Environmental justice screening after EJScreen: rebuilding the analysis from the source data

For years, an environmental justice (EJ) screen began the same way: open EJScreen, draw the project area, and read the percentile indexes. That workflow no longer exists. The U.S. Environmental Protection Agency (EPA) removed EJScreen from its website on February 5, 2025, the Council on Environmental Quality removed the Climate and Economic Justice Screening Tool (CEJST) the month before, and Executive Order 12898, the 1994 directive that first made environmental justice a federal mandate, was revoked by Executive Order 14173 in January 2025. For the first time in three decades, federal agencies are not explicitly tasked with advancing environmental justice.

The screening obligation itself, however, survived the loss of the tool. This post explains what still applies to a project, and how to rebuild a defensible EJ screen from the same authoritative data EJScreen was assembled from.

What still requires an EJ analysis

Three obligations outlived the executive orders, and together they are why the screen is still worth running. The first is Title VI of the Civil Rights Act of 1964, which remains in force and prohibits any recipient of federal financial assistance, including most state agencies and local governments, from discriminating on the basis of race, color, or national origin. One limit shapes how a screen should frame its result, because under the Supreme Court's decision in Alexander v. Sandoval (2001) only the federal government may bring a disparate-impact claim under Title VI, so the enforcement path is narrower than the standard itself. The second obligation is state law, since several states run their own EJ statutes and screening tools that the federal rollback did not touch, and California still operates CalEnviroScreen. A project's location, not the federal posture, decides whether a state requirement applies. The third is practical rather than legal, because many agency procedures, lenders, and grant programs still require a demographic and burden analysis as a condition of funding or review.

What EJScreen actually was

EJScreen was a combination, not a dataset of its own. It joined two kinds of public information and reported them as percentiles: demographic and socioeconomic indicators describing who lives in an area, drawn from the Census Bureau's American Community Survey (ACS), and environmental-burden indicators describing the regulated hazards nearby, drawn from EPA's own program data. The value it added was the join and the ranking, but the underlying inputs were always public and remain so today. The screen is therefore reproducible without the tool, because you can pull those same inputs directly from their sources.

Rebuilding the demographic picture

The ACS is the root source for the population side of the analysis, but you rarely have to process raw ACS tables, because two agency-maintained composites are built directly from it and are still published. The Social Vulnerability Index (SVI), from the Centers for Disease Control and Prevention, ranks every census tract on socioeconomic status, household composition, minority status, and housing and transportation, which gives a defensible and citable vulnerability measure without rebuilding it from scratch. The Low- to Moderate-Income Summary Data (LMISD), from the U.S. Department of Housing and Urban Development, reports the share of low- and moderate-income residents in each block group and is the standard income measure for federal program eligibility. Together these reconstruct the socioeconomic half of what EJScreen reported, from federal sources that remain in publication.

Rebuilding the environmental-burden picture

The burden half comes from EPA program data that was never part of EJScreen's removal and is still retrievable layer by layer. The National Priorities List identifies Superfund sites, the brownfields inventory maps contaminated and potentially contaminated parcels, the Facility Registry Service locates regulated facilities, and nonattainment areas mark where the air fails the National Ambient Air Quality Standards. Overlaid on the demographic layer and clipped to the project's area of analysis, these recreate the burden-and-population intersection that an EJ screen is meant to surface.

State the result honestly

A screen of this kind shows where vulnerable populations and environmental burdens concentrate near a project, and that finding is a screening input rather than a conclusion. It does not establish that the project causes a disproportionate impact, and it is not a Title VI finding, because both of those are legal and agency judgments that rest on far more than a desktop overlay. So a sound screen reports what the data shows, names the source and currency of each layer, and states the next step plainly. Where a project is federally funded, the Title VI standard applies and the analysis belongs in the administrative record; where the project sits in a state with its own EJ law, that state requirement may be the binding one and should be checked for the project's location.

Why this is worth assembling cleanly

Rebuilding an EJ screen by hand means pulling the SVI and income layers, adding the EPA burden datasets, normalizing them to a common geography, clipping everything to the area of analysis, and recording the currency of each source. None of that is difficult, but it is slow, it repeats on every project, and every manual step is another place for an error to enter the record. Avow's EJ and Title VI pack pulls those layers current from their official sources, clips them to the project area, records the currency of each, and returns the result with citations and symbology ready for QGIS and ArcGIS Pro. The judgment stays with the practitioner. The data gathering does not have to.