Replication: Are plastic bag regulations effective in reducing plastic litter? Evidence from shoreline cleanups

by Anna Papp and Kimberly L. Oremus

ReadMe last modified: May 7th, 2024

Code last modified: April 15th, 2024

Abstract

Plastic pollution poses threats to marine ecosystems and ecosystem services. While plastic bag bans and taxes are increasingly implemented worldwide, their effectiveness in reducing plastic litter remains unknown. Leveraging the patchwork of bag policies across different geographic scales in the United States and citizen science data on 97,774 U.S. shoreline cleanups, we assess the impact of these policies on plastic bag litter. We find that plastic bag policies lead to a 22-25% decrease in plastic bags as a share of total items collected at cleanups, with larger-scale bag bans and taxes further reducing shoreline litter. At a time when many jurisdictions are considering bag policies while others are preemptively prohibiting them, our study provides evidence that these policies can reduce shoreline plastic pollution.

---

Data

Download the following data:

• Policies: Compiled from Bag the Ban, the Retail Compliance Center, BagLaws.com, the Duke Nicholas Institute's Plastics Policy Inventory, and Wikipedia; and using MassGreen and Californians Against Waste to confirm legislation in Massachusetts and California.
  • Saved in data/policies directory, provided in directory.
• TIDES: Downloaded cleanup data for the United States from Ocean Conservancy website.
  • Download files for 2000-2009, 2010-2014, and then each separate year from 2015 until 2023.
  • Save files in the data/tides directory, as year.csv (and 2000-2009.csv, 2010-2014.csv)
• Shapefiles: Download shapefiles for processing cleanups and policies.
  • Download county shapefiles from here from the US Census Bureau; save files in the data/shapefiles directory, county shapefile should be in folder called county (files called cb_2018_us_county_500k.shp).
  • Download TIGER Zip Code tabulation areas from here from the US Census Bureau (through data.gov; save files in the data/shapefiles directory, zip codes shapefile folder and files should be called tl_2019_us_zcta510.
  • Download lakes from Natural Earth (North America supplement) and save in lakes folder in data/shapefiles directory.
  • Download rivers from USGS and save in rivers folder in data/shapefiles directory.
• Other: Helper files with US county and state fips codes, lists of US counties and zip codes in data/other directory, provided in the directory except as noted below.
  • Download zip code list and 2020 IRS population data from United States zip codes here and save as uszipcodes.csv in data/other directory.
  • Download demographic characteristics of zip codes from Social Explorer and save as raw_zip_characteristics.csv in data/other directory.

Refer to the .txt files in each data folder to ensure all necessary files are downloaded.

---

Code

Scripts to run entire pipeline:

• _00-02_create_data.R: This script runs all of Steps 0 - 2 and creates the final data used in the analysis (it requires all of the data outlined above).
  • Please see Section Data Code Details below for more details on file requirements and steps.
  • All final data (various spatio-temporal aggregations) should be saved in the data/processed/02_data_merged/ folder.
• _03_04a_main_figures.R: This script creates all the main figures.
  • See below for each separate file.
  • All outputs are saved in the figures/ and tables/ folders.
• _03_04b_appendix_figures_tables.R: This script creates all the appendix figures and tables.
  • See below for each separate file.
  • All outputs are saved in the figures/appendix/ and tables/appendix/ folders.

Main analysis files:

3. Step 03: 03_summary: This folder creates summary statistic figures and tables for both main text and appendix.
   • Main Figures:
     • Figure 1: 03a_figure1_data_summary.R (data summary)
     • Figure 3 (part a): 03b_figure3_spillover_demo.R - Note that this requires lake and river shapefiles (see above).
   • Appendix Figures and Tables:
     • Appendix Table S1: 03z_appendix_tabs01_cleanup_stats.R (cleanup summary statistics)
     • Appendix Figure S2 and Table S2: 03z_appendix_figs02_tabs02_repeat_cleanup.R (repeat cleanups)
     • Appendix Figure S3: 03z_appendix_figs03_policy_map.R (map of policies in the US)
     • Appendix Figure S4: 03z_appendix_figs04_cleanup_objects.R (chart with most common objects found in cleanups)
     • Appendix Figure S5: 03z_appendix_figs05_cleanup_year.R (chart with the number of cleanups per year)
4. Step 04: 04_analysis: This folder conducts the main regression analyses and robustness checks.
   • Main Figures and Tables:
     • Figure 2: 04a_figure2_main_results.R (main results and placebo outcomes)
     • Figure 3 (part b): 04b_figure3_spillover.R (spillover results)
     • Figure 4: 04c_figure4_policy_type.R (analysis by policy type)
   • Appendix Figures:
     • Appendix Figure S6 and Appendix Table S4: 04z_appendix_figs06_tabs04_other.R (other outcomes)
     • Appendix Figure S7: 04z_appendix_figs07_time_aggregation.R (robustness to various temporal aggregations of the data)
     • Appendix Figure S8: 04z_appendix_figs08_spatial_aggregation.R (robustness to various spatial aggregations of the data)
     • Appendix Figure S9 and S10: 04z_appendix_figs09_figs10_balances.R (robustness to using balanced panel)
     • Appendix Figure S11: 04z_appendix_figs11_pandemic.R (robustness to dropping the pandemic)
     • Appendix Figure S12: 04z_appendix_figs12_cleanuplocation.R (coastal vs. inland results)
     • Appendix Figure S13: 04z_appendix_figs13_state_details.R (state results)

