4% 30 1.0% 30

India 1.9% 31 0.7% 31

Spain 1.5% 32 0.2% 32


26/50

Table 5 shows bivariate correlations between early-

stage entrepreneurial activity (overall and high-expectation), established entrepreneurial activity

(overall and high-growth), and the anatomy o both

early-stage and established entrepreneurial activity(relative prevalence).

The table reveals several interesting patterns. First,

the adult-population prevalence o total early-stage

entrepreneurial activity (rst vertical column)correlates strongly with the adult-population pre-

valence o high-expectation TEA12, as well as withthe adult-population prevalence o established

entrepreneurial activity. However, overall TEA isnot associated with the adult-population prevalenceo high-growth established entrepreneurship. This

suggests the presence o actors that intervene toobscure the relationship between overall levels o

entrepreneurship and entrepreneurial rm growth. Onthe other hand, a positive correlation exists between

high-expectation TEA and high-growth established

entrepreneurship.

The correlations between adult-population prevalenceo entrepreneurial activity and the anatomy o

entrepreneurship (i.e., relative prevalence o high-

growth and high-expectation entrepreneurship) arerevealing. Overall, TEA correlates negatively and

signifcantly with the relative prevalence o both

high-expectation and high-growth entrepreneurialactivity. The adult-population prevalence o high-

expectation TEA, in contrast, correlatespositively with

high-expectation anatomy and has no relationshipwith high-growth anatomy. Also, the adult-population

prevalence o established entrepreneurs correlatesnegatively with both aspects o the anatomy o

entrepreneurship, whereas high-growth established

entrepreneurship correlatespositively with bothaspects o the anatomy o entrepreneurship.

This set o correlations suggests complex relationships

between overall population rates o entrepreneurship,

the anatomy o entrepreneurship, and the populationrates o high-growth and high-expectation

entrepreneurship. It appears that high-growth andhigh-expectation entrepreneurship may be quite

dierent phenomena, and each may be associatedwith and driven by dierent macroeconomic, cultural,

and demographic conditions. At the very least,

on the basis o correlations reported in Table 5, itcannot be claimed that increasing overall levels o

entrepreneurship would automatically lead to greaterentrepreneurial growth. Clearly, urther theory

development and empirical research is required to

ully understand the patterns observed in Table 5.

12 Total early-stage entrepreneurial activity.13 Methodological note: The correlations compare distribution means grouped by year and country. Pearson correlations, 2-tailed signicance;*** p < 0.001; ** p < 0.01; * p < 0.05.

Tle 5. Crreltins beteen Prelence Rtes High-Expecttin(Nscent nd Ne Entrepreners, 0 r Mre Expected Js) nd High-Grth

(Estlished Entrepreners, 0 r Mre Crrent Emplees) Entreprenership

4 5

Total Early-StageEntrepreneurial Activity TEA(Adult-Population Prevalence)

1 Overall activity

2 High-expectation 0.625(**)

EstablishedEntrepreneurial Activity(Adult-Population Prevalence)

3 Overall activity 0.653(**) 0.329(**)

0.287(**)4 High-Growth -0.018 0.241(**)

High-Expectation Activity(Relative Prevalence)

5 % TEA -0.135(*) 0.549(**) -0.199(**) 0.320(**)

High-Growth Activity(Relative Prevalence)

6 % Established Entrepreneurs -0.354(**) 0.000 -0.319(**) 0.661(**) 0.435(**)


27/50

6

The relationship between economic wealth and

high-expectation entrepreneurship is explored inFigure 10. Figure 10 shows the relationship between


entrepreneurship and a countrys real GDP percapita (PPP) in 2006. The data seem to exhibit a

reasonably linear pattern, dierent rom GEMsanalyses o overall levels o entrepreneurial activity

(Minniti et al., 2006). The t is not very strong,

however, indicating that actors other than GDPper capita exercise an important infuence on the


entrepreneurship. Even though a number o countriesall onto an almost direct line, the spread seems to

increase as a unction o GDP per capita.

