entrepreneurship and social...

DIMETIC Session October 9, 2008

Prof. Olav SorensonUniversity of Toronto

Entrepreneurship and social networks

1

Four studies

• Danish residents (1995-2001)

• U.S. Footwear manufacturers (1940-1989)• Sorenson & Audia, American Journal of Sociology, 2000

• U.S. Computer workstation manufacturers (1980-1996)

• U.S. Dedicated biotechnology firms (1978-1996)

• Stuart & Sorenson, Research Policy, 2003

2

Exp. 1: Footwear manufacturing 1940

MA-281

NY-264

MO-69

CA-13

IL-68

NH-69

PA-103

WI-56ME-60

Source: Sorenson & Audia, American Journal of Sociology, 2000

3

Footwear manufacturing 1989

CA-77

MA-51

ME-40

MO-48

NY-55

PA-67

WI-25

TX-36


4

Typical explanations

• Transportation costs

• Agglomeration externalities

• Pooled labor

• Shared suppliers

• Information spillovers

• Implication: Efficiency of spatial distribution

5

What do social networks do for entrepreneurs?

• Provide access to information

• Opportunity identification aided by access to private information which travels through social networks

• Firm building aided by access to accumulated tacit knowledge which travels through social networks

• Provide access to critical resources

• Resource mobilization facilitated by information transfer, the ability to monitor and sanction, and trust – all of which depend on social networks

• Offer social support

6

Social networks facilitate information access

• Burt & Raider (2002): Structurally diverse networks positively related to entrepreneurship

• Rezulli, Aldrich & Moody (Social Forces, 2003): Those at intersecting positions in different networks found firms at higher rates

• Elfring & Hulsink (Small Business Economics, 2003): Weak ties facilitate opportunity identification

7

Social networks facilitate resource access

• Financial capital: Fried & Hisrich (Financial Management, 1994), Hsu (Journal of Finance, 2004), Sorenson & Waguespack (Administrative Science Quarterly, forthcoming)

• Human capital

• Intellectual capital: Singh (Management Science, 2004), Sorenson, Rivkin & Fleming (Research Policy, 2006)

8

How do networks relate to geography?

• Social ties concentrate in space

• Both in physical space and social space (e.g., industry)

• Space shapes opportunities for contact, as well as the costs of maintaining a tie

• …leading industries to cluster

• Entrepreneurs most aware of opportunities and best able to formulate and execute plans for new businesses where they live in the industries in which they work

9

Evidence from the individual level

10

• IDA + supplemental information on entrepreneurs

• Spatial unit: 275 kommunes

Source: Dahl & Sorenson, “Home sweet home,” 2007

11

8,400 observations

Inexperienced Experienced

KommuneIndustry

12

• Conditional logit (McFadden choice model)

• One location actually chosen

• Compared to characteristics of 274 locations not chosen

• Control for founder characteristics

Location choice estimates

13

Entrepreneurs do not often move

Average effects Average effects Anchors Sirens

Home 0.914 0.388

Work 3.43 2.46

Distance to home -1.45 -1.47

Region tenure 0.252 0.247

Ln (rivals) 1.02 0.987 0.944

Population -0.231 -0.561 -0.015


14

• Discrete time probability of failure

• Include controls for firm heterogeneity

• Age - number of years since founding (splined)

• Limited liability - corporate form

• Ln (size) - ln of number of employees

• Includes controls for entrepreneur experience in industry

• X home - potential test for industry-specific social capital

Firm performance estimates

15

Firms founded by entrepreneurs with experience in the region survive longer

Home 0.176 0.130 0.189

Work -0.114 -0.069 -0.085

Distance to home 0.043 0.048 0.051

Region tenure -0.083 -0.064 -0.064

Ln (rivals) 0.066 0.066

Spin out -0.507 -0.389

Spin out X home -0.184


16

Firms founded by entrepreneurs with experience in the region recruit better employees and grow faster

No tenure Tenure < 6yrs Tenure > 5 years

Profits (t = 1) 33K DKK 84K DKK 102K DKK

Profits (t = 3) -125K DKK 60K DKK 84K DKK

Size (t = 1) 3.5 2.8 2.8

Size (t = 3) 6 4.9 2.5

% Local 16% 26% 32%

% Industry 18% 18% 20%


17

Evidence from the industry level

18

Footwear manufacturing 1989

CA-77

MA-51

ME-40

MO-48

NY-55

PA-67

WI-25

TX-36


19

Weighted density variables

• i is focal location

• j are all other location

• x is a vector of location characteristics

• Number of firms

• Exp.: size

• d is a vector of distances between i and j

• reduces to (usual) density when d = 0 for all ij pairs

20

Weighted density variables – Construction example

d = 10

d = 11

AB

d = 3

For location A, local density = 1.083 =

1 + 0 (0/11) + .083 (1/12)

For location B, local density = .341 =

0 + .091 (1/11) + .25 (1/4)

For location C, local density = 1.083 =

1 + 0 (0/4) + .083 (1/12)

C

21

Entry, exit and concentration

0

1

2

3

4

5

6

7

0 1 2 3 4 5 6 7 8 9

Entry rate

Exit rate

Mul

tiplie

r of

bas

e ra

te

Local density of shoe manufacturersSource: Sorenson & Audia, American Journal of Sociology, 2000

22

Time to equilibrium

1956

1956

1956

1963

1964

2047

Exit with random entry

Exit with no entry

Exit with structured entry

23

Exp. 2: Computer workstations in 1984

Source: Sorenson, 2004

24

Computer workstations in 1996


25

Local density and entry in workstations

0

5

10

15

20

25

30

35

40

0 10 20 30 40 50 60


Entry rate

Exit rateMul

tiplie

r of

bas

e ra

te

Local density of workstation manufacturers

26

Exp. 3: Biotechnology

Source: Stuart & Sorenson, Research Policy, 2003

27

Relative resource importance

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

BT VC FirmsVC Firms Patents Universities

Source: Stuart & Sorenson, Research Policy, 2003

28

Fig. 1.—The small world of markets and organizationsSource: Burt, American Journal of Sociology, 2004

29

entrepreneurship and social...

Documents