Table 3.
Panel regression results of the effect of having PHI on the health state of residents.
| Variables | Pooled model (1) | FE model (2) | RE model (3) |
|---|---|---|---|
| The ownership of PHI | 0.0101 *** (0.0031) | 0.0116 *** (0.0040) | 0.0101 *** (0.0031) |
| Gender | −0.0018 (0.0017) | — | −0.0018(0.0017) |
| Age | −0.0063*** (0.0012) | −0.0403*** (0.0034) | −0.0063*** (0.0012) |
| Marital status | 0.0018 (0.0025) | 0.0039 (0.0030) | 0.0018 (0.0025) |
| Education | 0.0033 (0.0022) | 0.0024 (0.0035) | 0.0033 (0.0022) |
| Registered permanent residence | −0.0078*** (0.0021) | −0.0121*** (0.0032) | −0.0078*** (0.0021) |
| Smoke | −0.0058*** (0.0018) | −0.0319*** (0.0054) | −0.0058*** (0.0018) |
| Drink | 0.0424*** (0.0017) | 0.0434*** (0.0022) | 0.0424*** (0.0018) |
| Social contact | 0.0436 *** (0.0018) | 0.0451*** (0.0021) | 0.0436*** (0.0017) |
| Disease | −0.0622*** (0.0017) | −0.0603 *** (0.0021) | −0.0621*** (0.0017) |
| Household income per capita | 0.0010 *** (0.0003) | 0.0008** (0.0003) | 0.0010*** (0.0003) |
| F test: F | 0.0000* | 195.09*** | 38.68*** |
| Hausman test: χ2 | 194.50*** | ||
| Observations | 44,397 |
Regression results were expressed by coefficients (standard error).
,
and
represented significant at 1%, 5% and 10% levels respectively.
“—”, The binary variable “Gender” was automatically missing in the panel regression.
The following analysis results were presented in the same way.