Analytics For Good: Pennsylvania Superfund Sites
Introduction
Superfund sites are designated locations across the United States that have been deemed by the EPA to be unsafe, toxic, and in need of cleanup from past pollution. The creation of these sites came out of a series of fires, accidents, and worrying medical trends for those leaving near heavily polluted areas. With mounting public pressure, the EPA began enlisting these locations on a National Priorities List to be addressed.
While some of these sites simply present harm to nature, the majority of sites introduce toxins into communities, which causes a spike in cancer, birth defects, and countless other negative outcomes for those who live around them. Understanding the impact of living near one of these sites is crucial for deciding where to live and what medical conditions to keep an eye on.
This is particularly important since there are way more sites than you might expect. There are over 90 in Pennsylvania alone and over 1,300 across the United States. This means there are 1,300 communities across the country that might have toxic drinking water, asbestos in their air, or some other major toxin in their environment. And this number continues to grow as the EPA approves new sites each year.
Superfund Sites in Pennsylvania
These sites are particularly common in Pennsylvania, and even more so around urbanized areas, as many of these polluted sites are the product of industrialization. From heavy metal factories to power plants, the source of pollution varies widely, but the majority of them were a product of industrialization before there were strict rules surrounding pollution.
It’s important to note that the dates above were when these sites were designated by the EPA as dangerous to human health, not when they were first polluted. This delta between site pollution and EPA designation has been a massive headache for finding who is responsible for the cleanup. Many of the companies that polluted are now obsolete, absorbed into massive corporations that have no ties to the original site, or simply won’t claim responsibility. This means the cost of clean-up is tossed between several parties.
To even begin the process of cleaning up a site, known as environmental remediation, all of these questions and more must be answered:
Is the toxin actually harmful to humans?
How does it enter our bodies (drinking water, air, radiation, etc.)?
Can it be cleaned up, and if so, how?
How long does it take for the disease to present after the first exposure?
Which populations are most susceptible to that toxin?
Only once all of these queries are confirmed, likely with massive amounts of data analysis, can a remediation plan and source of funding can be solidified. However, this is extremely difficult.
As you can see from the map above, most sites are concentrated around southeastern Pennsylvania and Philadelphia. Many of the points in that area are larger, which is proportional to the EPA’s site score as determined by their Hazard Ranking System. The higher the score, the greater risk the site presents to its immediate environment.
We’ve established what these superfund sites are, where they’re located, and why funding to clean them up is complicated, but why do they matter in the first place? Well, Superfund sites have two primary consequences for communities: health and real estate prices.
Health Impacts: Cancer Clusters
Quantifying the health impacts of these toxic sites is incredibly difficult. The primary cause of this is that every site is so vastly different. We know they present some sort of environmental toxin, but the causality and long-term effect of these toxins are both not well-studied and extremely variable. While cancer is the most common one, with many of the toxins being so diverse, the list of diseases they cause is similarly endless and uncharted.
For instance, while this data analysis primarily focuses on instances of cancer because it’s the most common, the probable diseases from the asbestos dumpsite in Ambler are asbestos-induced mesothelioma or asbestosis.
As you can see from this chart, Pennsylvania counties with Superfund sites have notably higher incidences of cancer than those without them. It’s also important to emphasize that not all toxins from Superfund sites result in cancer, and singling out the ones that do would likely paint an even stronger picture. However, as mentioned, the outcomes from exposure to these toxins are not definitive enough yet to split the data set as such.
From this graph, we can see there is a weak, positive correlation between the site score and cancer incidence near these sites, where the site score represents the potential threat these sites present to their surrounding area. It’s important to emphasize that these site scores range from moderately dangerous to very dangerous since the baseline is being the site of massive amounts of industrial waste, so not seeing a massive difference between these scores is expected.
Real Estate Prices: A Symbol of Oblivion
Prices of homes around Superfund sites tell two very distinct stories. For some, the toxin in the environment is well-understood and visible to residents. For these locations, the drop in real estate prices is closely tied to public visibility of the danger. A primary example of this is in Ambler.
Ambler, a town just 15 miles north of Penn’s campus, was the site of a massive asbestos factory. Much of the waste from this factory was dumped around the town in open piles as the danger of asbestos was not well understood when this occurred. However, even though the factory was shut down, the piles of asbestos remain littered haphazardly around the town.
