EDGI and the Right to Trust our Environmental Health

By Kelsey Breseman

I grew up in unincorporated Snohomish County, Washington—an area I often characterized, at the time, as the kind of place where people have guns and horses. As a kid, the woods behind my house felt like an endless adventure: a massive Pacific Northwest wetlands, where beavers would build ever-changing dams you might cross on foot (if you didn’t fall in), where stickerbushes grabbed at your clothes, and dripping thick underbrush would open up into spacious cedar groves that, even to a child, felt sacred. I drank tap water, danced in rainstorms, ate nettles and fiddleheads, blackberries and huckleberries from nearby roadsides. I didn’t discriminate between playing in a stream, and playing in a roadside ditch.

Only as an adult have I started to suspect the environment around me: the natural gas stove in the kitchen, the changing flavor of the water in the sink. I get short of breath since I had COVID, and I keep two air purifiers in different parts of the house for wildfire smoke season. I’m still eating nettles and blackberries from beside the road, but I’m always wondering: what’s on these? Have these been sprayed?

The world has changed, sure, but not as much as I have. Climate change and fires have worsened, but as far as we’ve measured, pollutants in the air have actually been on a multi-decade downward trend (EPA 2022), and though water quality is both subjective and mixed, it’s at least not obviously worsening over the same decades (USGS 2023). The trouble is this little modifier: “as far as we’ve measured.” Not every contaminant gets measured, and most that are measured, aren’t measured consistently by reliable sources (ProPublica 2021).

I work at the non-profit Environmental Data and Governance Initiative (we call it “edgy,” for EDGI). The project I’ve been working on at EDGI looks at the US Environmental Protection Agency (EPA) and its data: how they enforce the Clean Air Act, the Clean Water Act, and others; what those permits look like, what monitoring is done, and whose voices are trusted.

The trouble is, the closer I get to the data, to the system of measurement that leads into our system of governance, the clearer it is that these systems do not keep us safe.

EDGI is an action-oriented research collaborative advancing a modernized Environmental Right to Know (ERTK)—the belief that people should be able to know and make decisions about environmental conditions of concern, and that the collection and stewardship of environmental information should equip people, communities, and workers to protect their health and support the flourishing of surrounding ecosystems.

That’s a pretty expansive goal, compared to where we’re at. Unlike in Europe, which employs the precautionary principle in environmental governance, the United States uses a permission-to-pollute system. That means that every facility, from your local pharmacy, to a construction site, to an oil refinery, has official federal permission to pollute certain controlled substances, up to a certain amount and subject to certain monitoring requirements.

But as EDGI has noted in prior reports, not all harmful substances are controlled (Underhill et al. 2022), the amounts that are permitted are often not scientifically proven to protect humans or the environment (Paris et al. 2017), monitoring is typically self-reported by industry, not required to be an actual measurement (models and estimates are permitted, and the EPA has a low and decreasing rate of facility inspection (Fredrickson et al. 2019). Even when the EPA does inspect a permitted facility and finds a violation, the facility is unlikely to be penalized (Wylie et al. 2020). In all, this amounts to a system of environmental governance that fails at every level—from its permission-to-pollute premise through its ability to ensure facilities keep their pollution amounts within those permit amounts (Nost et al. 2020).

In a recent report, we delved into one further missed step: yes, sometimes data is wrong, ignored, or unenforced—but sometimes, it’s just missing. How Gaps and Disparities in EPA Data Undermine Climate and Environmental Justice Screening Tools takes a closer look at data missingness in EPA’s Enforcement and Compliance History Online database (ECHO). ECHO is a public interface to the data the EPA uses to track violations and enforce compliance of the Clean Water, Clean Air, Resource Conservation and Recovery, and Safe Drinking Water Acts (CWA, CAA, RCRA, SDWA). We analyze that data using open source code, a method that others can replicate and test, to make counts of missing data across these major programs.

We were able to identify a very high amount of missing data in ECHO. This is a major issue for environmental governance and public health; data that details pollution can be used to award resources or issue penalties. Data that’s simply not there means unknown threats to public health. Failure to submit reports looks like (and is characterized as) an administrative violation, but it’s really an administrative violation that conceals potential other harmful violations like unmeasured releases of pollutants. Even seemingly minor missing pieces of data—a typo in a facility’s location, for example— can mean that when officials set metrics for areas to support with funding under initiatives like Justice40, missing data gets left off the map. This can perpetuate patterns of environmental injustice in communities where that data is missing.

Here’s what we found in our Data Gaps report:

  • Over 19,000 facilities under key environmental protection laws are missing basic information such as their latitude and longitude. Nearly all—19,657 out of 19,675 (99.9%)—of these are SDWA-regulated facilities. Missing location information means that these facilities will be left out of any geographic analyses.

  • Data needed for basic environmental justice assessments, such as the percent minority population surrounding a facility or the Census block it resides in, is missing for 14% of the facilities in EPA’s most public-facing database. This increases to 83% of facilities regulated under SDWA.

