TABLE 3.

Parameter Estimates of Relational Event Models for

	Model 4	Model 5
Variable of theoretical interest:
Long term:
Embedding inertia	.29** (.02)	.26** (.02)
Embedding reciprocation	.19** (.02)
Dependence reciprocation		.02 (.01)
Transitive closure	.24** (.02)	.23** (.02)
Cyclic closure	.21** (.03)	.11** (.03)
Short term
Embedding inertia	.03** (.01)	.03* (.01)
Embedding reciprocation	.03* (.01)
Dependence reciprocation		.03** (.01)
Organizational control:
Size sender	1.68** (.14)	1.50** (.14)
Size difference	−.79** (.15)	−.52** (.14)
Occupancy recipient	−.08 (.09)	−.16 (.09)
Occupancy difference	−.03 (.09)	−.07 (.09)
Inefficiency recipient	−.02 (.04)	.00 (.04)
Inefficiency difference	−.11* (.05)	−.09* (.04)
Geographic distance	−.57** (.06)	−.66** (.06)
Niche overlap	.11** (.03)	.08* (.03)
Service differentiation	−.06 (.04)	.04 (.04)
Same LHU	1.11** (.10)	1.37** (.10)
Both urban	.66** (.08)	.64** (.08)
Sender urban	−2.26** (.25)	−2.48** (.26)
Recipient urban	.07 (.27)	.27 (.27)
Both public	2.47** (.52)	2.49** (.52)
Sender public	−1.52** (.51)	−1.93** (.51)
Recipient public	.13 (.46)	.63 (.46)
In-house capability	−.92** (.04)	−.90** (.04)
Bribery	3.55** (.34)	3.90** (.34)
Sender teaching	.45 (.24)	.74** (.24)
Recipient teaching	.46* (.18)	.41* (.18)
Endogenous structural control:
Long term:
Popularity sender	−.15 (.08)	.01 (.08)
Activity sender	−.03 (.06)	.01 (.06)
Popularity recipient	−.10 (.06)	−.01 (.06)
Activity recipient	−.06 (.05)	−.10* (.05)
Short term:
Popularity sender	−.03 (.05)	.05 (.05)
Activity sender	.08* (.04)	.10** (.04)
Popularity recipient	.03 (.04)	.05 (.04)
Activity recipient	−.03 (.04)	−.03 (.04)
Goodness of fit:
Null deviance	23,732
Residual deviance	13,968	14,066
Observed events (potential events)	1,736 (1,612,744)

Note.—SEs are in parentheses.

^*P < .05.

^**P < .01.

^***P < .001.