Figre 0. adlt-Ppltin Prelence High-Expecttin Entrepreneril actiit(Nscent nd Ne, 0 r Mre Expected Js)

1.60%

1.40%

1.20%

1.00%

0.80%

0.60%

0.40%

0.20%

0.00%

y=1E-07x+0.0028

R2 =0.1348

(China excluded from curve-fitting as outliar)

0 5000 10000 15000 20000 25000 30000 35000 40000 45000 50000

Real GDP (PPP) per Capita 2006 (USD)


28/50

The relationship between GDP per capita and the

anatomy o entrepreneurship is shown in Figure 11. Inthis gure, the relative prevalence o high-expectation

entrepreneurs (as percentage o a countrys population

o nascent and new entrepreneurs) is indicated. Again,a linear curve provides the best t with the data,

and the R2 is much higher than or adult-populationprevalence.

The dierences between adult-population prevalenceand relative prevalence patterns may be indicative

o the eect o opportunity costs on high-growthexpectations. Even though a mild linear pattern could

be observed or the relationship between per-capita

GDP and adult-population prevalence rate o high-expectation entrepreneurship, other actors also

seem to be at play. On the one hand, one potentialinfuence could be industry structural conditions,

which could exercise an infuence on the overall level

o entrepreneurial activity, at any level o growth

expectations. On the other hand, the link betweenper-capita GDP and the anatomy o entrepreneurship

may refect the eect o opportunity costs on high-

growth expectations. Remember that in high-incomecountries, nearly all entrepreneurial activity is

opportunity-driven, undertaken by individuals whoalready have a steady employment. For an employed

person to switch rom steady income to more uncertain

and volatile entrepreneurial income, a higherupside potential is probably required than in the

case o necessity entrepreneurship. I an individualhas no current income, the opportunity cost o the

entrepreneurial career option will be small, and the

required upside potential may not need to be as highor the individual to choose the entrepreneurial option.

Thus, opportunity costs could partly explain thepattern observed in Figure 11.

Figre . Entrepreneril antm: Reltie Prelence High-Expecttin actiit(Nscent nd Ne Entrepreners, 0 r Mre Expected Js, s Percent all Strt-up actiit)

(China excluded as outliar)

y=2E-06x+0.034

R2 =0.2909

18%

16%

14%

12%

10%

8%

6%

4%

2%

0%

0 5000 10000 15000 20000 25000 30000 35000 40000 45000 50000

Real GDP (PPP) per Capita 2006 (USD)


29/50

8

suMMaRy of CountRy CoMpaRison

There are signicant dierences between individualcountries in terms o both the adult-population

and relative prevalence o high-expectationentrepreneurial activity. Among high-income

countries, the dierence between the United Statesand Greeces adult-population prevalence rate o

high-expectation entrepreneurship is teenold.

There are surprisingly large dierences in high-

expectation entrepreneurship even among countrieswith similar GDP per capita and similar social

security and scal regimes. The comparison among

the Scandinavian countries may suggest theimportance o industry structural conditions as

a determinant o adult-population prevalence o

high-expectation entrepreneurship in high-incomecountries.

The two largest emerging economies in the world,

China and India, exhibit dramatically dierentlevels o high-expectation and high-growth

entrepreneurship. The dierence between Chinaand India is over sixold.

As a general rule, high-income countriesappear to perorm better in terms o high-

expectation entrepreneurship than in terms ooverall entrepreneurial activity rates. O high-

income countries, the United States, Israel,

Iceland, and Canada exhibit the highest adult-population prevalence rates o high-expectation

entrepreneurship.

The United States ranks higher in terms o high-

expectation entrepreneurship than in terms o high-growth (established) entrepreneurship. In terms o

high-growth established entrepreneurship (adult-

population prevalence o established entrepreneurswho currently have 20 or more employees), the

United States ranks behind Japan. Note, however,that high-expectations and high-growth are quite

dierent phenomena and cannot be directly

compared with one another.

