High-Growth EntrepreneurshipDavid B. AudretschPrepared for the OECDCopenhagen, March 2012

3/23

/201

2

Research Questions

• What constitutes a “high-growth firm”?• How prevalent are high-growth firms?• What is their (economic) impact?• What are determinants of high growth firms?

• Firm-specific• Locational• What are policy implications?

3/23

/201

2

What Constitutes a High Growth Firm?

• “All enterprises with average annualized growth greater than twenty percent per annum, over a three-year period, and with ten or more employees at the beginning of the observation period. Growth is thus measured by the number of employees and by turnover.”• the OECD-Eurostat Manual on Business

Demography Statistics (2007)

3/23

/201

2

Gazelle Firms

• “All enterprises up to five years old with average annualized growth greater than twenty percent per annum over a three-year period, and with ten or more employees at the beginning of the observation period.” • OECD-Eurostat Manual on Business

Demography Statistics (2007)

3/23

/201

2

Prevalence• Less than 5 percent of firms in U.S. (Birch and

Medoff , 1994)• Between 2-4 percent of firms in U.K. (BERR,

2008)• Less than one percent of enterprises in most

countries (OECD, 2007)• Less than two percent of turnover in most

countries (OECD, 2007)

3/23

/201

2

Economic Impact• Birch and Medoff (1994 )1988-1992, around 70 percent of

all new jobs in the United States created by existing firms (rather than new startups) were accounted for by only four percent of the firms. This same four percent of the firms accounted of 60 percent of all new jobs in the entire U.S. economy.

• U.K. government study finds between two to four percent of all firms account for most of the growth in employment (BERR, 2008)

• Account for high share of employment created in any time period• OECD (2007)

3/23

/201

2

Determinants

• Theoretical Framework• Empirical Evidence• Firm Specific• Locational Specific

3/23

/201

2

Theoretical Framework – Gibrat’s Law• Underlying Assumption: Opportunities are

randomly distributed• Sizeit = (1 +et) Sizeit-1

• Prediction – Firm growth is unpredictable, randomly distributed and not specific to firm or locational characteristics

3/23

/201

2

Framework of Knowledge Spillover Theory of Entrepreneurship• Knowledge created in one organizational context

but not fully commercialized triggers entrepreneurial startups

• Entrepreneurship provides conduit for spillover of knowledge from organization creating knowledge to new firm commercializing it

3/23

/201

2

Framework of Knowledge Spillover Theory of Entrepreneurship

• New & firms account for high share of employment created

• Prediction that high growth should be systematically related to • High knowledge contexts (firm & locational specific)• Negatively related to firm age (firm specific)• Negatively related to firm size (firm specific)• (Contrary to Gibrat’s Law)

3/23

/201

2

Empirical Evidence on Firm Growth• For largest firms, Gibrat’s Law holds

• Not systematically related to firm-specific characteristics of size and age

• For broader distribution of firm size, • Growth rates are higher for younger enterprises

• Growth rates are higher for smaller enterprises• Growth rates are even higher for small and young enterprises in

knowledge-intensive industries• Caves , Journal of Economic Literature (1998)• Sutton, Journal of Economic Literature (1997)

3/23

/201

2

Empirical Evidence• Consistent with Jovanovic’s theory of noisy

selection (1982) and the knowledge spillover theory of entrepreneurship

• Robust across countries• Caves , Journal of Economic Literature (1998)• Sutton, Journal of Economic Literature (1997)

3/23

/201

2

Entrepreneurship & Firm Growth

Incumbent Firm

Performance- Returns- Wages

Survival Trajectory

Failure Trajectory

Time

A

B

B

B

C

D

3/23

/201

2

Temporal Impact of Entrepreneurship on Employment Growth in the United States (Source: Acs and Mueller, 2007)

0.0

0.5

1.0

1.5

2.0

2.5

Impa

ct o

f new

bus

ines

s fo

rmat

ion

on e

mpl

oym

ent c

hang

e

0 1 2 3 4 5 6Lag (year)

3/23

/201

2

Determinants of High-Growth Firms• Firm-Specific Determinants

• High Growth Firms Young• High Growth Firms Small• Birch and Medoff (1994), Henrekson and

Johansson (2010), Storey (1994)

3/23

/201

2

Firm-Specific Determinants

• Henrekson and Johansson (2010, p. 1), “net employment growth rather is generated by a few rapidly growing firms—so-called gazelles—that are not necessarily small and young. Gazelles are found to be outstanding job creators. They create all or a large share of net new jobs. On average, gazelles are younger and smaller than other firms, but it is young age more than small size that is associated with rapid growth.”

