Table 2.
Regression discontinuity regression results
| 1 | 2 | 3 | 4 | 5 | |
|---|---|---|---|---|---|
| All | [-6m, +6m] | [-3m, +3m] | Parametric | Local linear | |
| Panel A: total outpatient visit | |||||
| Copayment abolition | 0.471*** | 0.404*** | 0.247*** | 0.383*** | 0.322 |
| (0.013) | (0.014) | (0.010) | (0.020) | (0.258) | |
| [60.16%] | [49.78%] | [28.12%] | [46.67%] | [37.99%] | |
| R2 | 0.955 | 0.975 | 0.973 | 0.964 | |
| Panel B: total healthcare cost | |||||
| Copayment abolition | 0.215*** | 0.031 | −0.103*** | 0.021 | −0.045 |
| (0.013) | (0.018) | (0.016) | (0.021) | (0.266) | |
| [23.99%] | [3.15%] | [−9.79%] | [2.12%] | [−4.40%] | |
| R2 | 0.892 | 0.770 | 0.650 | 0.964 | |
| Panel C: healthcare cost per-visit | |||||
| Copayment abolition | −0.255*** | −0.373*** | −0.350*** | −0.363*** | −0.264*** |
| (0.015) | (0.012) | (0.013) | (0.008) | (0.018) | |
| [−22.51%] | [−31.13%] | [−29.53%] | [−30.44%] | [−23.20%] | |
| R2 | 0.653 | 0.984 | 0.984 | 0.964 | |
| Panel C-1: healthcare cost per visit (NHI-sharing) | |||||
| Copayment abolition | −0.233*** | −0.364*** | −0.357*** | −0.352*** | −0.273*** |
| (0.015) | (0.013) | (0.015) | (0.009) | (0.018) | |
| [−20.78%] | [−30.51%] | [−30.02%] | [−29.67%] | [−23.89%] | |
| R2 | 0.567 | 0.982 | 0.981 | 0.964 | |
| Panel C-2: healthcare cost per visit (patient-sharing) | |||||
| Copayment abolition | −0.306*** | −0.392*** | −0.332*** | −0.386*** | −0.246*** |
| (0.014) | (0.011) | (0.011) | (0.010) | (0.021) | |
| [−26.36%] | [−32.43%] | [−28.25%] | [−32.02%] | [−21.81%] | |
| R2 | 0.808 | 0.986 | 0.985 | 0.964 | |
| Observations | 2688 | 192 | 96 | 2688 | 2688 |
Columns 1–3 show the regression results from a fixed effects model. The regression results from a parametric model and a local linear model are presented in Columns 4–5. The average age variable is included as a covariate. *** Significant at the 1% level, ** Significant at the 5% level, * Significant at the 10% level.