Research Article
Analysis of Household Consumption Behavior and Indebted Self-Selection Effects: Case Study of Thailand
Table 3
Estimation results of copula-based endogenous switching model.
(a) Estimation of sample selection equation |
| Parameter | Case 1 (all kinds of debt) | Case 2 (housing/land debt) | Frank–Joe | Dynamic Frank–Joe | R–Joe90–Joe | Dynamic R–Joe90–Joe |
| Intercept | −3.7010 | −3.8470 | −7.0430 | −7.2670 | (−20.468) | (−9.783) | (−20.489) | (−25.896) | ln asset | 0.1543 | 0.1478 | 0.2047 | 0.2056 | (23.194) | (19.866) | (14.431) | (14.514) | ln exp | 0.5055 | 0.5557 | 0.2990 | 0.2938 | (18.229) | (13.293) | (6.640) | (6.537) | ln children | 0.0521 | 0.0502 | 0.0028 | 0.0033 | (20.057) | (18.87) | (0.725) | (0.834) | ln income | −0.1782 | −0.2010 | 0.1975 | 0.2232 | (−7.740) | (−8.021) | (5.708) | (6.233) | ln age | −0.3141 | −0.3196 | −0.5420 | −0.5400 | (−9.773) | (−5.525) | (−8.8090) | (−8.981) |
|
|
(b) Estimation of consumption function for households without debt |
| Parameter | Case 1 (all kinds of debt) | Case 2 (housing/land debt) | Frank–Joe | Dynamic Frank–Joe | R–Joe90–Joe | Dynamic R–Joe90–Joe |
| Intercept | 0.9033 | 0.9211 | 0.8981 | 0.8983 | (248.576) | (276.193) | (279.083) | (279.348) | ln ad income | 239.7000 | 203.00 | 299.1000 | 296.1000 | (1752.608) | (112.866) | (1111.661) | (1798.407) | ln saving rate | −0.9526 | −0.9575 | −0.9390 | −0.9387 | (−894.060) | (−769.568) | (−1321.572) | (−1329.810) | ln age | −0.0009 | −0.0009 | −0.0011 | −0.0011 | (−5.793) | (−7.158) | (−8.761) | (−8.794) | ln age 2 | 0.0009 | 0.0009 | 0.0012 | 0.0012 | (6.647) | (7.574) | (9.596) | (9.633) | ln family size | 0.0113 | 0.0096 | 0.0139 | 0.0140 | (28.464) | (27.538) | (46.928) | (47.333) | ln workers | −0.0130 | −0.0116 | −0.0131 | −0.0131 | (−22.412) | (−22.492) | (−29.343) | (−29.293) | ln good workers | 0.0092 | 0.0085 | 0.0065 | 0.0065 | (10.191) | (10.591) | (10.021) | (10.078) | | 0.0437 | 0.0374 | 0.0459 | 0.0606 | (128.787) | (108.801) | (196.144) | (67.739) |
|
|
(c) Estimation of consumption function for indebted households |
| Parameter | Case 1 (all kinds of debt) | Case 2 (housing/land debt) | Frank–Joe | Dynamic Frank–Joe | R–Joe90–Joe | Dynamic R–Joe90–Joe |
| Intercept | 0.8645 | 0.8653 | 0.8405 | 0.8344 | (156.582) | (163.123) | (40.575) | (41.127) | ln ad income | 273.300 | 271.100 | 346.03 | 339.6000 | (984.908) | (102.664) | (196.600) | (519.557) | ln saving rate | −0.9380 | −0.9384 | −0.9250 | −0.9179 | (−897.758) | (−896.498) | (−200.871) | (−197.062) | ln age | −0.0013 | −0.0013 | −0.0076 | −0.0073 | (−5.906) | (−5.994) | (−9.055) | (−9.330) | ln age 2 | 0.0009 | 0.0009 | −0.0023 | −0.0020 | (4.558) | (4.622) | (−2.798) | (−2.481) | ln family size | 0.0167 | 0.0166 | 0.0018 | 0.0015 | (40.483) | (40.153) | (13.212) | (2.005) | ln workers | −0.0072 | −0.0072 | 0.0208 | 0.0171 | (−11.611) | (−11.547) | (14.990) | (13.230) | ln good workers | 0.0056 | 0.0056 | −0.0086 | −0.0090 | (6.981) | (6.961) | (−4.244) | (−4.521) | ln debt ratio | −0.0001 | −0.0001 | 0.0139 | 0.0139 | (−24.850) | (−23.894) | (5.592) | (5.680) | | 0.0504 | 0.0474 | 0.0569 | 0.0549 | (154.600) | (148.663) | (70.121) | (68.3470) |
|
|
(d) Estimation of dependence parameters |
| Parameter | Case 1 (all kinds of debt) | Case 2 (housing/land debt) | Frank–Joe | Dynamic Frank–Joe | R–Joe90–Joe | Dynamic R–Joe90–Joe |
| | - - - | 0.2384 | - - - | 7.9160 | (0.392) | (87.687) | | - - - | −0.2164 | - - - | −0.7761 | (−3.542) | (−60.734) | | −2.265 | - - - | −1.99 | - - - | (−24.838) | (−22.146) | | - - - | 3.412 | - - - | −3.1370 | (1.789) | (−3.692) | | - - - | 0.3242 | - - - | 0.2278 | (−1.701) | (2.765) | | 2.123 | - - - | 1.467 | - - - | (61.076) | (55.174) | log | 67190.74 | 68144.1 | 74992.16 | 75081.88 | AIC | −132169.0 | −136230.2 | −149930.3 | −150105.8 |
|
|
Note. “ ” denotes significance at 1% level. Source. Computation. |