3/23

/201

2

Contradictory Evidence• Acs, Parsons and Tracy (2008) • American Corporate Statistical Library (ACSL),

from Corporate Research Board• 1994-2006• Linked to DMI file from Dun & Bradstreet, the

United States Bureau of Labor Statistics’ Industry Occupation Mix, and the PUMS file from the United States Census Bureau

3/23

/201

2

Key Findings of Acs, Parsons & Tracy (2008) • Most high impact firms are small• Large high-impact firms account for most of the

employment creation• High-impact firms are not young (typical high-impact

firm not a startup)• Mean age 25 years old• Survived startup & adolescent phases prior to being

classified as high impact• High-impact firms found in most sectors of economy

3/23

/201

2

Table 1: U.S. Gazelles

Number of Employees Period Number of

GazellesJob Change Revenue Change ($1,000s)

1-19

1994-1998 309,160 3,018,440 $577,533,025

1998-2002 301,275 3,573,918 $716,504,242

2002-2006 283,308 2,883,475 $589,072,471

20-499

1994-1998 43,342 3,014,683 $762,963,829

1998-2002 42,390 3,291,048 $957,923,241

2002-2006 39,617 2,130,682 $1,014,653,361

500-plus

1994-1998 1,547 5,063,517 $1,195,977,664

1998-2002 1,665 4,515,417 $1,841,396,607

2002-2006 1,485 2,514,558 $1,663,635,336

Total

1994-1998 354,049 11,096,640 $2,536,474,518

1998-2002 345,330 11,380,383 $3,515,824,090

2002-2006 324,410 7,528,715 $3,267,361,168

3/23

/201

2

Number of Employees PeriodNumber of High-Impact

FirmsJob Change Revenue Change ($1,000s)

1-19

1994-1998 327,397 3,170,729 $346,038,292

1998-2002 278,190 3,577,111 $423,042,570

2002-2006 359,289 4,041,099 $425,041,975

20-499

1994-1998 23,464 2,788,969 $503,059,203

1998-2002 20,601 2,966,647 $570,102,604

2002-2006 16,523 2,001,835 $549,674,434

500-plus

1994-1998 1,253 5,501,049 $1,110,073,562

1998-2002 1,182 5,192,558 $1,657,759,197

2002-2006 793 2,966,826 $1,060,128,527

Total

1994-1998 352,114 11,460,747 $1,959,171,057

1998-2002 299,973 11,736,316 $2,650,904,371

2002-2006 376,605 9,009,760 $2,034,844,936

3/23

/201

2

1994-1998 1998-2002 2002-2006

Firm Size (No. of

Employees)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Firm Size (No. of

Employees)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Firm Size (No. of Employees)

1-19 20-499500-

plus1-19 20-499

500-

plus1-19 20-499 500-plus

Age of Firm

0-4 2.83 0.67 0.56 4.13 0.9 1.35 5.55 0.89 0.38

5-7 16.72 7.94 4.89 22.42 9.89 9.73 23.26 10.19 6.2

8-10 16.81 11.49 7.94 15.46 11.56 7.7 17.3 13.04 10.63

11-14 17.85 16.82 14.6 15.08 13.92 9.98 14.34 13.82 10.76

15-19 15.22 16.19 13.95 13.75 16.09 15.57 11.95 14.41 13.04

20-24 10.51 11.49 9.22 9.61 11.68 11.68 8.59 12.44 9.75

25-29 6.75 9.13 9.3 6.24 8.43 6.77 6.09 8.62 7.72

30-39 6.62 9.96 11.39 6.54 10.72 10.58 6.74 10.97 10.89

40-49 3.32 6.12 6.82 2.98 5.75 5.33 2.67 5.47 6.96

50-69 2.42 6.31 10.67 2.4 6.3 8.63 2.27 5.46 9.49

70-99 0.95 3.9 10.67 0.94 3.4 7.02 0.86 3.2 7.85

100-plus 0 0 0 0.45 1.36 5.67 0.39 1.48 6.33

3/23

/201

2

Additional Evidence• United Kingdom 2008 study by Department for

Business Enterprise and Regulatory Reform (BERR)

• Broad range of sectors• entrepreneurs & management teams with higher

skill levels & educational attainment• greater propensity to hold intellectual property

and intangible assets, including trademarks

3/23

/201

2

Additional Evidence• Superior access to finance (high prevalence of

venture capital finance)• Cultural context promoting high growth• High social capital component – networks,

partnerships, relationships & linkages to other firms and institutions ( supply chains, formal strategic alliances)• BERR (2008)

3/23

/201

2

Characteristics of Entrepreneur• High level of human capital (education)

• BERR (2008); Baum et al. (2001); Baum &Locke (2004); Vivek et al. (2009)

• Experience as entrepreneur• Baum &Locke (2004)

• Experience as employee in high growth firm• Klepper (2009 ); Agarwal et al. (2004)

