How we calculate PressurePoint Scores and what the data means.
The PressurePoint Score quantifies rate-increase vulnerability for a utility's residential customers by combining two core factors: the utility's effective rate and the poverty rate of its service territory. The poverty rate is defined differently for electric and gas, because exposure to the two fuels is different.
Every household is an electric customer, so a utility's electric score uses the plain poverty rate of its service territory: the share of people below the federal poverty line (ACS PUMS, POVPIP under 100), with no heating-fuel filter, weighted by occupied housing units across the territory. Every poor household in the territory is exposed to that utility's electric rate.
Gas is different: only households that heat with natural gas are exposed to a gas utility's rate. So the gas score uses a fuel-weighted poverty rate, the share of households in the territory that both heat with natural gas and live below the federal poverty line. The U.S. reference lines shown on utility pages (12.5% electric, 4.8% gas) follow the same definitions.
A higher score indicates greater regulatory exposure — customers are paying elevated rates in areas with more poor households exposed to that fuel's cost. Because the poverty inputs differ by fuel, the same geographic area can produce different electric and gas scores.
| Tier | Percentile Range | Interpretation |
|---|---|---|
| Most Exposed | Top 20% of cohort | Maximum regulatory exposure — high rates meet high poverty |
| At Risk | 60th–80th percentile | Significant exposure requiring proactive action |
| Watch List | 40th–60th percentile | Moderate exposure — defensible with demonstrated action |
| Defensible | Bottom 40% | Below-average regulatory exposure |
PressurePoint analyzes two parallel datasets:
For utilities that serve both fuels in the same territory (64 combination utilities, e.g., ConEd, BGE, PG&E), each fuel's score is calculated independently and displayed on a single utility page with a fuel toggle.
Excluded: Cooperatives, Retail Power Marketers, State-owned, Federal, Behind the Meter, and Political Subdivisions. These categories are excluded because the tool's framing around rate-regulated customer relationships doesn't apply cleanly (853 electric utilities and 2 gas cooperatives excluded).
Utilities are grouped into six cohorts by ownership type and customer count. Ranking is done within each cohort, so an IOU Large score is not directly comparable to a Muni Small score in absolute terms.
The gas dataset has no Muni Major utilities (no municipal gas distributor meets the Major threshold); the JSON ships the key with an empty list to keep the frontend schema stable.
A CCA is a public agency that buys electricity generation on behalf of the residents of its member communities. The incumbent investor-owned utility (PG&E, SCE, or SDG&E) continues to own the wires and bills customers separately for delivery, so a CCA sets and bills only the generation portion of the bill (roughly one third of a full residential bill). It does not control delivery.
PressurePoint covers the 25 California CCAs identified in EIA Form 861 (ownership type "Community Choice Aggregator", service type "Energy", meaning generation). They are a single flat group with no size tiers, and they are walled off from the IOU and Municipal rankings: a CCA's generation-only figures are never mixed into full-bill cohorts.
Because a CCA bills generation only, it is benchmarked on its generation rate (not a full bill) and a CCA-only PressurePoint Score (generation rate times CCA-weighted poverty), ranked only against other California CCAs.
Parent (holding company) pages roll up a holding company's subsidiary utilities using the same formula as a single utility: the rate is weighted by residential sales, poverty and bill are weighted by residential customers, and the score is the weighted rate times the weighted poverty rate. A parent is ranked only against other holding companies (56 in the electric cohort, 60 in gas), never inside a single-utility cohort. Subsidiaries keep their own cohorts and their own rankings; the parent rollup is a separate, additional view.
The estimated monthly residential bill is calculated from each utility's own reported revenue, sales, and customer count — so the bill reflects both the local rate and local consumption patterns (dense urban markets like ConEd have lower per-household consumption than high-AC Southern markets).
For utilities that report across multiple service types (bundled and delivery-only for electric; firm residential for gas), all filings are aggregated to produce a single utility-level figure per state. Gas consumption is reported in Mcf (thousand cubic feet); electric in kWh. CCAs are the exception; they show an Estimated Monthly Generation Cost instead of a full bill (see CCA Profile Metrics below).
