Skip to main content
. 2016 Feb;151:147–156. doi: 10.1016/j.socscimed.2016.01.013

Table 4.

Determinants of technical efficiency.

Variables Model 1 (No quality) β Model 2 (quality as output) β Model 3 (quality as input) β Model 4 (structural quality -input/process quality – output) β
HIVSTIFAC −0.014 −0.008 −0.0214 0.028
HIVSTI MCH 0.083* 0.050* 0.085** 0.095**
HIVSTICS 0.039 0.009 0.029 0.017
HIVSTIR −0.096** −0.085** −0.078** −0.046**
FUINT −0.004 0.007 0.004 0.020
LN(POP) 0.004 −0.009 0.006 0.003
PROPHIV −0.103 −0145* −0.119 −0.035
PROPCS −0.413** −0.058 −0.345** −0.070
PUBLIC 0.353** 0.161** 0.322*** 0.069
Other facilities 0.190** 0.124** 0.158** 0.119**
RURAL −0.062 −0.164*** −0.045 −0.038
Year2011 0.077* 0.069** 0.073** 0.0456**
Sigma 0.178*** 0.135*** 0.148*** 0.157***
Log-likelihood 25.57 46.42 35.88 34.71
Mean VIF 2.81

Dependent variable: DEA bias-corrected efficiency scores from models 1–4.

∗∗∗, ∗∗, denote significance at 1%, 5% and 10% levels. Confidence intervals obtained from 1000 bootstrap interactions.