Break Free From Plastic Brand Audit Analysis 2019-2020
Matthew
Abstract
This report reviews an analysis of the Break Free From Plastic brand audit data for 2019–2020. Using R, we load the public dataset, perform data tidying operations, and use ggplot2 to visualize and study patterns in global plastic pollution. The findings demonstrate how a small set of corporate actors dominate labeled waste counts and underscore the persistence of PET plastics (polyethylene terephthalate) packaging in community cleanups worldwide. This analysis provides evidence for targeted policy interventions and enhanced corporate accountability measures.
Introduction
Plastic pollution imposes significant ecological and social costs on communities worldwide. Mismanaged plastics accumulate in waterways, soils, and food chains, contributing to long-term environmental degradation and rising cleanup costs. Brand audits conducted by volunteers provide visibility into the origins of collected plastic items.
The Break Free From Plastic (BFFP) global brand audit dataset is a publicly available databases linking waste items directly to corporate producers. By examining patterns in this dataset, we can better understand which demographics are most heavily represented in environmental litter. This evidence-based approach informs policy interventions, supports accountability mechanisms, and provides data to justify extended producer responsibility initiatives.
This analysis focuses on identifying the most problematic plastic types, the companies responsible for the highest volumes of branded waste, and geographic patterns in plastic pollution across 2019 and 2020.
Methodology
The analysis begins by importing the publicly hosted 2019–2020 BFFP dataset through GitHub’s TidyTuesday repository. We use R’s tidyverse to transform unstructured data into structured, data suitable for examination and visualization.
The raw dataset contains information on plastic waste collected during brand audits, including the country of collection, year, parent company responsible, and counts by plastic type (PET, HDPE, LDPE, PP, PS, PVC, and others).
## # A tibble: 133,800 × 6
## country year parent_company volunteers plastic_type count
## <chr> <dbl> <chr> <dbl> <chr> <dbl>
## 1 Argentina 2019 Grand Total 243 empty 0
## 2 Argentina 2019 Grand Total 243 hdpe 215
## 3 Argentina 2019 Grand Total 243 ldpe 55
## 4 Argentina 2019 Grand Total 243 o 607
## 5 Argentina 2019 Grand Total 243 pet 1376
## 6 Argentina 2019 Grand Total 243 pp 281
## 7 Argentina 2019 Grand Total 243 ps 116
## 8 Argentina 2019 Grand Total 243 pvc 18
## 9 Argentina 2019 Grand Total 243 grand_total 2668
## 10 Argentina 2019 Grand Total 243 num_events 4
## # ℹ 133,790 more rowsTo do analysis across different plastic types, we reshape the data
from wide format to long format using pivot_longer(). This
transformation allows us to treat plastic type as a categorical
variable, and we create one row per observation, making it easier to
filter by plastic type, aggregate by company, and generate meaningful
visualizations.
## # A tibble: 107 × 3
## country year total_count
## <chr> <dbl> <dbl>
## 1 Argentina 2019 11772
## 2 Argentina 2020 7596
## 3 Armenia 2020 20
## 4 Australia 2019 14
## 5 Australia 2020 3957
## 6 Bangladesh 2019 90
## 7 Bangladesh 2020 4908
## 8 Benin 2019 19954
## 9 Benin 2020 688
## 10 Bhutan 2019 14005
## # ℹ 97 more rowsThis country-level summary provides context for understanding which regions face the most severe branded plastic pollution challenges.
## # A tibble: 2,750 × 7
## country year parent_company plastic_type count large_count country_lower
## <chr> <dbl> <chr> <chr> <dbl> <lgl> <chr>
## 1 Argentina 2019 Grand Total empty 0 FALSE argentina
## 2 Argentina 2019 Grand Total hdpe 215 TRUE argentina
## 3 Argentina 2019 Grand Total ldpe 55 FALSE argentina
## 4 Argentina 2019 Grand Total o 607 TRUE argentina
## 5 Argentina 2019 Grand Total pet 1376 TRUE argentina
## 6 Argentina 2019 Grand Total pp 281 TRUE argentina
## 7 Argentina 2019 Grand Total ps 116 TRUE argentina
## 8 Argentina 2019 Grand Total pvc 18 FALSE argentina
## 9 Argentina 2019 Grand Total grand_total 2668 TRUE argentina
## 10 Argentina 2019 Grand Total num_events 4 FALSE argentina
## # ℹ 2,740 more rowsResults
The following table presents total plastic counts by country for 2019, sorted alphabetically. This summary reveals substantial variation in audit participation and documented waste across nations, reflecting both actual pollution levels and differences in volunteer engagement.
