Results: Difference-in-Differences Nearest Neighbor Matching
| 2012 | 2013 | 2014 | |
|---|---|---|---|
| Panel A: Mean Difference | −0.0084 | −0.0127 | 0.0261 |
| t-Statistic | −0.4783 | −0.6838 | 1.4050 |
| 95% confidence interval | [−0.043; 0.026] | [−0.049; 0.024] | [−0.010; 0.062] |
| Panel B Variables | |||
| Old | 0.0118 | 0.0446* | −0.0043 |
| (0.0205) | (0.0230) | (0.0251) | |
| [−0.028; 0.052] | [−0.001; 0.090] | [−0.054; 0.045] | |
| Refurbishment (>2008) | 0.0012 | 0.0108 | 0.0285 |
| (0.0246) | (0.0276) | (0.0269) | |
| [−0.047; 0.049] | [−0.043; 0.065] | [−0.024; 0.081] | |
| Observations | 5,419 | 5,151 | 5,280 |
| Adjusted R-squared | 0.22 | 0.25 | 0.22 |
Note: Panel A reports the difference in means across bias-corrected estimates from the first stage and the corresponding t-statistic and 95% confidence intervals for the three different years. Panel B shows the results for the dummy variables “Old” and “Refurbishment (>2008)” while controlling for a set of individual housing characteristics and municipality fixed effects for 2012–2014. Each year is estimated in a separate regression. The standard errors are in parentheses and clustered on municipality level, and the 95% confidence intervals are in square brackets.
↵* p < 0.10.