Table 6.
Second stage results from censored tobit regression
| Explanatory variable | Model 1a | Model 2b | ||
|---|---|---|---|---|
| Censored tobit model CRS | t-value | Censored tobit model VRS | t-value | |
| OPD visits per year (2016) | - 6.78 *10–06 | (-0.57) | -8.98*10–06 | (-0.97) |
| HC distance to closest provincial referral hospital (km) | 0.003 | (0.65) | 0.001 | (0.40) |
| HC distance to closest referral hospital (km) | 0.012 | (1.25) | 0.018* | (2.41) |
| Health centers with beds (binary) | 0.063 | (0.32) | -0.022 | (-0.15) |
| Population coverage by health center (No. of persons) | - 4.36*10–06 | (-0.29) | -2.78*10–06 | (-0.24) |
| Health center size (m2) | 0.002*** | (4.17) | 0.001** | (2.80) |
| Quality Score (%) | -0.009* | (-2.36) | -0.008** | (-2.80) |
| Constant | 0.034 | (0.09) | 0.155 | (0.53) |
| Sigma | 0.30 | (0.02) | 0.21 | (0.04) |
| Observations summary | ||||
| Number of observations | 43 | 43 | ||
| Log likelihood | -17.81 | -8.87 | ||
| LR chi2 | 33.66 (DF = 7) | 27.07 (DF = 7) | ||
| Prob > chi2 | 0.0000 | 0.0003 | ||
| Pseudo R2 | 0.4859 | 0.6042 | ||
| Left-censored observations | 20 (at thetaCRS <= 0) | 25 (at thetaVRS <= 0) | ||
| Uncensored observations | 20 | 18 | ||
| Right-censored observations | 3 (at thetaCRS >= 1) | 0 | ||
Abbreviations: CRS constant returns to scale, HC health center, OPD outpatient department, VRS variable returns to scale
p-value
* p < 0.05
** p < 0.01
*** p < 0.001
a Dependent variable representing CRS inefficiency score
b Dependent variable representing VRS inefficiency score