So why not just pick up the waste and transport it somewhere far away? Disturbing the piles converts asbestos into its friable form, in which the microscopic crystals become airborne and spread for miles. This presents a massive health hazard to Ambler, surrounding towns, and workers that would have to handle the waste as friable asbestos crystals and pierce lung tissue and lead to several types of lung diseases. The waste is here to stay, and in a particularly visible form. This has significant implications for real estate prices.
Above is the Housing Price Index (HPI) of Ambler, PA over the past few decades with two key years highlighted. The HPI is an index that represents the annual total value of home purchases. After the EPA began to phase out asbestos (which was later overturned due to pressure from lobbyists), housing prices plateaued and even slightly declined. Similarly, after the EPA designated the town’s asbestos waste as a superfund site, housing prices sharply fell.
It’s important to note that this part of this decline was due to the 2009 financial crisis, however, the HPI of Ambler is still far below the area average and never reached the same climb in HPI following the recession.
But the BoRit asbestos dumpsite wasn’t always something that people viewed as inherently dangerous. Similar to how we didn’t know cigarettes were harmful until the 1950s or that lead paint was dangerous until the 1980s, we didn’t view asbestos as a toxin in the same way we do today.
“All the asbestos fibers [were] floating in the air. When she came back into her office, she used to have to brush her suit off.”
— Carol DiPietro, recounting growing up in Ambler, PA
The combination of the rapid change in public awareness of asbestos and the visibility of physical piles caused such a significant impact on real estate prices. However, for most Superfund sites around Pennsylvania, the enemy is entirely invisible. Most don’t know they have sites near them because they might think these locations are taped-off, closely-guarded vaults. In actuality, they don’t look like Chernobyl, they look like fields, forests, or lakes. There’s one just a few miles from Penn’s campus–I can see it from my high-rise dorm window–and I doubt many Penn students recognize that.
While sites like Ambler have asbestos, a compound that most understand to be harmful to human health, piled high across the town, many emit invisible radiation or quietly leech toxins into our drinking water. For those, residents are completely unaware, and that’s how we arrive at this graph.
This helps explain why it’s actually more expensive to live near Superfund sites: we are oblivious to an invisible enemy. The HPI of houses near Superfund sites happens to be higher than those without a Superfund site, which can most likely be attributed to higher housing prices closer to Philadelphia where many of these sites are clustered, compared to the rest of the state.
This gets to the root of the problem that between company cover-ups of the pollution, poor attempts to communicate the implications of these sites, and public officials not looking to cause an economic recession in their area by spreading fear, the American public is almost entirely clueless and oblivious to these sites.
Takeaways: So How is This Data for Good?
The combination of oblivious home-buyers and declining funds being sent to the EPA means that these sites continue to be enlisted yet the rate of remediation continues to fall.
“You just didn’t think about it, and you didn’t know.”
— Victor Romano, recounting living in Ambler
Even with the vast advancement of scientific innovation to help identify and remediate the poisons at these sites, this data points to a problem: there is a massive gap between the danger these sites pose to our health and the public perception of these locations. With funding sources drying up, it’s important to remain aware of the impacts of these sites, vote for those who advocate for the removal of the toxins at these sites, and educate those who live near these sites of what medical conditions they should keep an eye out for. The creation of these sites coincided with the birth of the environmental justice movement, and letting these sites go untouched due to lack of funding and awareness would mean ending an era of massive environmental cleanup over the past few decades.
Process
Web scraping:
While a couple of the datasets were downloaded, the majority of what’s used in this project was web scraped from the EPA site. This involved both iterating through tables and using links to traverse through several pages and collect relevant information.
Data Cleaning:
After downloading and importing the data in RStudio, the data was cleaned to remove erroneous details that were a project from web scraping. This included cleaning up dates using lubridate, removing extra whitespace, and deleting any NAs in the dataset. This data was then joined. Data was also filtered nationwide statistics to only match regions within Pennsylvania. To learn more, please see my code for further explanations and methods.
Sources
Superfund Data:
Detailed Superfund Sites — EPA
Housing Data:
Cancer Data:
Historical Facts:
Living in the Town Asbestos Build — Science History