  • Nationally, the typical facility regulated under each of these environmental protection laws is missing…

    • 86% of CWA-specific information

    • 86% of RCRA-specific information

    • 71% of CAA-specific information

    • 40% of SDWA-specific information

  • Facilities in majority-minority communities typically have more information missing than facilities in majority-white communities, for all acts except SDWA.

  • Facilities already screened by EPA as particularly susceptible for environmental injustice are missing substantially more data, on average, than other facilities. This is especially true for Clean Water Act inspection-related data.

  • 78% of all facilities regulated under the CWA are missing inspection counts, but only 75% of facilities in majority white areas, rising to 83% of facilities in majority-minority areas.

  • Western states including Texas, New Mexico, Colorado, Utah, and Nevada are much worse when it comes to inspection data completeness for facilities in majority-minority communities.

In short, a great deal of critical data is missing, and potentially in ways that worsen existing environmental justice disparities by reducing the data about polluting facilities in majority-minority areas.

As practitioners, awareness of this data missingness is important: we must present our analyses in ways that indicate data quality and completeness, rather than taking the dataset as-is. We must understand and communicate that all of our graphs have error bars, and the harms we can document are minimum harms. We must understand that data missingness does not impact all communities equally, and analyze not only the data, but also the gaps it leaves.

As advocates for justice, we must listen to impacted communities, following their lead in asking for stronger environmental protection from toxins, greater awareness of and participation in the permitting process, and to be taken seriously when conducting their own accountability checks—whether reporting on unpleasant odors or using laboratory-grade measurements of chemicals in the air and water. Changing the logics and infrastructures of the way the United States governs pollutants requires the work of many organizations and approaches, working in coalition.

As a member of the Coming Clean network for environmental justice, EDGI seeks to expand the traditional “right to know” framework into the “right to know, participate and decide” (Louisville Charter). But what does that look like, and how do we get there?

EDGI is a highly intersectional site of work. Our core teams include both staff and volunteer contributors, professionals from across fields such as technology and law, as well as academics spanning environmental health, social sciences, geography, and history as fields of study. One of our most essential ways of working is through coauthorship, a model which not only allows us to incorporate the nuances and expertise from these many perspectives into our public outputs, but also reflexively trains us as collaborators and researchers to incorporate, for example, social sciences framing into our data work.

As a technologist and data scientist, I have been profoundly influenced by the coauthorship model in my understanding of why and how to engage local communities as collaborators. EDGI pursues federal-level governance objectives, but grounds that work by testing our assumptions and tools at the local level.

That line of thought influenced me to reach out to my local grassroots environmental group, Green Snohomish, to show them EPA tools and the tools we’ve been working on at EDGI. I met with them for the first time recently. Green Snohomish is primarily engaged with the issues that are right in front of them: litter down by the river, lead weights in fishing and hunting equipment—why, they asked me, doesn’t the EPA just ban lead in all recreational activities?

I couldn’t answer that, but (using our open source Watershed Notebook) I could show them a table of lead releases—instances, quantities—self-submitted by facilities who have permits to release lead in the Snohomish watershed. Between recreational and industrial lead pollution, which has the bigger impact? I don’t know, and according to EPA’s How’s My Waterway tool, there is only one monitoring location in the watershed with any current data, and that one measures only water level and clarity—not lead.

But lead is bad. Everybody knows that, even if they don’t know the exact quantities (I showed them EPA’s Integrated Risk Assessment tool so they could have a reference point, but health is not my expertise). The group immediately started talking about where they could get test kits, and do citizen science. They asked me: is there something we can use to test fish that we’ve caught, and see if it’s safe? Again, I don’t know.

What’s more—should I have to know? Should they? Is the world we want one where groups of volunteers all over the country feel like they need to personally manage monitors that check for chemicals in their water, and soil, and air in order to know that it’s safe to drink, eat, breathe?

Or is this something we consider a duty of government—a public good, either establishing a minimal baseline, like Medicaid, or providing a full-fledged service, like the library system? Because—if this is work performed at personal expense, by volunteers, we can be certain that safety will be applied only in communities that can afford to buy the tools, put in the time, and learn this expertise.

Imagine, instead, enough—and transparent enough—monitoring of polluting facilities that checking on the quality of air, drinking water, and water bodies, was as simple as checking the weather. No expertise required, just: what’s there? Am I in harm’s way? What can I do about it?

EDGI works both on tools that help now—like the open-source Notebooks of the Environmental Enforcement Watch, the people-centric history of US environmental governance housed at A People’s EPA, and ongoing monitoring and public comment on critical digital tools—and initiatives designed to begin reshaping our governance structures, such as researching and demonstrating ideas for data accessibility, participating in federal digital tool redesign, and long-term projects to situate EPA’s work in the broader context of the environmental justice movement.

Both EDGI and SEHN envision a world beyond permission-to-pollute environmental governance, where the health of humans and the environment are held sacred and kept safe—to the extent, perhaps, that we and future generations can take the right of environmental safety for granted.

We would love to partner with SEHN supporters in our ongoing work—please reach out to us at contact@envirodatagov.org if you would like to connect.