## # A tibble: 52 × 3
## country year total_count
## <chr> <dbl> <dbl>
## 1 Argentina 2019 11772
## 2 Australia 2019 14
## 3 Bangladesh 2019 90
## 4 Benin 2019 19954
## 5 Bhutan 2019 14005
## 6 Brazil 2019 19338
## 7 Bulgaria 2019 69
## 8 Burkina Faso 2019 23299
## 9 Cameroon 2019 35900
## 10 Canada 2019 30
## # ℹ 42 more rowsCountries with established environmental advocacy networks and higher levels of civic engagement typically report larger plastic counts, although this may reflect audit intensity rather than pollution severity alone.
The histogram below illustrates the distribution of plastic item counts across all entries in the dataset. We apply a logarithmic scale to the x-axis to better visualize the highly skewed distribution.
The distribution demonstrates that most audit entries involve small quantities of items, typically fewer than 10 pieces per company-type combination. However, a small number of entries exceed even thousands of items, indicating that certain brands contribute disproportionately to plastic pollution. This pattern underscores the concentration of responsibility among a limited set of corporate actors.
This bar chart identifies the parent companies most frequently found in plastic waste during Argentina’s 2019 brand audits. The data reveals leaders in branded pollution.
Notably, unlabeled items (“null”) represent the largest category, suggesting that significant portions of plastic waste lack clear branding or have degraded beyond recognition. Among branded items, major multinational beverages, like Coke-a-Cola, and consumer goods companies dominate the rankings. While some of these corporations have announced cleanup initiatives and sustainability commitments, their products continue to represent substantial portions of environmental plastic pollution.
The following visualization examines PET (polyethylene terephthalate) waste specifically, as PET is one of the most common plastic types used in beverage bottles and food packaging. Despite being technically recyclable, PET frequently escapes collection systems and persists in the environment.
This chart reveals consistency in the corporate actors responsible for PET pollution across both years. Global beverage companies consistently appear at the top of PET waste rankings, reflecting the dominance of single-use plastic bottles in their business models.
To understand which materials pose the greatest pollution challenges, we aggregate counts by plastic type using both dplyr and data.table approaches.
## Top plastic types (dplyr):## # A tibble: 10 × 2
## plastic_type total
## <chr> <dbl>
## 1 grand_total 1204956
## 2 o 650583
## 3 num_events 446488
## 4 pet 275700
## 5 ldpe 116640
## 6 pp 97696
## 7 hdpe 35742
## 8 ps 21243
## 9 empty 4174
## 10 pvc 3169##
## Top plastic types (data.table):## plastic_type total
## <char> <num>
## 1: grand_total 1204956
## 2: o 650583
## 3: num_events 446488
## 4: pet 275700
## 5: ldpe 116640
## 6: pp 97696
## 7: hdpe 35742
## 8: ps 21243
## 9: empty 4174
## 10: pvc 3169The unlabeled category dominates, but among identifiable plastic types, PET is the most prevalent material in branded waste audits. This finding aligns with the beverage industry’s heavy use of PET bottles. Other significant plastic types include polypropylene (PP) and high-density polyethylene (HDPE), commonly used in packaging and containers.
Conclusion
The analysis reveals that plastic waste counts are highly skewed: most records involve small quantities, while a few companies exceed thousands of items per audit. Beverage producers dominate PET totals, reinforcing calls for extended producer responsibility and refill systems.
This analysis has several limitations that should be acknowledged. Brand audit data depends on volunteer effort and may not represent all geographic regions or pollution contexts equally. Detection bias favors branded items over generic or degraded plastics. The dataset captures only items found during specific cleanup events, not the total scope of plastic pollution. Despite these constraints, the patterns revealed are consistent and significant enough to inform policy discussions.
First, the distribution of plastic waste counts is highly skewed: most audit entries correspond to small numbers of items, while a handful of companies account for disproportionately large amounts of branded waste. Second, PET packaging remains one of the most dominant and environmentally persistent forms of plastic pollution, despite its technical recyclability. Third, a small group of multinational corporations repeatedly appear at the top of branded waste tallies across countries and years.
These patterns reinforce the need for stronger extended producer responsibility initiatives, deposit-return systems, refill and reuse programs, and improved corporate accountability. Voluntary corporate commitments have proven insufficient to address the scale of plastic pollution documented in these audits. Regulatory interventions that create structural incentives for waste reduction and circular economy practices offer the most promising path forward.
References
Break Free From Plastic. (2021). Brand Audit Data 2019-2020. Retrieved from TidyTuesday GitHub repository.
Data Source: https://github.com/rfordatascience/tidytuesday/tree/master/data/2021/2021-01-26