Table 5.
Regression results and model comparison.
| Dependant variables (Insurance) | Pooled OLS (n = 26,561) | Fixed effects linear (n = 26,561) | Random effects GLS (n = 26,561) | Fixed effects logit (n = 6011) | Fixed effects linear (n = 6011) |
|---|---|---|---|---|---|
| Open degree 1 | 0.0548 (1.24) | 0.0190 (0.40) | 0.0499 (1.13) | 0.352* (1.67) | 0.101 (0.34) |
| Open degree 2 | −0.00471 (−0.11) | −0.0231 (−0.51) | −0.0100 (−0.22) | −0.227 (−0.78) | −0.0256 (−0.14) |
| Collective 1 | −0.0627* (−2.22) | −0.0518 (−1.72) | −0.0605** (−2.13) | −0.546*** (−5.56) | −0.170* (−2.01) |
| Collective 2 | −0.0956** (−2.36) | −0.0798* (−1.89) | −0.0924** (−2.26) | −1.025*** (−5.51) | −0.278* (−2.09) |
| SOE | 0.544*** (8.74) | 0.249*** (9.04) | 0.504*** (8.62) | 2.101*** (7.08) | 0.548*** (3.48) |
| Collective | 0.243*** (6.95) | 0.236*** (8.37) | 0.246*** (7.13) | 1.662*** (5.21) | 0.494** (3.09) |
| Other units | −0.0813*** (−4.00) | −0.0270 (−1.41) | −0.0797*** (−4.02) | −0.472*** (−2.71) | −0.120 (−1.49) |
| Unemployed | −0.114*** (−4.27) | −0.077*** (−4.12) | −0.108*** (−4.52) | −0.705** (−2.48) | −0.166*** (−3.47) |
| SOE*Open Degree 1 | −0.122** (−2.53) | −0.0857** (−2.60) | −0.117** (−2.56) | −1.011*** (−3.98) | −0.240 (−1.54) |
| SOE*Open Degree 2 | −0.201** (−2.65) | −0.00406 (−0.09) | −0.168** (−2.38) | −0.594** (−2.00) | −0.0990 (−0.81) |
| Collective*Open Degree 1 | −0.198*** (−4.41) | −0.157*** (−5.40) | −0.192*** (−4.39) | −0.824*** (−2.89) | −0.274* (−1.97) |
| Collective*Open Degree 2 | −0.0365 (−0.78) | −0.0744* (−2.03) | −0.0470 (−1.02) | −0.698** (−2.35) | −0.212* (−1.98) |
| Reg*Open Degree 1 | −0.0633 | −0.0516* | −0.0614 | −0.673*** | −0.155 |
| (Reg = 1 if urban; = 0 if rural) | (−1.29) | (−2.30) | (−1.39) | (−2.98) | (−0.97) |
| Reg*Open Degree 2 | 0.0154 (0.27) | −0.151** (−2.35) | −0.00220 (−0.04) | −1.235*** (−3.90) | −0.258* (−1.97) |
| Registration status(urban = 1) | 0.145** (2.76) | 0.127** (2.60) | 0.156*** (2.96) | 1.113*** (4.05) | 0.230** (2.39) |
| Age | 0.000675 (0.69) | 0.0116 (0.66) | 0.00123 (1.31) | 0.231* (1.80) | 0.0495 (0.71) |
| Age square | 2.56e−05* (2.24) | −3.77e−05 (−0.77) | 1.89e−05* (1.73) | −0.00073* (−1.77) | −0.00019 (−1.15) |
| Gender(male = 1) | 0.00225 (0.39) | 0.00424 (0.68) | |||
| Number of HH rosters contributed to income | −0.0149*** | 0.00154 | −0.0129*** | 0.0226 | 0.00717 |
| (−3.57) | (0.42) | (−3.60) | (0.51) | (0.41) | |
| Education | 0.0342*** (7.69) | 0.0125* (2.12) | 0.0353*** (8.08) | 0.0722 (1.34) | 0.0245 (1.13) |
| Household income(adjusted) | 0.0532** (2.96) | 0.0143** (2.74) | 0.0463*** (2.94) | 0.246*** (4.34) | 0.0676** (3.12) |
| Temporary health(bad = 1) | 0.0452*** (6.04) | 0.0162 (1.31) | 0.0399*** (5.51) | 0.289* (1.95) | 0.0735 (1.65) |
| Marital Status(married = 1) | 0.0272** (3.29) | 0.0309* (2.28) | 0.0277*** (3.43) | 0.271 (1.63) | 0.0646 (1.81) |
| 1993 | −0.0586 (−1.81) | −0.0665 (−1.14) | −0.0575* (−1.82) | −0.954*** (−3.43) | −0.247 (−1.41) |
| 1997 | −0.0412 (−1.05) | −0.0599 (−0.44) | −0.0386 (−0.97) | −1.131 (−1.37) | −0.232 (−0.48) |
| 2000 | −0.0338 (−0.84) | −0.0966 (−0.52) | −0.0362 (−0.89) | −1.731 | −0.366 (−0.53) |
| Observations | 26,561 | 26,561 | 26,561 | 6011 | 6011 |
| R-squared | 0.46 | 0.08 | 0.17 | ||
| Number of id | 12,264 | 12,264 | 2017 | 2017 | |
Note: (1) Sample size is 26,561 for the first three models (first three columns) and 6011 for the fixed effects logit model. The last column gives the results for the fixed effects model using the smaller sample.
(2) The baseline group for employment group (SOE, Collective, Other Units, and Unemployed) is “others” including retired, housewife, student, and others with no job.
(3)***p < 0.01, ** p < 0.05, * p < 0.1. All are two-tailed tests. The numbers in the parenthesis below the coefficients are t-statistics.
(4) The standard errors are cluster-robust standard errors that cluster on the province.
(5) The above (2)-(4) apply to Table 6 as well.