Skip to main content
. 2022 Sep 15;38(5):1119–1143. doi: 10.1007/s10680-022-09642-3

Table 4.

Negative binomial models on migration

Model 1 2 3 4
Log local social capitali

0.466

(0.319)

 − 0.0527

(0.228)
Log bridging social capitalij

2.061***

(0.0473)

2.067***

(0.0473)
Log bonding social capitali

 − 0.667**

(0.252)

0.127

(0.175)
Log populationi

2.724***

(0.270)

2.405***

(0.151)

2.697***

(0.278)

2.301***

(0.164)
Log populationj

2.235***

(0.134)

2.050***

(0.122)

2.224***

(0.132)

2.050***

(0.122)
Log iWiWi

 − 1.333***

(0.462)

 − 2.208***

(0.156)

 − 0.626**

(0.273)

 − 2.124***

(0.371)
Log iWiWj

 − 0.245*

(0.128)

 − 1.784***

(0.116)

 − 0.239*

(0.127)

 − 1.788***

(0.116)
Log distanceij

 − 3.666***

(0.0795)

 − 0.521***

(0.0908)

 − 3.658***

(0.0794)

 − 0.512***

(0.0908)
Pseudo-R2 0.255 0.313 0.256 0.313
N (subregion pairs) 30,448 30,448 30,448 30,448

Coefficients, clustered standard errors in parentheses. Dependent variable: migrationij,t+1

Additional controls: population rate by urbanization, share of selected demographic groups (gender × age categories), subregion characteristics, and same county dummy. Full regression table is displayed in Appendix 2

***p < 0.01, **p < 0.05, *p < 0.1