Global Plastics AI Policy Tool

Countries are exploring ways to reduce the impact of plastic. This tool explores different policy interventions both regionally and globally.

⚠️ Warning: Your browser may not be supported or is out of date. Though we try to test across multiple devices and browsers, we have not tried it on your specific version. If you run into issues, please consider using Firefox or updating your browser.

⚠️ Warning: An application update is available. Please click here to update the application.

Loading please wait...

ℹ️ This is the detailed continuation of our interactive introduction which overviews the opportunities of a global plastics treaty.

🏁 Start here:

Exploring plastics outcomes under different policy scenarios, this tool starts with mismanaged waste like plastic in the ocean. This overview tab allows you to combine different high level policies to reduce waste. Note that this is a "pre-print" model. Things may shift during the publication process.

Overview Showing Plastic as Million Metric Tons.


Global 2049 Plastics Projections


Impact of Policies on Global 2049 Plastics Projections


ℹ️ Your policy selections on the overview tab and this tab are kept in sync.

🏁 Start here:

Exposing additional metrics like sectorial consumption at a regional level, this tab refines your policy options from the first tab, allowing you full control to craft your own unique interventinos.

Details Showing in at as .

Deatils View

Mismanaged Waste

Incinerated Waste

Landfill Waste

GHG (Experimental)



Start of Life


Timeseries description


Production Emissions Consumption Emissions Landfill Waste Recycling
var test = in.prototype + 10;


This tab allows you to configure the behavior of the application. These choices are independent of both your policy selections and changes to financial and system assumptions. It intends to support ergenomic use of the application per user preference and, as appropriate, use of adaptive technology. Note that changes to these settings will be saved to a small text file on your machine called a cookie. If you return to the application later and the file is still present, you will be given the option to restore your settings a few seconds after the application finishes loading.


Some views like the details tab may show a policy panel next to visualization describing that policy. Some users may prefer a linear view where the visualization takes up the whole window width.


By default, the color scheme emphasizes color differences between series (example: to better differentiate end of life fates by color). This may benefit some color blind and low vision users. However, you may prefer a high contrast color scheme which empahsizes difference from colors to the background, helping other low vision users.


The headers allow for interactivity, manipulating the entire visulization from a single component which also describes the current page's contents. This may help some low motor users or those where navigating elements may take extra time. Some adaptive technologies will read these interactive headers as intended while others provide overly long names. Users may instead select a shorter "static" option which removes the header interactive elements.


For major charts, users can instead request a table instead of a graphical visualization. Note that all data may be exported as well.

Code Editors

Navigating to details about a policy will let users change the code describing the logic for that intervention. This embedded editor environment can be exited by pressing esc. However, some users not wishing to edit the code may prefer to disable these components for a more streamlined experience.


Controlling what subset of available visualizations are visibile on the details tab, the tool offers "advanced scrolling" which may cause scrollable areas inside other scrollable areas. However, some users including those with motor impairments, may prefer to only have more simplified scrolling with a single scrollable area.


There are some download artifacts are avilable under the CC-BY-NC 4.0 License.

Business as Usual

The Business as Usual CSV File describes longitudinal projections if no policy intervention is used. This is the same as looking at the results while leaving all of the policy levers in their default values and leaving the overview checkboxes unchecked.

Policy Summaries

The Policy Scenarios CSV File describes the 2050 outcomes (unless another year is specified) for the following common policy selections where a single policy is included in the simulation at a time:
  • banPsPackaging: The result of having banned polystyrene packaging.
  • banSingleUse: The result of having reduced single use packaging by 90%.
  • capVirgin: The result of having capped virgin plastic production to 2025 business as ususal levels.
  • minimumRecycledContent: The result of a 30% minimum recycled content mandate for new products.
  • minimumRecyclingRate: The result of a 30% minimum recycling collection rate mandate.
  • recyclingInvestment: The result of 100B USD investment in recycling collection and infrastructure.
  • reducedAdditives: The result of 60% reduction in additives.
  • taxVirigin: The result of a low consumption tax (see tool for more details).
  • wasteInvestment: The result of 100B USD investment in non-recycling waste management including incineration and landfill.
Furthermore, this CSV includes the following which describe complex multi-intervention scenarios:
  • lowAmbition: 30% reduction in single use packaging, 30% reduction in additives, 20% minimum recycling rate mandate, 20% minimum recycled content mandate, 10 billion USD investment in plastic recycling, and 25 billion USD investment in waste infrastructure.
  • highAmbition: All interventions or "checkboxes" on the overview tab at full strength.
Finally, this CSV also provides these benchmarks for reference:
  • businessAsUsual: 2050 results without any policy intervention.
  • businessAsUsual2024: 2024 results without any policy intervention.
Keep in mind that policy outcomes cannot be added together because some policies may overlap with each other like for mimnum recycled content and minimum recycling (collection) rate. Therefore, to combine policies, users need to either leverage this web application or use the stand alone engine.

