The Unintended Consequences of Infrastructure Development Antonella Bancalari

The start of the implementation phase is set as the year in which the first project in a given district was started, and the completion is set as the year in which all sewerage projects in a given district had been completed. [...] The rate of completion of the implementation phase was low, with only 13% of treated municipalities completing all infrastructure works by the end of the study period, the highest percentage being in 2015 (see Table C1 in the appendix). [...] The weights on the 2×2 DDs are proportional to timing group sizes and the variance of the treatment dummy in each pair, which is highest for units treated in the middle of the panel (Goodman-Bacon, 2021). [...] The correction is not pronounced because the problematic 2× 2 DD estimates (when using ‘always treated’ and ‘already treated’ as the control group) were assigned low weights in the traditional TWFE (see Figure D3 in the appendix).7 7The low weight of the 2×2 DD estimates when using ‘always treated’ is due to only 6.84% of treated districts initiating works at the beginning of the analysis, as show. [...] The point estimates of the remaining lags are either the same as or slightly lower than the OLS estimates, likely due to the differences in how weights are determined (weights in OLS are proportional to timing group sizes and the variance of the treatment dummy, while in CS they only depend on timing group sizes).8 ii Adding controls Tables D4 and D5 in the appendix present a variety of specificat.