In terms o the anatomy o entrepreneurship,high-income countries do better than middle- and

low-income countries. Singapore stands out as the

country with the highest relative prevalence ohigh-expectation entrepreneurship.

The patterns between per-capita GDP and adult-

population and both adult-population and relativeprevalence o high-expectation entrepreneurship

may suggest the eect o career opportunity costs as

a contributing actor to the high relative prevalenceo high-expectation entrepreneurs in high-income

countries.


30/50

Prfle High-Expecttin Entrepreners

An individuals decision to launch a new venture

is aected by both contextual actors (i.e., theentrepreneurs environment) and his or her personal

characteristics and skills. The size and shape o an

opportunity (and related growth expectation) isnot dened by the environment alone, but rather by

the match between opportunity and the personalcharacteristics o the individual considering it (Shane,

2000). A given opportunity may represent signicant

growth potential or some individuals (or example,those possessing the right skills and social capital to

eectively exploit it) but only a small potential orothers (i.e., those whose human and social capital is

poorly suited to exploit the opportunity). Thus, growth

expectations and realized growth may be infuencedby individuals characteristics, and it may be possible,

at least to some extent, to identiy characteristicsassociated with high-growth expectations and realized

entrepreneurial growth.

Because the GEM data contain demographic data

on nascent, new, and established entrepreneurs,they oer a good opportunity to explore the prole

o high-expectation and high-growth entrepreneurs.Table 6 provides a summary o the proles o

nascent, new, and established entrepreneurs in

high-income countries, as compared against thegeneral adult population. The table is organized

in three major columns, with one each or nascent,new, and established entrepreneurs. The prole o

high-expectation and high-growth entrepreneurs is

presented in the middle o each major column. For

nascent and new entrepreneurs, high-expectationentrepreneurs (20 or more expected jobs in ve years)are compared rst against the general population o

nascent or new entrepreneurs (let subcolumn), andsecond against the adult-age population in general

(right subcolumn). For established entrepreneurs

(owner-managers o rms over 42 months old), high-growth entrepreneurs (those who currently have 20

or more employees) are similarly compared againstthe general population o established entrepreneurs

and adult-age population in general. Statistical

signicances o the dierences are also indicated.

Because o the large size o the GEM sample, most

dierences are statistically signicant. In a sampleo this magnitude, a dierence between means can be

statistically signicant without being very meaningulin practice. A closer examination o the table reveals

many interesting patterns, however.

In terms o age prole, nascent high-expectationentrepreneurs are most heavily biased toward

young individuals, whereas established high-growth

entrepreneurs are biased toward older individuals. Inall, 18.8% o nascent high-expectation entrepreneurs

were rom 18 to 24 years old, whereas only 3.3% oestablished high-growth entrepreneurs were observed

in this age bracket. In contrast, the two oldest age

categories are overrepresented among high-growth

established entrepreneurs and under-representedamong nascent and new entrepreneurs.

While women are underrepresented among allcategories o entrepreneurs, they are particularly

underrepresented among high-expectation and high-growth entrepreneurs. Only some one-third o all

categories o entrepreneurs were women, whereas less

than one-quarter o the high-expectation and high-growth entrepreneurs were women.

For secondary education, lower levels o education

are underrepresented among all categories o

entrepreneurs, but particularly so among high-expectation and high-growth entrepreneurs. With

the exception o graduate experience, nascent andnew high-expectation entrepreneurs tend to have

higher education, perhaps refecting the higheropportunity costs aced by this group, as highly

educated individuals typically have more alternative

employment opportunities.

Also, household income distinguishes entrepreneurshigh-expectation and high-growth entrepreneurs,

in particularrom the general population. Only

one-th o high-expectation nascent entrepreneursbelonged to the lowest third household income

category, and less than one-tenth o high-expectationand high-growth new and established entrepreneurs.

Some 43% o nascent high-expectation entrepreneurs

belonged to the highest income tier, as did as many as

63% o high-growth established entrepreneurs.