---

Data Code Details

Step 0: 00_cleanup: Compile cleanup data, match to geographic boundaries, and then aggregate to various spatio-temporal levels.

a. 00a_cleanup_data_country.R: compile cleanups and match to county

Details:

• requires:
  • data/shapefiles/county/cb_2018_us_county_500k.shp (county shapefile)
  • data/other/us-state-ansi-fips.csv
  • data/other/statefips.csv (state fips codes)
  • data/other/us-county-ansi-fips.csv (county fips codes)
• produces:
  • data/processed/00_data_cleanup_county.rda

b. 00b_cleanup_data_zip.R: match cleanups to zip code

Details:

• requires:
  • data/processed/00_data_cleanup_country.rda (from previous step)
  • data/shapefiles/tl_2019_us_zcta510/tl_2019_us_zcta510.shp (zip code tabulation area shapefile)
• produces:
  • data/processed/00_data_cleanup_county_zip.rda

c. 00c_cleanup_data_cell.R: create 0.1/0.01/0.001 degree cells

Details:

• requires:
  • data/processed/00_data_cleanup_county_zip.rda (from previous step)
• produces:
  • data/processed/00_data_cleanup_county_zip_cell.rda

d. 00d_cleanup_data_clean.R: clean (normalized variables, outliers, etc.)

Details:

• requires:
  • data/processed/00_dat_cleanup_county_zip_cell.rda (from previous step)
• produces:
  • data/processed/00_data_cleanup.rda
  • data/processed/00_data_cleanup_locations.csv (for Google Earth Engine distance calculations)

e. 00e_cleanup_distance.R: calculate the distance to water bodies

• requires:
  • data/processed/00_data_cleanup.rda (from previous step)
  • data/other/distanceCoast.csv (to replicate, use GEE script)
  • data/other/distanceRivers.csv (to replicate, use GEE script)
  • data/other/distanceLakes.csv (to replicate, use GEE script)
• produces:
  • data/processed/00_data_cleanup.rda

f. 00f_cleanup_final.R: aggregate to various levels (e.g., cell, month)

• requires:
  • data/processed/00_data_cleanup.rda (from previous step)
• produces:
  • data/processed/00_data_intermediate/... (separate files for each aggregation)

Step 1: 01_policy: Compile bag policies and match to zip codes.

a. 01a_policy_county_zip.R: aggregate policies

• requires:
  • data/other/uscounties.csv (list of US counties)
  • data/other/uszipcodes.csv (list of US zip codes)
  • data/shapefiles/tl_2019_us_zcta510/tl_2019_us_zcta510.shp (zip code tabulation area shapefile)
  • data/policies/policies.xlsx (list of bag policies)
• produces:
  • data/processed/01_zip_policy.rda
  • data/processed/01_county_policy.rda

b. 01b_zip_characteristics.R: add demographic characteristics to zip codes

• requires:
  • data/processed/01_zip_policy.rda (from previous step)
  • data/other/raw_zip_characteristics.csv (downloaded from Social Explorer)
• produces:
  • data/processed/01_zip_policy_characteristics.rda

c. 01c_neighbor_zip.R: find all the neighbors of each zip code

• requires:
  • data/shapefiles/tl_2019_us_zcta510/tl_2019_us_zcta510.shp (zip code tabulation area shapefile)
• produces:
  • data/processed/01_zip_neighbors_list.rda

d. 01d_neighbor_zip_policy.R: find neighbors, neighbors-of-neighbors, etc. of treated zip codes

• requires:
  • data/processed/01_zip_neighbors_list.rda (from previous step)
  • data/processed/01_zip_policy.rda (policy data, from step 1a.)
• produces:
  • data/processed/01_zip_neighbors_policy.rda

Step 2: 02_merge: Merge cleanup and policy data

a. 01a_merge.R: merge all policies (helper script)

• requires:
  • data/processed/01_zip_policy_characteristics.rda (from step 1b)
  • data/processed/01_zip_neighbors_policy.rda (from step 1d)
  • data/processed/01_county_policy.rda (from step 1a)
  • data/processed/00_data_intermediate/... (separate file for each aggregation, from step 0f)
• produces:
  • data/processed/02_data_merged/... (separate file for each aggregation)