Welcome back to Mapistry Labs, where we make EPA enforcement data easy to understand for environmental teams.
In our last edition, we looked at high-level enforcement trends for the Clean Air Act (CAA). This left some questions unanswered, specifically around penalties. In our second edition, we dive deeper into CAA penalties, breaking down data by industry and state, so you can better understand how your specific sites are affected.
Skip the penalties. Get a demo of our environmental software.
If you'd rather get the short version, here are the four key takeaways from Mapistry Labs #2:
For industry definitions, please review the NAICS website. If you’d like to share our charts, you can download them here.
To start, let’s take a look at how fines differ between industries. After all, you want to understand how your specific industry is changing.
In 2023, the manufacturing industry received over 50% of CAA all enforcement activities with a penalty. Mining (19%) and utilities (9%) were the two other most commonly named industries.
Now, we don’t know if that’s just because manufacturing, mining, and utilities are the biggest industries. The biggest industry could just see more action because there are more businesses, right?
To control for industry size, we calculated the number of cases per 100,000 businesses. We also aggregated data into 5-year averages to reduce fluctuations.
What we found is that manufacturing, mining, and utilities are still the top offenders, even controlling for industry size. But the mining industry, which is comparably small compared to manufacturing (33k vs. 664k registered businesses), is seeing a penalty density that's 8 times higher than the manufacturing industry.
Although businesses in the manufacturing and utilities industry receive fewer penalties per 100,000 than those in mining, they have a penalty density that's 16 and 32 times higher than all other industries, respectively.
In our last piece, we saw that penalties exploded in 2023, but didn’t know where the increase happened. We calculated 5-year-average penalties from 1984 to 2023 by industry to see where the biggest changes happened. Like before, mining, manufacturing, and utilities showed the biggest changes in average penalty payments. The wholesale and retail trade industries also saw explosive penalty growth.
Here are our findings for each industry:
Overall, it’s important to note that some industries fall into each of these categories that you wouldn’t necessarily expect. For example, oil refineries fall under manufacturing, while oil & gas extraction is categorized as a mining business, within NAICS. We will dive deeper into each of these industries in future issues.
Similar to our industry approach, we normalized the number of penalties issued by calculating a per-hundred-thousand rate for each state. The higher the number, the higher the likelihood of a penalty.
The result for each state could be interpreted in different ways. For one states with a high penalty density could be the home for many high-risk industries (mining, manufacturing, etc. See above). For states like California, it’s likely that stricter enforcement is a cause of increased penalty density.
The states with the highest penalties were New Mexico, Utah, and North Dakota. As mentioned earlier, New Mexico saw some of the largest penalties, like the $45.5 million Ameredev penalty for burning natural gas.
On the right side you’ll also see states where penalties have rapidly increased or shrunken, with New Mexico leading the pack, and Arizona as the fastest-shrinking state.
We analyzed the publicly available data from the EPA’s Integrated Compliance Information System for Air (ICIS-Air). For a detailed description of the dataset, please visit the EPA’s website.
4,320 formal actions were not part of the analysis for NAICS codes because their industry was not classified (99—Nonclassifiable Establishments).
We removed data up to the year 1980 because some years were missing data. We also excluded 2024 data as it is incomplete.
We analyzed data using Row Zero and visualized it with the help of Google Sheets and Figma.