While the generally higher wealth o established

high-growth entrepreneurs probably refects theirentrepreneurial success, the wealth o nascent and

new high-expectation entrepreneurs may have more

to do with social stratication o growth opportunities.It may be that wealthy individuals, thanks to their

social capital, may both get to see more growthopportunities, and their wealth, human capital, and

societal connections may also make them better

equipped to pursue such opportunities.

Not surprisingly, entrepreneurs in all categories

are more likely to have either ull- or part-timeemployment than the general population, and high-

growth and high-expectation entrepreneurs areparticularly likely to do so.


31/50

0

Entrepreneurial activity variables set all categories

o entrepreneurs apart rom the general population.With the exception o recent experience on business

closure, the activity variables also set high-

expectation and high-growth entrepreneurs apartrom other entrepreneurs. High-expectation and

high-growth entrepreneurs are twice as likely asother entrepreneurs to have made business angel

investments14 during the past three years, are more

likely to have known other entrepreneurs during thepast two years, and are more likely to expect to start

a new business within the next three years. There isno notable dierence between high-growth and high-

expectation and other entrepreneurs with regard to

experience on business closure. Interestingly, nascentand new high-expectation entrepreneurs are twice as

likely as established high-growth entrepreneurs tohave shut down a business during the past 12 months.

In addition, attitude variables set both nascent and

new high-expectation entrepreneurs apart rom both

the population o entrepreneurs as well as rom thegeneral population. Nascent and new high-expectation

entrepreneurs are less likely to be inhibited by earo ailure, are more optimistic concerning start-up

opportunities, and are more likely to believe that they

have sucient skills to start a new rm. However,established high-growth entrepreneurs do not dier

markedly rom all established entrepreneurs in

terms o their attitudes. Perception o opportunities,

thus, may have more to do with the initiation o newhigh-expectation businesses than with actual growth,

and it may well be a refection o a certain degree o

overcondence. Or, it may also be that entrepreneursattitudes change as they gather rsthand business

experience.

The distribution o high-expectation and high-growth

entrepreneurs across business sectors providesinsight into the eect o industry structure on high-

growth entrepreneurship. We can observe thathigh-expectation and high-growth entrepreneurs

are underrepresented in the agriculture, shery,

orestry, and hunting sectors (i.e., primary agriculturaloutput). High-expectation new entrepreneurs and

high-growth established entrepreneurs are slightlyoverrepresented in the manuacturing sector and

in transportation, communication, and utilities.Nascent and new high-expectation entrepreneurs

are underrepresented in retail, hotel, and restaurant

sectors, but not so or high-growth establishedentrepreneurs. Interestingly, nascent high-expectation

entrepreneurs are overrepresented in the nancial,insurance, and real-estate sectors, but high-growth

established entrepreneurs are underrepresented in

these sectors. There is also slight underrepresentationo high-expectation and high-growth new and

established entrepreneurs in consumer services.

14 A business angel investment reers to a situation in which a private individual invests his or her own unds to new privately-heldentrepreneurial companies started by others. This activity does not include investments in public share oerings, however.

Prfle High-Expecttin Entrepreners


32/50


33/50


34/50

Prfle High-Expecttin Entreprene

The demographic characteristics o high-expectation

and high-growth entrepreneurs show largely similarpatterns in middle- and low-income countries (GDP

per capita less than USD 20,000 in 2006) as in

high-income countries. As in high-income countries,nascent and new high-expectation entrepreneurs

are overrepresented in the youngest age bracket andunderrepresented in the oldest age brackets. Unlike in

high-income countries, the overrepresentation o high-

growth established entrepreneurs peaks in the agebracket o 35 to 44 years and tapers o toward older

age brackets.

The gender distribution also shows similar patterns in

middle- and low-income countries as in high-incomecountries, although emales are less underrepresented

among entrepreneurs in middle- and low-incomecountries than in high-income countries.

As in high-income countries, entrepreneurs o any

kind are underrepresented in the lowest education

bracket and tend to be overrepresented in highereducation brackets. The distinguishing eect o

higher education appears particularly pronounced orhigh-expectation new entrepreneurs and high-growth

established entrepreneurs.