Pipeline Detailed Outputs

The output of the machine learning pipeline including a comprehensive SQLite database can be found in the Data Pipeline Zip File.

Stand-Alone Engine

Additional JSON files are also available for download and this tool's simulation can be run outside the browser from the command line. See the stand alone execution README for more details.

Guide Introduction

This guide provides some high level observations from the data as material to complement the tool and serves as a continuation of our interactive introduction. Each section provides a different perspective into these results. The sections:
  • Introduction: This section.
  • Consumption: Consumption projections under business as usual.
  • Waste: Waste projections under business as usual.
  • Policy: Overview of different common policy options.
  • Resources: Links to resources for further reading.


Annual global plastic production in business as usual is set to increase more than 22.3 percent from today to 2050 (612.0 million metrics tons in 2024 to 748.6 MMT) in the business as usual scenario.
Timeseries visualization showing plastic consumption increasing over time and in most regions.


About 73.5 million metric tons of mismanaged waste is generated per year (2024). That is expected to rise about 64.6 percent in 2050 (120.9 MMT). This business as usual scenario sees the greatest mismanaged burden in "rest of world" countries. Specifically, RoW's mismanaged waste is 3.7 times greater than NAFTA, EU 30, and China combined in 2050. Indeed, it accounts for about 78.9 percent of global mismanaged waste in that projection.
Timeseries visualization showing plastic waste increasing over time and in most regions.


These numbers could change under different policy scenarios. Five of the options that seem to have a large impact on mismanaged volume:
  • A Minimum Recycling Content mandate at 40 percent would change annual mismanaged waste by about -46.7 perecent in 2050 (64.4 MMT vs 120.9 MMT without any intervention).
  • If capping plastic production to 2025 levels, this simulation expects a change in annual mismanaged waste of -20.9 percent in 2050 (120.9 BAU MMT in 2050 to 95.6 MMT).
  • A 90 percent reduction on other single-use packaging beyond polystyrene would change mismanaged waste by -16.6 MMT globally, by 2050.
  • With this in mind, a consumer tax on packaging would change mismanaged waste by -9.8 MMT globally, by 2050.
  • Taxes could fund investment. A 100 billion investment in recycling would increase recycling by 99.7 percent (253.5 vs BAU 126.9 MMT) and change mismanaged waste (91.1 with investment vs 120.9 MMT) by -24.7 percent.Meanwhile, 100 billion for waste infrastructure changes mismanaged waste by -69.1 percent (37.4 with investment vs 120.9 MMT).
In addition to these policies which have a large volume impact, consider the following which may also be important in other ways:
  • A ban on Polystyrene Packaging would only change mismanaged waste by -0.6 MMT but addresses a type of often mismanaged polymer found frequently occurring in ocean plastics mass.
  • There are some concerns about additives even if their volume is not large. Removing 90 percent of additives would change mismanaged waste by -1.1 MMT but these still may be important materials to remove.
  • Though banning it changes mismanaged waste only by -0.5 MMT (in part since some nations have already started reducing waste imports), a policy to ban waste trade may still serve equity goals.

