Washing away greenwashing

• 

  Interactive visualization

• 

  Interface allows exploration of companies and industry sectors

Inspiration

We wanted to create a project that addressed a problem in climate change, however, we have found it disappointing that many solutions to climate change are addressed at individual responsibility, when increasing evidence shows that the role of corporations is much greater. We therefore set out to build something that could help investors and consumers hold companies to account.

We noticed that as consumers, it takes a lot of effort and expertise to dig into the details of a company's climate plan. It is difficult to find out whether a company has made a climate pledge, if they have a concrete and actionable plan to work towards their target, if their reporting mechanisms hold themselves accountable, and whether they use "easy routes" like purchasing carbon credits. We believe companies may make use of this lack of transparent data to make claims that paint themselves in a more favourable light, and we wanted to highlight any disconnect between a company's claims and the concrete steps they actually take.

What it does

Our interactive display illustrates the disconnect between a company's marketing about their climate action and the concrete steps it has taken towards making and sticking to a pledge to the environment. We take data collected from the Net Zero Tracker initiative on Canadian companies, and assign each one a score which summarizes how robust their climate actions are. We then scrape the environmental policy page of the companies, and use two NLP methods (sentiment analysis and scoring based on similarity to a list of keywords) to create a "boast score" to capture how these companies present their environmental actions to the public.

How we built it

webscraping: python, urllib sentiment analysis: BERT, pytorch, word2vec, gensim interactive visualization: plotly dash, python, pandas

Challenges we ran into

Some companies had links to pdf reports instead of having a web page for policies.

Accomplishments that we're proud of

The whole thing!

What we learned

As expected, some companies boasted more about their actions than their Net Zero tracker scores deserved. However, some companies were surprisingly modest!

What's next for Washing away greenwashing

Optimizing the scoring and data collection procedures, automatic online updates to keep up with companies that do better/worse.

Built With

• dash
• gensim
• pandas
• python
• pytorch