TABLE 4.

Parameter Estimates of Relational Event Models for Within-Specialty Transfers between 2005 and 2007

	Model 6	Model 7
Variable of theoretical interest:
Long term:
Embedding inertia	.35** (.04)	.37** (.04)
Embedding reciprocation	.03 (.06)
Dependence reciprocation		−.14** (.03)
Transitive closure	.24** (.03)	.22** (.03)
Cyclic closure	−.05 (.06)	−.07 (.06)
Short term:
Embedding inertia	.00 (.02)	.01 (.02)
Embedding reciprocation	.01 (.03)
Dependence reciprocation		−.01 (.02)
Organizational control:
Size sender	1.02** (.28)	.89** (.27)
Size difference	−1.06** (.27)	−1.15** (.27)
Occupancy recipient	.25 (.22)	.14 (.22)
Occupancy difference	.20 (.19)	.07 (.19)
Inefficiency recipient	−.24 (.15)	−.29 (.15)
Inefficiency difference	−.30 (.19)	−.30 (.19)
Geographic distance	−.05 (.10)	−.05 (.10)
Niche overlap	.08 (.08)	.06 (.08)
Service differentiation	−.43** (.07)	−.41** (.07)
Same LHU	.91** (.19)	1.47** (.20)
Both urban	−.79** (.15)	−.82** (.15)
Sender urban	−.38 (.43)	−.042 (.44)
Recipient urban	−1.05* (.47)	−1.09* (.48)
Both public	.61 (1.10)	.78 (1.11)
Sender public	.81 (1.11)	.83 (1.11)
Recipient public	.29 (1.10)	.25 (1.10)
In-house capability	2.01** (.27)	2.04** (.27)
Bribery	2.27** (.56)	1.78** (.56)
Sender teaching	−.64 (.38)	−.62 (.38)
Recipient teaching	.52 (.31)	.55 (.31)
Endogenous structural control:
Long term:
Popularity sender	.22 (.16)	.31* (.16)
Activity sender	.09 (.09)	.13 (.09)
Popularity recipient	.04 (.12)	.03 (.117)
Activity recipient	−.18* (.09)	−.14 (.087)
Short term:
Popularity sender	−.18 (.11)	−.15 (.11)
Activity sender	.16* (.07)	.15* (.07)
Popularity recipient	.06 (.06)	.07 (.07)
Activity recipient	.07 (.07)	.06 (.07)
Goodness of fit:
Null deviance	7,628
Residual deviance	4,337	4,323
Observed events (potential events)	558 (518,382)

Note.—SEs are in parentheses.

^*P < .05.

^**P < .01.

^***P < .001.