With this background in mind, this tool allows users to explore these policies in various scenarios like the following:

  • Without intervention, the 3470 MMT global mismanaged waste produced between 2011 and 2050 would tower roughly 2.4 miles into the sky (3.8 km) if placed over Manhattan in New York City (using plastic bottles to represent that waste). In 2050 alone, 120.9 MMT would be produced, reaching 0.08 miles into the sky (0.13 km).
  • With a low ambition treaty, this tower decreases to 2753 MMT and would reach 1.89 miles into the sky (3.04 km) if placed over Manhattan in New York City (using plastic bottles to represent that waste). In 2050 alone, 80.1 MMT would be produced, reaching 0.05 miles into the sky (0.09 km).
  • With a high ambition treaty using all of these policies, this tower decreases to 1360 MMT and would reach 0.9 miles into the sky (1.5 km) if placed over Manhattan in New York City (using plastic bottles to represent that waste). In 2050 alone, 7.8 MMT would be produced, reaching 0.01 miles into the sky (0.01 km).

Note about recycling: in addition to a Minimum Recycling Content (MRC) mandate, one may also consider a Minimum Recycling Rate (MRR). A 30 percent MRR changes annual mismanaged waste by -20.9 in 2050 (95.7 MMT). In contrast, recall that a MRC at 30 percent would change annual mismanaged waste by about -46.7 perecent in 2050 (64.4 MMT vs 120.9 MMT without any intervention). All that said, these two policies could complement each other and be implemented together. Note MRC dictates how much of new plastic products must come from recycled materials. Meanwhile, MRR dictates how much of plastic waste must be collected for recycling regardless of if those materials are used.


To understand these trends, consider reviewing other resources on variables that may be related to plastics behavior: These variables are considered in the machine learning model used to build this tool's projections.
Timeseries visualization showing total consumption, per-capita consumption, and consumption divided by GDP.


This project allows users to explore different plastics outcomes under various policy scenarios, offering the ability to craft highly customized policy packages and to build new policy interventions through an embedded programming environment. This tool is both a continuation of our interactive introduction and is a deployment of open source code built as a joint project of: Open source on GitHub at SchmidtDSE/plastics-prototype.

