Warning: Removed 73 rows containing non-finite outside the scale range
(`stat_boxplot()`).
Industry Specific Sustainability in the US
Introduction
For our final, my area of interest lies in exploring sustainability within specific industries. Different sectors approach environmental, social, and governance (ESG) policies and industries such as energy, manufacturing, retail, and technology, in different ways and have unique ecological impacts and strategies for implementing sustainable practices. Understanding these industry-specific efforts can be incredibly helpful to identify best practices, gaps, and innovations that drive systemic change. I want to understand if there are any patterns between ESG scores and sectors/industries, then later supplement that with some data
I found a dataset from Kaggle that showcases companies from the S&P 500 index, which outlines the ESG performances and risk profiles of the companies that fall within this category.
Data Dictionary:
| Variable Name | Description |
|---|---|
| Stock Symbol | The unique stock symbol associated with the company |
| Name | The official name of the company. |
| Address | The primary address of the company’s headquarters. |
| Sector | The sector of the economy in which the company operates. |
| Industry | The specific industry to which the company belongs. |
| Full Time Employees | The total count of full-time employees working within the company. |
| Description | A concise overview of the company’s core business and activities. |
| Total ESG Risk Score | An aggregate score evaluating the company’s overall ESG risk. |
| Environmental Risk Score | A score indicating the company’s environmental sustainability and impact. |
| Governance Risk Score | A score reflecting the quality of the company’s governance structure. |
| Social Risk Score | A score assessing the company’s societal and employee-related practices. |
| Controversy Level | The level of controversies associated with the company’s ESG practices. |
| Controversy Score | A numerical representation of the extent of ESG-related controversies. |
| ESG Risk Percentile | The company’s rank in terms of ESG risk compared to others. |
| ESG Risk Level | A categorical indication of the company’s ESG risk level. |
Summary Statistics
| Statistic | Mean | Max | Min |
|---|---|---|---|
| Environment Risk Score | 5.74 | 25 | 0 |
| Social Risk Score | 22.5 | 22.5 | 0.8 |
| Governance Risk Score | 6.73 | 19.4 | 3 |
| Total ESG Risk Score | 21.53 | 41.7 | 7.1 |
When looking at the summary stats, it’s important to understand the scale that companies are awarded within. They’re looked at as low controversy, moderate controversy, and high controversy. The higher the ESG risk score, the more harmful it is. Typically above or around a score of 40 indicates a severe risk.
Descriptive Analysis
Through looking at the Total ESG risk Distribution by Sector, we’re able to view how each sector’s distribution looks according to the sector. This way, we’re able to view the differences in the risks associated with each sector, and which are high and low risk. According to the data and graph we plotted, Basic Materials, Energy, and Utilities are the sectors that have the highest medians when it comes to Total ESG Risk. The sector with the lowest Total ESG Risk on the S&P 500 list is Real Estate, which also has the smallest range. Technology has a few outliers that have higher risk scores, which could be associated with largest technology companies. This informs us of the sectors that have a higher impact and risk on ESG standards, and implies that there may need to be some specific, targeted policies to help reduce the impacts of them.
This gives us a good foundation and understanding of what sectors the S&P 500 are composed off. To me, it seems important to understand the composition, and if there are a large amount of companies that are contributing to pollution or another types of waste/pollution and have high ESG risk scores. Upon looking at the graph, it seems like Technology has the highest total number of businesses classified under the Sector, but has a relatively lower ESG risk score than some of the other sectors. Industrials, the sector with the second largest number of businesses, has a somewhat high median and a large distribution. Energy, Basic Materials, and Energy, the sectors with the highest medians, don’t have as many businesses classified under their sector, which might be positive, as more businesses might only increase their risk.
Warning: Removed 1 row containing missing values or values outside the scale range
(`geom_bar()`).
I wanted to observe, within the top 3 sectors with the highest ESG Risk Scores, what the distribution of scores among the industries. The top 3 in this graph fall within the Energy sector, with Oil & Gas E&P, Oil & Gas Integrated, and Oil & Gas Refining & Marketing. This makes sense, since typically, oil and gas companies tend to have a harmful impact on many aspects of ESP. Many other industries on this graph are results of different types of harvesting, mining, and other types of material gathering that often impacts the environment or workers livelihoods.
I wanted to check to see how many industries from the S&P 500 database fell into what controversy level categories. Many weren’t reported, but I felt as though it was worthwhile to show how many industries fell into this. Many of these industries, often large, very profitable organizations, fell into the moderate and significant controversy levels. It looks like it’s pretty unusual for a majority of the companies to have extreme levels, like high, none, or severe. It seems like they tend to fall within the moderate/in between classifications. This makes sense - it would seem odd for a large corporation to have no risk associated with their ESG policies, or for a company to legally get away with having a completely detrimental, unaddressed, unregulated impact.
Warning: Removed 1 row containing non-finite outside the scale range
(`stat_smooth()`).Warning: Removed 1 row containing missing values or values outside the scale range
(`geom_point()`).
I wanted to connect this data with the data that I scraped from the US Climate Alliance website, which is concerned primarily with actions with positive environmental impacts. To do so, I wanted to see how much the Total ESG risk is impacted by the Environmental Risk Score, looking at Sector. Clearly, there is a pretty clear, positive correlation between Total Risk Score and Environmental Risk Score. Lower Environmental Risk scores typically result in lower ESG Risk scores, and vice versa.
Secondary Data Analysis
Using an API taken from the US Climate Alliance, we can observe the sentiments in different states (those who have opted into the alliance) and the interests and emotions behind their policies. After looking at data from the past 5 years, along with related articles and policies. we can observe the sentiment and trends from some of the more climate-friendly states (California, New York, and Vermont). My goal is to see how the ESG risks associated industries and policies enforced align. After creating URLs for our desired scraping, performing any cleansing like data types and deleting columns we don’t need, we scraped our data and exported it into a CSV, which we’ll use here to ensure we’re using static data. We’ll be using the BING lexicon for our text analysis.
For the sentiment analysis, I wanted to see what type of emotion was typically associated with language used in climate action and policies. I looked specifically at the policies that had “climate-governance” as a sector. Climate Governance is a large part of ESG policy, so I figured this would be a good intersection between the two data sources. Upon running the graph, we see that a majority of the terms in Climate Governance actions are associated with the emotions “anticipation”, “positive”, “trust”, and “joy”. For future-looking policies that are created to better the environment or protect certain policies, people, or nature initiatives, seeing the emotions “trust” and “anticipation” make sense. The policies should use language that creates a sense of trust with their constituents and have them looking forward to the changes implemented. It’s also good that there is a positive emotional correlation with the language - the changes are being described in a positive way, for the betterment, protection, and progression of certain initiates.