In middle- and low-income countries, household

income is similarly associated with the incidenceo entrepreneurship as in high-income countries.

Again, both social stratication and wealth-creation

mechanisms may be at play here, with wealth

likely aecting both exposure to opportunities andwillingness to pursue them, which, when successul,enhances the wealth o the high-growth entrepreneur.

Both employment status and entrepreneurial activity

characteristics display similar general trends in

middle- and low-income countries as in high-incomecountries. In middle- and low-income countries, high-

growth established entrepreneurs are not signicantlymore likely than other established entrepreneurs to

report business angel activity. Unlike in high-income

countries, they are more likely than all establishedentrepreneurs to have shut down a business during

the past 12 months. For entrepreneurial attitudes,

as in high-income countries, established high-growthentrepreneurs are not more likely to perceive good

business opportunities than is the general populationin middle- and low-income countries.

suMMaRy

Education and household income, as well asentrepreneurial activities and attitudes, are

signicantly associated with high-expectation andhigh-growth entrepreneurship.


are better educated than other entrepreneurs and

the general population. In high-, middle-, and low-income countries, high-expectation and high-growth

entrepreneurs are signicantly more likely to havegraduate experience than reerence groups.

High-expectation and high-growth entrepreneursare likely to be wealthier than other entrepreneurs

and the general population, in high-, middle-, and

low-income countries.

High-expectation and high-growth entrepreneursare more likely to exhibit entrepreneurial attitudes

and activities than other entrepreneurs andthe general population, with a ew exceptions.

Established high-growth entrepreneurs are notmore likely than other established entrepreneurs

to perceive good opportunities or start-up activity,

whereas nascent and new entrepreneurs are.

The uneven distribution o high-expectation, high-growth, and other entrepreneurs across industry

sectors suggests the eect o industry structural

conditions on entrepreneurial growth. High-expectation and high-growth entrepreneurs are

generally overrepresented in the manuacturing,transportation, communication, and utilities sectors,

but underrepresented in agriculture and consumer

services.


35/50

4

High-Expecttin Entreprenership nd Ntinl Entrepreneril Frmerk

Cnditins

GEMs adult population surveys provide the base

data or the study o country-level entrepreneurialactivity. In addition to adult population surveys, GEM

also collects data describing the national context

o entrepreneurial activity. This data is collectedby means o a questionnaire survey conducted

among some 35 to 50 experts knowledgeable aboutentrepreneurship in the national context. These

survey data are used to compute multi-item scales

describing various aspects o national entrepreneurialramework conditions.15 Two types o ramework

conditions are considered: rst, those that can bemore or less directly infuenced by government policy,

and second, those describing the national culture and

society in general.

The rst group o policy indicators measures theollowing aspects:

Availability o unding or new and growing rms

(including venture capital, debt unding, subsidies,

business angel nance, and capital market nancethrough initial public oerings)

Government policy priorities with regard to

entrepreneurship at national and regional

government levels

Degree o regulatory burden within the economy(e.g., ease o obtaining permits and licenses, scal

burden, and consistency and predictability o

regulatory control)

Availability and quality o support programs andinitiatives or new and growing rms

The degree to which primary and secondary

education support the development o

entrepreneurial skills and initiative

The degree to which post-secondary education andvocational training support the development o

entrepreneurial skills and initiative.

Eciency o technology transer rom research to

entrepreneurial rms, quality o national science

and technology base, and the openness o this baseor access by new and growing rms

Availability and accessibility o proessional and

business services or new and growing rms

Dynamism and change in domestic consumer and

business-to-business markets

Ease o market entry by new and growing rms,

and eectiveness and enorcement o antitrustlegislation

Quality o physical inrastructure relevant

or new business activity (e.g., roads, utilities,

communications inrastructure)

Protection o intellectual property rights

All the indicators above can be more or less directly

infuenced by purposeul government intervention.However, there are also cultural and societal

conditions and valuations in national contex

Documents

GEM High Growth Report 07b