Model Information

Note that the current model should be treated as "in pre-print" and, in addition to expanded documentation / discussion, predictions may evolve as this project goes into publication. That said, this effort uses a mixture of techniques to accomplish both projection of business as usual as well as simulate different policy interventions.
  • Consumption is modeled through random forest and uses historic data as well as GDP per-capita (USD 2010 PPP) and population projections.
  • Waste is modeled through random forest and uses historic data as well as GDP per-capita (USD 2010 PPP) and population projections.
  • Goods / materials trade is modeled through random forest and uses historic data as well as GDP per-capita (USD 2010 PPP) and population projections.
  • Waste trade is modeled through random forest and uses historic data as well as GDP per-capita (USD 2010 PPP), population projections, and a flag indicating before / after implementation of China's National Sword.
  • Interventions are mechanistic as described in their intervention supporting documentation.
Focusing on the machine learning, performance evaluation is reported in detail across the entire sweep of models beyond the random forest option. However, focusing on the hidden test set performance for the model configurations in production as mean absolute error:
  • Consumption sees a test MAE of 1.0 MMT. It should be mentioned that there are a relatively large number of classes of consumption.
  • Waste sees a test MAE of 0.01 MMT. This tool highlights that there are only 4 classes of fates.
  • Goods trade sees a test MAE of 1.57 MMT. This project notes an important degree of change in this metric over time in the historic dataset.
  • Waste trade sees a test MAE of 1.15 MMT. It is worth mentioning that this metric has had major shifts due to policy like China's National Sword.
Of course, these measures report on instances randomly selected to go into a hidden test set. These "in-sample" errors may under-estimate what one would see in practice when making temporally displaced "out of sample" predictions in future years. Therefore the model pipeline also runs an experiment in which it hides 2019 and 2020 from the training before using 2019 as validation. This in mind, we observe the following out of sample MAEs for the production model configuration:
  • Consumption sees an out of sample MAE of 1.15 MMT.
  • Waste sees an out of sample MAE of 0.02 MMT.
  • Goods trade sees an out of sample MAE of 2.06 MMT.
  • Waste trade sees an out of sample MAE of 1.56 MMT.
Additionally, for a limited number of features, models sectorize trade masses as ratios to overall net trade with the following performance where the difference in available input data may explain task performance:
  • Hidden test set MAE: 0.2
  • Out of sample test MAE: 0.08
Finally, the modeling team further monitors various segments of the dataset to ensure adequate performance for all sectors, regions, fates, train / test / validation splits, etc. In addition to offering an opportunity for domain experts to evaluate model results, this activity can also provide segment error estimations. One such method of monitoring is to retrain the selected random forest configuration with many different train / validation sets to determine the range of expected performance with some differences anticipated due to chance. All this in mind, consider the following region-level performance given that dimension's particularly notable importance to equity:
  • Consumption: As the actual response vaiable in the model is change in consumption, results are reported as MAE in delta MMT. That said, this MAE stays under an acceptable 2.5 MMT for all regions in the trials though EU 30 and NAFTA tend to stay under 1 MMT whereas China and rest of world may be around 1.5 MMT in practice.
  • Waste: As the actual response variable in the model is percent of waste, results are reported as MAE in percentage points. These stay under an acceptable 3% in all trials though China and EU 30 are seen going above 1% while RoW and NAFTA tend to stay below 1% MMT.
  • Goods trade: As the actual response vaiable in the model is percent traded, results are reported as MAE in percentage points. These stay under an acceptable 4% (note this is larger than some other trade numbers) in all cases with only China being notable with MAE over 2%
  • Waste trade: As the actual response variable in the model is precent traded, results are reported as MAE in percentage points. These stay under an acceptable 3% without any region having a notably divergent distribution of MAEs.
Note that these describe a range of validation MAEs seen across 100 trials of model training with each having a randomly selected test / validation split. Regardless, though any modeling effort leaves room for improvement, these error estimations generally suggest that modeling maintains reasonable error across its important dimensions. Still, this high level overview ignores some deeper modeling information made available by this project. Therefore, for further documentation and detailed results as well as instructions on how to train / run the model on your own, please see: Model version 2024.15 (pre-release). Last updated 2024-05-17. Major edits after initial release include:
  • Added packaging-only versions of some levers on 2024-05-25.
  • Updates to capex / opex costs for incineration and landfill on 2024-05-22.
  • Resolved bug (overall results generally unchanged) in determining trade effects and minor imprecision in virgin cap on 2024-05-21.
  • Fix to EU30 definition in future population projections on 2024-05-18.
  • Updates to EU30 definition and incorporation of new model training data on 2024-05-17.
  • Column names improved on 2024-02-14.
  • Polymer-level tracking and GHG (under feature flag) added on 2024-02-18.
  • Fully comphrehensive tracking of secondary consumption with new polymer-level data added on 2024-02-21.
  • Made global GHG public on 2024-02-22.
  • Production polymer balancing to ensure trade on 2024-03-07.
  • Simplification of polymer-level trade which is slightly less noisy on 2024-03-08.
  • Update default assumption for added material required for reuse on 2023-03-27.
  • Update reuse with addressable products parameter on 2024-04-05.
See open source model pipeline where a Docker image is available.

Contact Information

We have multiple methods of following up:
  • For general communication about this tool including for inquires regarding privacy and data, please send an email to our project inbox.
  • For feature requests, bug reports, and code contributions, please visit the project issue tracker. Thank you for your contribution.
  • To contact the UC Santa Barbara team specifically, see the BOSL contact form.
  • To contact the UC Berkeley team, see the DSE contact form.


This webpage respects your privacy and collects as little information as possible while still providing a positive user experience. By using the application, you agree to the following.

Data Collection

This application records standard server access logs for security and stability reasons, including to prevent abuse. This is a common practice employed by many websites necessary for maintaing application function. This information includes:
  • IP addresses which describe from which location and internet connection the application is accessed.
  • User agent strings which provide basic information about the device and software ("browser") used to access the site.
  • The requested URLs which are unique identifiers for data and other resources requested from our servers.
These logs are maintained for security reasons and to ensure strong application performance but are anonymized within 7 days of collection. These anonymized statistics are maintained indefinently as aggregated information for us to understand usage patterns in order to improve the service and maintain its performance. The anonymized data may also include error or bug reports. That said, and no other presonal identifying information is collected including email address, phone number, race, ethnicity, age, etc. Furthermore, please note that:
  • Personally identifying information is never shared or sold.
  • Anonymized data do not include specific IP addresses.
  • IP address is not retained past 7 days except for security / abuse prevention purposes if potential unusual behavior is deteced like in the case of a large number of requests.
  • No information we collect is used for advertising.
  • Though we do not ask users their age to respect their privacy, this webpage is intended for those 18 years and older.

