Urban Water Conservation Policies in the United States

This repository contains scripts for reproducing the analysis in J.M. Gilligan et al., "Urban Water Conservation Policies in the United States."

Pre-requisites

To run the replications, you will need to install R (version 3.4.0 or above), RStudio, LaTeX (preferably the TeXLive or MikTeX distributions).

You can obtain R for Windows, Linux, MacOS, and other operating systems from https://www.r-project.org or https://cran.rstudio.com/ (get R version 3.4.0 or later). You can obtain RStudio for Windows, Linux, or MacOS from https://www.rstudio.com. You can obtain a LaTeX distribution from https://ctan.org/starter. If you are running Windows, you will also need to install RTools, which you can obtain at https://cran.rstudio.com/bin/windows/Rtools/ (get RTools version 3.5 or later).

and the following R packages:

• devtools
• extrafont
• fontcm
• geosphere
• ggthemes
• gridExtra
• hms
• janitor
• knitr
• loo
• maps
• ncdf4
• pacman
• parallel
• raster
• reshape2
• rstan
• tidyverse
• RColorBrewer
• viridis
• xtable
• zoo

You can install these packages by running the following command from R or from the R console in RStudio:

install.packages(c("devtools", "extrafont", "fontcm", "geosphere", "ggthemes", "gridExtra", "hms", "janitor", "knitr", 
                   "loo", "maps", "ncdf4", "pacman", "parallel", "raster", "reshape2", "rstan", "tidyverse", 
                   "RColorBrewer", "viridis", "xtable", "zoo"))

You will also need to install the following packages from github (e.g., using the devtools::install_github function):

• dkahle/gmap
• jonathan-g/jgmcmc@jgmcmc
• jonathan-g/jgally@jgally
• slowkow/ggrepel

You can install these packages using the following commands from R or the R Console in RStudio:

devtools::install_github("dkahle/gmap")
devtools::install_github("jonathan-g/jgmcmc@jgmcmc")
devtools::install_github("jonathan-g/jgally@jgally")
devtools::install_github("slowkow/ggrepel")

Performing Analyses and Generating Manuscript:

Once you have everything installed, open RStudio and set the global options from the "Tools/Global Options" menu, go to "Sweave", and choose "Weave Rnw files using" to "knitr" and "Typeset LaTeX into PDF using" to "pdfLaTeX".

Next, open the file urban_water_conservation_policies.Rnw in RStudio. You can set the relevant processing options in this file (see below) and then click on the "Compile PDF" button in the editing pane or use the "File/Compile PDF" menu option. You can also compile from the R Console by typing knit("urban_water_conservation_policies.Rnw", envir=new.env())

Similarly, you can generate the Supporting Information by opening urban_water_conservation_policies_si.Rnw in RStudio and compiling to PDF using the "Compile PDF" button, the "File/Compile PDF" menu option, or by typing knit("urban_water_conservation_policies_si.Rnw", envir=new.env()) at the R Console.

Processing Options

You can control which parts of the analysis you reproduce by changing the variables at the beginning of the file urban_water_conservation_policies.Rnw (legal values are TRUE or FALSE:

• do_cache: Turn on knitr's data-caching mechanism. This variable should be set to FALSE. If you set it to TRUE, this may speed up subsequent runs if you are editing and re-running code in the scripts or urban_water_conservation_policies.Rnw, but my experience with this is that caching introduces a lot of potential for errors and doesn't speed things up noticeably if you are fitting the large test_models and pooled_test_models data sets.
• download_data_sets: Download fresh copies of the data sets (from the supporting information) from https://doi.org/10.6084/m9.figshare.5714944. These data sets contain the VWCI scores and all the covariates necessary to perform the regression analyses described in the paper.
• recalc_pvi: Recalculate the PVI scores for cities and states. This requires you to obtain the raw state and county-level election returns for the 2008 and 2012 presidential elections (available from CQ Press, as cited in the paper). I downloaded the returns in groups, by state:
  • 2008_al_fl.csv and 2012_al_fl.csv for Alabama through Florida
  • 2008_ga_ky.csv and 2012_ga_ky.csv for Georgia through Kentucky
  • 2008_la_ne.csv and 2012_la_ne.csv for Louisiana through Nebraska
  • 2008_nv_nm.csv and 2012_nv_nm.csv for Nevada through New Mexico
  • 2008_ny_sd.csv and 2012_ny_sd.csv for New York through South Dakota
  • 2008_tn_wy.csv and 2012_tn_wy.csv for Tennessee through Wyoming These files should reside in the data/proprietary/ directory.
• recalc_explanatory: Build a data frame with the following explanatory variables
  • population
  • population growth rate
  • surface water use as a fraction of public water supply
  • Democratic share of the two-party vote in the 2008 and 2012 presidential elections.
  • National Democratic vote share.
  • Partisan Voting Index (PVI)
  • Population density
  • Population density growth rate
  • MSA area
• recalc_vwci: Assign FIPS codes to MSA-level VWCI scores (for geolocation).
• recalc_city_coords: Recalculate the geographic location of the MSAs.
• recalc_climate: Recalculate the average temperature, precipitation, and Köppen aridity index for cities and states.
• recalc_msa_variables: Assemble a data-frame with MSA-level variables and covariates.
• recalc_standardized_data: Transform raw MSA- and state-level variables into standardized forms, as described in the paper.
• resample_models: Run the regression analysis for VWCI, Requirements, and Rebates using the baseline regression model described in the paper (hierarchical beta-binomial, with varying intercept).
• resample_test_models: Run the alternate regresion models described in the "Robustness Checks" sections of the paper
  and supporting information.
• resample_pooled_test_models: Run the baseline and alternate regression models using the alternate normalization of MSA-level covariates (normalized relative to all MSAs, as opposed to representing the difference between the MSA and the state).
• force_resampling: Force re-running all regression models that are selected by resample_models, resample_test_models, and resample_pooled_test_models. This is only necessary if do_cache is set to TRUE, which I do not recommend; if do_cache is set to FALSE, this has no effect and all selected models are automatically re-sampled.