3/23

/201

2

Characteristics of Entrepreneur• High levels of experience in industry

• Baum et al. (2001); Baum &Locke (2004)• Gender (male)

• BERR (2008

3/23

/201

2

Characteristics of Founding Team of Entrepreneurs• Size of founding team• Stability of the team members• Time together as a team• Heterogeneity of background• Cohesiveness

• Eisenhardt & Schoonhoven, 1990

3/23

/201

2

Locational Characteristics• No tradition in research & management

• Journal of Economic Literature surveys by Sutton (1997) and Caves (1998)

• Existence of cluster or agglomeration of complementary economic activity & supporting institutions-- Porter (1998)

• Empirical evidence identifying higher growth rates for entrepreneurial startups within a cluster

3/23

/201

2

Empirical Evidence• Empirical evidence identifying higher growth

rates for entrepreneurial startups within a cluster• Gilbert et al. (2006 & 2008); Lechner and

Dowling (2003) • Geographic proximity facilitates accessing and

absorbing localized knowledge spillovers-Jacobs (1969); Jaffe et al. (1993); Audretsch & Feldman (1996)

3/23

/201

2

Localized Spillover Conduits• Worker mobility

• Almeida and Kogut (1999); Saxenian (1990); Lee, Miller, Hancock and Rowen (2000)

• Entrepreneurial startups (Audretsch, Keilbach & Lehmann, 2006)

• Localized networks, linkages & social capital• Saxenian (1990)

3/23

/201

2

Empirical Evidence• Acs, Parsons and Tracy (2008) • High-impact firms found in almost every U.S.

location• City• SMSA• State• Region

3/23

/201

2

Empirical Evidence• Role of Geographic Proximity to Urban Area

• Location with close geographic proximity to urban area important• High impact firms found not only in urban areas• Importance of urban area decreasing over time• No discernible difference in spatial location of high- and low-

impact firms

3/23

/201

2

Table 4a. High-Impact Firm Geographic Location

Distance from Central Business

District (Miles)

1994-1998 1998-2002 2002-2006

Number Percent Number Percent Number Percent

In CBD 36,758 10.48 28,085 9.38 33,249 8.84

1-5 31,771 9.06 27,547 9.20 33,966 9.03

6-10 59,279 16.90 50,357 16.82 63,458 16.88

11-15 35,154 10.02 31,476 10.52 39,269 10.45

16-20 26,307 7.50 23,018 7.69 30,169 8.02

21-25 27,998 7.98 24,197 8.08 30,383 8.08

26-30 15,579 4.44 13,507 4.51 18,014 4.79

31-35 10,377 2.96 9,661 3.23 12,866 3.42

36-40 10,180 2.90 8,941 2.99 11,046 2.94

41 or more 14,432 4.12 15,004 5.01 19,515 5.19

Rural 82,840 23.62 67,549 22.57 84,008 22.35

3/23

/201

2

Table 4b. Low-Impact Firm Geographic Location

Distance from Central Business

District (Miles)

1994-1998 1998-2002 2002-2006

Number Percent Number Percent Number Percent

In CBD 983,126 9.83 1,197,286 8.24 1,345,903 7.92

1-5 879,598 8.79 1,318,135 9.07 1,538,320 9.05

6-10 1,660,875 16.60 2,461,005 16.93 2,921,467 17.19

11-15 984,786 9.85 1,513,943 10.41 1,794,170 10.55

16-20 722,589 7.22 1,122,682 7.72 1,359,973 8.00

21-25 762,361 7.62 1,180,531 8.12 1,373,575 8.08

26-30 438,348 4.38 662,607 4.56 801,096 4.71

31-35 290,937 2.91 443,464 3.05 562,935 3.31

36-40 279,359 2.79 411,190 2.83 483,402 2.84

41 or more 434,649 4.35 714,863 4.92 877,225 5.16

Rural 2,566,109 25.65 3,513,281 24.16 3,941,502 23.19

3/23

/201

2

Policy Implications• Promote entrepreneurship capital

• Audretsch, Lehmann & Keilbach (2006)• Promote access to finance

• Lerner & Gompers (2010)• “There is strong evidence that a heavy regulatory burden

negatively impacts new companies’ into the market and thereby contributes to reduced competitive pressure and less entrepreneurship.”• Swedish Agency for Growth Policy Analysis (2010, p. 8)

3/23

/201

2

Conclusions• High impact entrepreneurship plays key role in

growth & job creation in OECD• Systematic firm-specific characteristics of post-

adolescent & large firms contribute the most to employment growth

• Entrepreneurial characteristics of human capital, experience, access to finance & social capital important

• Policy can facilitate high impact entrepreneurship

3/23

/201

2