Natural Gas Companies are Choosing Poor Counties To Host Their Toxic Infrastructure
Why Census data is too flawed a way to account for compressor stations’ impact on minority communities
Compressor stations, or large, industrial gas-fired turbines that exist every couple hundred miles along a pipeline route, release toxic chemicals that increase health risks for the individuals that live near them. People within a one-mile radius of these stations are more likely to be afflicted with respiratory problems, cognitive impairments, leukemia, and a number of cancers.
Many poor and minority communities think that they have been targeted by natural gas companies as hosts for this polluting infrastructure, and they might be onto something. Eighty-three percent of the zip codes and 90 percent of the counties with compressor stations are below the national median household income, according to my examination of relevant data — median household income, percent African American, and population density for all zip codes and counties in the U.S. from the Census, joined with ArcGIS’s inventory of compressor stations in the U.S., which includes both the zip code and the county where they are located.
A somewhat common argument made by energy companies is that compressor stations are intentionally placed in rural areas. Such an arrangement could explain the lower-than-average median household income in most compressor station counties, as rural areas tend to be poorer than densely populated ones.
Yet my study found that the average population density for these counties is instead characteristic of small-to-medium sized metro counties, as defined by the National Center for Health Statistics 2013 Urban–Rural Classification Scheme for Counties. The average population density in the U.S. is, in fact, lower than the average density for these counties.
In terms of race and compressor station counties, the data isn’t as dramatic as it is for class. I found that 25 percent of all US counties have compressor stations in them, and 25 percent of US counties where the concentration of African Americans is 25 percent or higher have compressor stations in them. 33 percent of US counties where the concentration of African Americans is 50 percent or higher have compressor stations in them.
That said, some pipeline routes tell a dramatically different story.
70 pipeline companies account for the over 1300 compressor stations in operation nationwide. Only 1.5 percent of these stations are located in neighborhoods that are much more African American than their surrounding counties: but 30 percent of them belong to a single pipeline, Southern Natural Gas.
When comparing zip codes to counties, I found where compressor stations are located, on the whole, have slightly higher median household incomes than those of their overall counties. Bucking that trend are all seven compressor stations along the Algonquin Gas Transmission route. They have median household incomes that are on average $26,000 lower than those of their surrounding counties.
These patterns suggest that compressor stations placed in significantly poorer or blacker neighborhoods, though not part of a large trend, can’t merely be explained as a series of one-offs. A number of them were placed by the same planners and the same teams.
The fact that most compressor station counties are already poor likely explains why there was not a significant difference between zip codes and counties for median household income. If widespread structural discrimination is at play, it’s not between zip codes within a county. Rather, it could be between communities on a pre-determined pipeline route. Or it could be between counties or broader areas before the routes are planned. Because natural gas companies are notoriously vague about how they plan their pipeline routes, it’s difficult to know exactly what to study. The results of my study neither debunk nor suggest patterned discrimination across the natural gas industry: but they do suggest that future studies need to use Census data in a more sophisticated way, if at all.
Some communities and researchers say that Census data is too flawed a way to account for compressor stations’ impact on minority communities. The town of Union Hill, Virginia has proof.
A historically black community founded by freed slaves, Union Hill has been overtaken in the past five years by a community battle over Dominion Energy, which plans to build a compressor station in the neighborhood. Dominion Energy and FERC, the agency that commissions environmental and neighborhood impact assessments for proposed natural gas infrastructure, allege, citing Census data, that the community is 40% African American.
“I live here. I know who lives here. I know my neighbors,” local activist Chad Oba said in response to the 40% figure put forth by the energy company.
Led by researcher Lakshmi Fjord from the University of Virginia, Oba and her neighbors personally knocked on every door within a 1-mile radius of the proposed compressor station. They instead found that 75% of residents were African American. The EPA warns against impact studies using Census tracts for this exact reason.
In a 2016 report, the EPA stated that minority and low-income communities “may reside in tightly clustered communities, rather than being evenly distributed throughout the general population. Selecting a geographic unit of analysis (e.g., county, state, or region) without sufficient justification may portray [these] population percentages inaccurately by artificially diluting their representation within the selected unit of analysis.
Without regulations enforcing this, and without another easily accessible measurement unit, it remains the way that FERC analyzes the communities that energy companies wish to enter. It remains the statistical source that these companies parrot. And it remains, regrettably, the measurement used for this study.
In “Dominion Energy & Environmental Racism: a case study in how to lie with maps”, analyst Stephen Metts breaks down in painstaking detail why and how Dominion Energy used improper data analysis methods to erase the blackness of Union Hill. His study not only demonstrates that the community has many more black people than the company claims — it shows that the population ring 0.5–1 mile from the compressor station has a much higher proportion of African American residents than the rings that are further away.
What I’ve learned from Metts, a professional researcher whose study was exponentially more sophisticated and precise than mine, is that just as public data can be used as a tool to empower oppressed communities, the lack of it — or the misuse of it — can also be used by those in power to perpetuate their agenda. On the one hand, my methods of using publicly available data that any citizen can access, and crunching the numbers that anyone with a working knowledge of spreadsheets could crunch, couldn’t reveal any patterns of widespread racial discrimination by the industry. On the other hand, the incredibly sophisticated methods that Metts used, aren’t being used by energy companies themselves, even though they’re one of the few entities with the resources and manpower capable of commissioning such studies.
Looking forward, I want to investigate how, exactly, energy companies plan pipeline routes and compressor station locations. Learning which comes first, how, and why, can give us better clues about what to study next.
