Skip to contents

A set of R functions that complement my masters thesis.

There, I analyze the effect of landfills on the nearby population. I vary 2 parameters:

  • inner radius
    defines the dummy variable of interest, living close to a landfill.
  • outer radius
    excludes population that lives too far from any landfill.

With this package, one can create donut_lists: lists of regressions with varying inner and outer radius. There are also functions to interpret the regressions.

Installation

Run the following line in R to install this package:

remotes::install_github("sarahzeller/donutsR")

Functions

  • donut_models creates a donut_list.
  • plot_significance plots the summary for one parameter of interest.
  • donut_summary outputs a modelsummary table, specifically tailored to donut_models.

donut_models takes a data frame with a distance parameter dist_km, which determines the distance to a relevant geometry for all or some rows. It also takes two distance vectors: for the inner radius (inner) and the outer (outer). The function then performs a feols regression with clustered standard errors. It outputs a donut_list, which includes additional list element with the distance parameters. Additional parameters can be used for the following robustness checks:

  • Landfill-cluster-bootstrapped standard errors for the treatment dummy
  • Conley standard errors for all independent variables
  • Excluding the n largest clusters
  • Excluding observations within k distance-units of the landfills

Application

An example of where I used this package can be found here. It is the online appendix of my masters thesis, where I analyse the impact of landfills on residents. The analysis, plots and tables are built with this package.