Data Storage

We save very limited data to your device and that data is not used to track you.
  • Cookies are small text files saved to your computer to maintain data such as preferences between website sessions. This application uses essential cookies to support the core function of the application to, for example, prevent showing explanatory text multiple times and improve user experience. These cookies are required by the application to function.
  • There are no non-essential cookies for this application and cookies are not used for tracking, advertising, or analytics.
  • Some browsers may save temporary copies of our data and code for performance reasons through a process called "caching" which can be controlled through your browser settings. This application may use "cache-busters" to inform your browser when new versions of application resources are available.
  • Local storage is used, for example, to save the application on your computer to support weak internet connections or offline usage. This includes the local storage of the Progressive Web Application if enabled and supported. Other use of local storage is optional in specific application features but is engaged only after your confirmation. For example, an optional additional use of local storage includes saving and loading policy scenarios.
By using this application, you agree to the transfer of standard server access logs and anonymized statistics as described above to be transferred to data centers based within the United States of America.

Security and Access

We take security seriously. Communication between your device and our servers is encrypted with secure socket layer SSL. Access to non-anonymized logs and the configuration / code for the application is limited to the current maintainers of the project, automated systems we've constructed for running the application, and our subprocessors. Note that anonymized data may be shared with project partners for the purposes of tracking project success and impact.


See DreamHost CDPA for more information about our web host / subprocessor. Anonymized bug reports may be managed with with Sentry.

External sites

Note that external sites linked from this application maintain their own privacy policies, terms of service, and licensing. This website is not responsible for any external content. We define external content to include any content at any domain other than


This statement's content was last updated on 2024-01-19. Updates to this policy will be posted to this page.

License and Terms

This project's source is available on GitHub at SchmidtDSE/plastics-prototype. See below for licensing details: This application uses a number of open source components as described in the project readme. A robots.txt is available.


The copyright for this software is "(c) 2023 Regents of University of California" and our central inbox can be used to contact us with any inquires. A big thank you to the extended team:
  • Elijah Baker BOSL, UCSB: Policy research.
  • Nivedita Biyani BOSL, UCSB: Policy research, policy modeling, business as usual modeling, plastics domain knowledge.
  • Carl Boettiger DSE, UC Berkeley: Advisor.
  • Magali de Bruyn DSE, UC Berkeley: Guidance on programming portals, CI / CD, and DSL feedback.
  • Roland Geyer UCSB: Principal investigator, business as usual modeling, policy modeling, policy research, plastics domain knowledge.
  • Kevin Koy DSE, UC Berkeley: Executive leadership, advisor, overview tab design.
  • Ciera Martinez DSE, UC Berkeley: Program management, communications, outreach, advisor, tutorial and new user flow.
  • Douglas McCauley BOSL, DSE, UCSB, UC Berkeley: Executive leadership, policy modeling, policy reserach, communications, outreach, media, data visualization, UX / design.
  • Molly Morse BOSL, UCSB: Program management, communications, outreach, media, data visualization (landing page), UX / design (landing page), overview tab design.
  • Linda Nakasone Community: Community provided bug reports.
  • Neil Nathan BOSL, UCSB: Communications, outreach, media, data visualization (landing page), UX / design (landing page), overview tab design.
  • Sam Pottinger DSE, UC Berkeley: Software engineering, lead machine learning / AI, business as usual modeling, policy modeling, lead UX / design (this tool), lead data visualization (this tool), CI / CD, language engineering.
  • ThoughtLab Agency: Development for the landing page.
  • Kelly Wang BOSL, UCSB: Communications
A humans.txt with additional detail is available with the above people in alphabetical order except for Sam who is serving as the web contact and, thus, listed first. Note that other collaborators may have contributed to the broader constellation of related projects and are listed in related publications.


Publications are still in progress. Please cite the pre-print at 10.48550/arXiv.2312.11359. See also our CITATION.cff file. Thank you!

You can add your current custom policy as a new policy option, creating a checkbox for it on the menu. Alternatively, you can also enter a link to add a shared custom policy, adding it to the options menu with a new checkbox.

Multiple exports are available as CSV files:

These downloads will reflect any policies or other configuration choices currently active / selected in the tool.