Skip to main content
. 2021 Apr 26;79:59. doi: 10.1186/s13690-021-00582-8

Table 5.

The influence of late retirement on difficulty in physical functioning and problems of cognitive status (Heckman two-stage regression)

Dependent variables
Difficulty in physical functioning (2018) Problems of cognitive status (2018)
Coef. S.E. 95% CI Coef. S.E. 95% CI
Late retirement status of 2015 −0.353** 0.023 [−0.399, − 0.308] −0.069 ** 0.026 [−0.119, − 0.018]
Social behavior − 0.130 ** 0.020 [− 0.169, − 0.091] − 0.084 ** 0.023 [− 0.129, − 0.039]
Health insurance − 0.045 0.042 [− 0.127, 0.037] 0.121 ** 0.043 [0.036, 0.205]
Gender
Male REF. REF.
Female −0.317 ** 0.060 [−0.435, − 0.199] −0.207 ** 0.059 [−0.323, − 0.091]
Age − 0.047 ** 0.007 [− 0.060, − 0.033] − 0.011 0.007 [− 0.025, 0.002]
Education
Less than lower secondary REF. REF.
Upper, secondary & vocational training −0.194 ** 0.039 [− 0.271, − 0.116] −0.077 0.049 [−0.174, 0.020]
Tertiary −0.284 ** 0.084 [− 0.448, − 0.120] −0.095 0.111 [−0.313, 0.123]
Marriage
Married REF. REF.
Partnered −0.030 0.046 [−0.120, 0.061] − 0.015 0.059 [−0.131, 0.100]
Separated 0.205 0.229 [−0.244, 0.654] −0.036 0.185 [−0.399, 0.327]
Divorced −0.030 0.128 [−0.282, 0.221] 0.046 0.177 [−0.301, 0.392]
Widowed 0.011 0.031 [−0.049, 0.071] − 0.002 0.034 [−0.068, 0.064]
Never married 0.017 0.099 [−0.177, 0.211] 0.002 0.156 [−0.305, 0.308]
Residence
Central city/town REF. REF.
Urban-rural integration zone 0.137 ** 0.049 [0.041, 0.234] 0.025 0.057 [−0.086, 0.136]
Rural 0.269 ** 0.030 [0.211, 0.328] 0.161 ** 0.034 [0.094, 0.228]
Special zone 0.096 0.208 [−0.311, 0.503] −0.086 0.184 [−0.446, 0.274]
Inverse mills ratio 1.061 ** 0.104 [0.856, 1.265] 0.384 ** 0.102 [0.184, 0.584]
Intercept 4.353 ** 0.435 [3.500, 5.207] 4.010 ** 0.428 [3.172, 4.849]
Observations 3757 3223
F-statistics 69.27 10.44
[P-value] [0.000] [0.000]

Notes: The Heckman two-stage regression (1st stage) is conducted by regressing late retirement (2015) on self-reported health (2015), health care services (2015), age and gender, according to which the Inverse Mills Ratio (IMR) is obtained and applied in the 2nd stage regression.

*p < 0.05, **p < 0.01