Brownfields Don’t Land Randomly: What My Buffalo Research Still Shows in 2025
At the conclusion of my Graduate Studies,
I wanted to research a topic that would allow me to have all the resources I need, professional assistance, Buffalo intellectuals, data, and motivational support because when you’re a grad student, you need it! I was interning at two locations. One of which helped me immensely where I was able to use ArcGIS to plot the points and map all the brownfields. In this study you will find some interesting and relevant data. I researched mapped active brownfield sites across Buffalo and compared their proximity to community-level vulnerability indicators using ZIP-code analysis. These indicators were not chosen at random. They are commonly used in environmental justice research and federal screening tools. I obtained survey data through various outlets in buffalo to obtain community insight, which shows some quite interesting results. Overall, this research was very fun! Yeah, 🤨 Ok nerd! Nevertheless, this blog post is a revisit to the research since its creation in 2022. I will dig even deeper into the conclusions made and provide a personal and intellectual suggestions for updating to the data in 2025.
In short, the vulnerability indicators used include:
Household poverty and SNAP participation
Households speaking a language other than English
Households receiving SNAP regardless of poverty status
High-risk lead areas where data were available
Brownfield locations were sourced from the New York State Department of Environmental Conservation. Demographic data came from the U.S. Census Bureau. Statistical significance was tested using regression analysis at a 95 percent confidence level.
What the Data Actually Showed
The results were not ambiguous.
There were statistically significant correlations between brownfield proximity and multiple vulnerability indicators, particularly poverty and SNAP participation. The p-values were not borderline. They were extremely small, meaning the likelihood that these patterns occurred by chance was very low.
In plain terms, brownfields in Buffalo are more likely to be located near communities that are already economically stressed and socially vulnerable.
One area where the study was appropriately cautious involved lead exposure. While spatial overlap between high-risk lead areas, poverty, and brownfields was visible, a full statistical test could not be conducted due to data limitations at the time. That limitation was clearly stated and still matters.
What matters more is that the lack of a lead regression does not weaken the central conclusion. It simply means that lead exposure was not overclaimed.
Was the Original Conclusion Accurate?
Yes. The conclusion matched the data.
What it did not do was fully own the strength of the findings.
The language framed the results as an “initial step” rather than stating plainly that structural inequity was already evident at the ZIP-code level. That was a tone choice, not a data problem. As a graduate student, caution is rewarded. In policy and research spaces, clarity matters more.
If the same data were written up today, the conclusion would be firmer, not different.
What Changed by 2025
The data did not suddenly reverse itself. Buffalo did not magically decontaminate its industrial legacy, and vulnerable communities did not relocate en masse.
What changed is the policy context.
Federal recognition expanded
The EPA’s Justice40 Initiative formally acknowledges that low-income and marginalized communities bear disproportionate environmental burdens. That aligns directly with the patterns identified in this study.
Analytical tools improved
Updated versions of EPA EJScreen and cumulative impacts frameworks now use multiple overlapping vulnerability indicators, similar to the approach used in this research.
Language became explicit
What was once cautiously framed as “potential disproportionality” is now openly discussed as systemic environmental injustice in federal planning documents.
In other words, the field caught up.
What the Research Still Gets Right
This work remains relevant because it demonstrates three things that are still true in 2025.
Brownfields cluster
They are not evenly distributed across urban space.
Vulnerability compounds
Poverty, language barriers, and environmental risk overlap rather than operate in isolation.
Community input matters
Survey responses showed that residents are aware of environmental conditions but lack confidence in institutional response, a finding echoed in national environmental justice literature.
Where the Framing Needs Updating
If this research were rewritten today, one issue would need clearer acknowledgment.
Environmental gentrification risk
Green redevelopment and phytoremediation can improve neighborhoods, but without anti-displacement protections, cleanup can push out the very residents who bore the burden. That tension needs to be named explicitly in 2025.
The solution is not less cleanup. It is smarter cleanup paired with housing protections and community-led planning.
Final Takeaway
The conclusion of the original research was correct. The data was sound. The methods were appropriate.
What changed is not the reality on the ground, but the confidence with which we are now allowed to say what the data shows.
Brownfields do not land randomly.
Environmental burdens follow social vulnerability.
And mapping those patterns is no longer controversial, it is expected.
This research did not age out. It aged into relevance. More relevant than ever.
“Tell me in the comments: What are your thoughts on these findings. What do you think this tells us about any location that has high brownfield areas? What are the implications of their respective location?”
