Table 4. Probit model estimation results for household dataset, Sri Lanka.
| Variable | Estimate | Robust SE | Marginal effect |
|---|---|---|---|
| Constant | -5.5081*** | 0.3007 | |
| Socio-economic and demographic characteristics | |||
| Chro_ill_patients | -0.2545** | 0.1049 | -0.0142 |
| Males_HH | -0.2358*** | 0.0867 | -0.0137 |
| Elders_HH | 0.4850*** | 0.1273 | 0.0281 |
| Pr_hh_working | -0.2691*** | 0.0490 | -0.0156 |
| Male_headed | 0.0952* | 0.0536 | 0.0053 |
| Head_age | -0.0083*** | 0.0017 | -0.0005 |
| Maritalstatus_head | 0.1231** | 0.0559 | 0.0067 |
| Edu_level | -0.0867*** | 0.0045 | -0.0050 |
| Ethnicity | |||
| Sinhala | 3.1586*** | 0.1979 | 0.1374 |
| Sri Lankan Tamil | 3.4641*** | 0.2501 | 0.8294 |
| Indian Tamil | 3.3206*** | 0.2599 | 0.8730 |
| Sri Lankan Moors | 3.3576*** | 0.2873 | 0.8531 |
| Malay | 3.7948*** | 0.4459 | 0.9415 |
| Religion | |||
| Buddhist | 2.5392*** | 0.1765 | 0.1217 |
| Hindu | 2.4259*** | 0.1454 | 0.5263 |
| Islam | 2.3439*** | 0.2160 | 0.5536 |
| Roman Catholic/Other Christian | 2.3536*** | 0.1823 | 0.5693 |
| Health_exp | -3.4029*** | 0.6826 | -0.1971 |
| Head_chronic | -3.4029*** | 0.6826 | -0.1971 |
| Geographical location | |||
| Urban | -0.2470*** | 0.0681 | -0.0122 |
| Colombo | -1.0571*** | 0.1472 | -0.0291 |
| Gampaha | -0.8875*** | 0.1264 | -0.0266 |
| Kalutara | -0.7511*** | 0.1219 | -0.0234 |
| Kandy | -0.4586*** | 0.1087 | -0.0181 |
| Matale | -0.5810*** | 0.1301 | -0.0200 |
| Nuwara_Eliya | -0.6106*** | 0.1271 | -0.0208 |
| Galle | -0.6154*** | 0.1187 | -0.0214 |
| Matara | -0.2921*** | 0.1109 | -0.0131 |
| Hambantota | -0.9533*** | 0.1470 | -0.0249 |
| Jaffna | -0.3730*** | 0.0978 | -0.0154 |
| Mannar | -1.3527*** | 0.2389 | -0.0255 |
| Vavuniya | -1.2931*** | 0.2101 | -0.0254 |
| Batticaloa | -0.4363*** | 0.0953 | -0.0171 |
| Ampara | -0.8530*** | 0.1274 | -0.0239 |
| Trincomalee | -0.3839*** | 0.1117 | -0.0156 |
| Kurunegala | -0.6865*** | 0.1145 | -0.0231 |
| Puttalam | -0.7693*** | 0.1413 | -0.0229 |
| Anuradhapura | -0.5435*** | 0.1263 | -0.0195 |
| Polonnaruwa | -0.8560*** | 0.1560 | -0.0235 |
| Badulla | -0.1808 | 0.1141 | -0.0089 |
| Moneragala | -0.3603*** | 0.1238 | -0.0150 |
| Ratnapura | -0.3331*** | 0.1071 | -0.0144 |
| Kegalle | -0.3087*** | 0.1129 | -0.0136 |
| Type of chronic illness | |||
| High_sev_diseases | 0.2213** | 0.0997 | 0.0139 |
| Brain_diseases | 0.5346*** | 0.0959 | 0.0501 |
| Ent_diseases | 0.2308* | 0.1177 | 0.0165 |
| Otherdiseases | 0.2600** | 0.1038 | 0.0190 |
| Area under ROC curve | 0.7969 | ||
| Pseudo R2 | 0.2437 | ||
| Log likelihood | -3230.4914 | ||
| No. of observation | 20,696 | ||
Source: Author’s calculation based on the DCS [64].
Note
***significant at the 1% level
** significant at the 5% level; significant at the 10% level; Variable ‘Burgher’ was dropped with 26 observations because of estimability; Variables ‘Buddhist’, ‘Kilinochchi’ and ‘Rural’ were removed from the stepwise regression because p> = 0.15.