This is my master thesis project. I analyze the effect of the 2013 announcement of powerline construction projects on federal (BTW) and state election (LTW) voting outcomes in German municipalities.
For this, I geocode the endpoints of proposed powerlines given in the 2013 Federal Demand Plan Act (BBPLG). I then construct a shapefile which has the (direct) line between the endpoints and examine, whether these lines intersect with municipalities. This then gives me my treatment indictor, which I use to estimate Differences-in-Differences (DiD) Regression models of the effect of powerline construction projects on voting results
- pl_corridors calculates power lines routes and returns a shape file
- pl_treatment calculate the treatment indicator (powerline intersects with municipality)
- btw_/ltw_data_prep prepare the federal and state election data for analysis
- controls_prep prepares a set of control variables
- btw_/ltw_eda provide exploratory analysis for federal and state elections
- pl_eda provides exploratory analysis of the treatment variable
- cs_did provides the DiD Models
Raw data cannot be provided due to large file sizes. To allow replication, I plan to add a script that pulls the data from the Genesis/Regionaldatenbank APIs at a later point.