Table 6.
Difference-in-differences estimates for “placebo” reform effect on ACSCs hospital admission: construct artificial/placebo treatment municipalities and placebo reform year (cluster standard errors in municipalities are in the parentheses)
| Variable | Placebo reform effecta | Artificial/placebo treatment municipalities bReform = 1 (year = 2012 and 2013) Reform = 0 (year = 2010 and 2011) |
Artificial/placebo treatment municipalities bReform = 1 (year = 2013) Reform = 0 (year = 2010 and 2011) |
|---|---|---|---|
| Placebo_Reform | − 0.0438*** (0.110) | – | – |
| Treat | − 2.743*** (0.295) | – | – |
| Placebo_Reform × Treat | − 0.0072 (0.0247) | – | – |
| Reform | – | − 0.0008 (0.0121) | – |
| Placebo_Treat | – | 0.7305*** (0.0195) | – |
| Reform × Placebo_Treat | – | − 0.0040 (0.0154) | – |
| Reform1c | – | – | − 0.0137 (0.0127) |
| Placebo_Treat | – | – | − 1.044*** (0.0547) |
| Reform1 × Placebo_Treat | – | – | − 0.0037 (0.0175) |
| Number of observation | 48,652 | 98,976 | 74,015 |
| R-squared | 0.34 | 0.36 | 0.36 |
| Year fixed-effects | Yes | Yes | Yes |
| Municipality fixed-effects | Yes | Yes | Yes |
All the models are also control for the variables included in Table 4. DID coefficients (i.e. Reform × Treat/Reform × Placebo_Treat/Reform1 × Placebo_Treat) tested against one-sided alternatives
‘*’, ‘**’ and ‘***’ represents significance level at the 10%, 5% and 1% level respectively
aPlacebo reform = 1, if year = 2011 and Placebo reform = 0, if year = 2010
bThe Placebo treatment municipalities are created arbitrarily/randomly using Norwegian county numbers (see the map in Appendix). For example, municipalities belongs to the first three counties (1–3) are considered to be in the treatment municipality and the next 3 (4–6) considered as control municipality, and doing the same procedure for the rest of the counties
cReform1 = 1 if after Reform includes year 2013; Reform13 = 0 if before Reform includes year 2010–2011