Skip to main content

Table 2 Number of friends and choice of self-employment, 2SLS Regressions

From: Social-family network and self-employment: evidence from temporary rural–urban migrants in China

 

(1)

(2)

Sample C

Sample D

 

First stage

Endogenous independent variable: Number of friends

Instrumental variable: Log distance when first migrated

Log distance when first migrated

-.5271***

(.1802)

-.5584***

(.1873)

Male

3.4932***

(1.4230)

3.4404**

(1.4714)

Age

-.2555***

(.0862)

-.2499***

(.0787)

Years of schooling

1.5546***

(.2726)

1.5806***

(.2776)

Married

-.5510

(1.9689)

-.4040

(2.0111)

Number of children

.2496

(.6306)

.1399

(.6755)

Years since first migrated out of rural area

.0613

(.1471)

.0557

(.1490)

Constant

Yes

Yes

Home province fixed effects

Yes

Yes

First-stage F statistic

8.56

8.89

 

Second stage

Dependent variable: = 1 if self-employed; = 0 otherwise

Number of friends

.0133**

(.0058)

.0117**

(.0049)

Male

-.0505

(.0331)

-.0435

(.0304)

Age

-.0030

(.0025)

-.0036

(.0022)

Years of schooling

-.0296***

(.0100)

-.0268***

(.0090)

Married

.1608***

(.0383)

.1577***

(.0362)

Number of children

.0318*

(.0177)

.0305**

(.0151)

Years since first migrated out of rural area

.0083***

(.0028)

.0071***

(.0024)

Constant

Yes

Yes

Home province fixed effects

Yes

Yes

No. of observations

2,379

2,336

OLS coefficient of “number of friends”

.00039*

(.00024)

.00041*

(.00023)

  1. Outliers above the 99th percentile (with more than 200 total contacts) are excluded from the regressions. Sample C includes all household heads aged between 16 and 70 years who initially were not self-employed and who are not in their first jobs in urban sectors. Sample D only includes all household heads aged between 16 and 70 years who initially were not self-employed and who are not in their first jobs in urban sectors, excluding those who have chosen self-employment because they could not find wage-earning jobs.
  2. Standard errors in parentheses are robust to heteroskedasticity and clustered at the city level.
  3. ***statistically significant at the 1 percent level; **statistically significant at the 5 percent level; *statistically significant at the 10 percent level.