Utility pages have a rate-basis toggle with two views. The default, All-in (customer pays), divides residential revenue by residential volume for customers on the utility's bundled (basic/default) service: the rate an actual customer pays, including supply. The alternative, Utility-billed (excludes third-party supply), divides all of the utility's residential revenue by all residential volume delivered, which excludes supply bought from a competitive supplier, a Community Choice Aggregator, or a gas marketer. Utility-billed therefore understates the rate for customers who buy supply from someone other than the utility. Rankings, the displayed bill, and the 5-year rate change all follow the selected basis.
For electric utilities, the two views come from EIA-861 residential sales in kWh: bundled service versus bundled plus delivery-only service. 43 utilities produce a materially different rate depending on which basis is selected; utilities whose customers all take supply from the utility are identical on both views. The four FirstEnergy Pennsylvania utilities (Met-Ed, Penelec, Penn Power, West Penn Power) cannot be split by basis because EIA reports them as one combined 2024 filing; this is flagged on their pages.
For gas utilities, All-in uses EIA-176 residential sales (line 1010): revenue and volume for customers who buy their gas from the utility, commodity included. Utility-billed adds residential gas transported for others (line 1110), the delivery-only service for customers who buy gas from a marketer, to both revenue and volume. 32 gas utilities move by at least $0.25 per Mcf, mostly in Ohio and New York. Two utilities, Atlanta Gas Light and Ohio Gas Company, report no residential sales of their own because every customer buys from a marketer, so an all-in rate cannot be computed for them; their pages show the utility-billed figure in both views and say so.
Rates, bills, and customer counts are full-year 2024 averages from EIA-861 (electric) and EIA-176 (gas). The 5-year rate change compares 2024 to 2019 on the same basis. Poverty is the ACS five-year estimate for 2020–2024. Disconnection figures are EIA-112 2024. Because these are annual averages, a utility's current tariff can differ from the figure shown, especially where supply rates reset seasonally: Massachusetts basic service resets twice a year, and gas rates generally follow a seasonal pattern.
A CCA profile replaces the full-bill metrics with generation-only equivalents, because the CCA controls only the generation slice of the bill.
This is the generation charge only; the local utility bills delivery separately, so it is not a full monthly bill.
For CCAs, the full Monthly Bill, the "vs national average" comparison, and the State Rank are intentionally not shown, because each would imply a full-bill comparison that does not apply to a generation-only provider.
This is the single most important input to the score. Both fuels use a territory-weighted poverty rate derived from household-level Census microdata (ACS PUMS, POVPIP under 100), but the underlying poverty definition differs by fuel.
The electric poverty rate is person-weighted poverty for occupied housing units in the territory, with no heating-fuel filter. Every household is an electric customer, so every poor household in the territory is exposed to that utility's electric rate; there is no fuel selection to make.
The gas poverty rate we report is not "all poor households in the area." It's the share of households in the territory that both:
This matters because a gas utility's real rate-case exposure comes from customers who must use that fuel. In a region where most poor households heat with propane or oil, a gas utility's score appropriately reflects a smaller at-risk population. In regions where gas-heating is dominant, gas-weighted poverty closely tracks overall poverty.
Poverty (plain for electric, fuel-weighted for gas) is computed from the actual geography a utility serves. The method varies by utility type:
(county occupied housing units) × (tract share of PUMA). Summed across all counties in the territory and divided by total weight. Result: a county-household-weighted mean of PUMA-level poverty, scoped precisely to the utility's service area.Because PUMAs can differ substantially from surrounding regions, a utility's territory-weighted poverty can diverge sharply from broader geographic averages:
Percentage change in the utility's residential rate from 2019 to 2024. This trajectory metric highlights utilities facing accelerating regulatory exposure, even if their current score is moderate. Displayed-only — does not enter the pressure score formula.
The "vs national average" callouts on utility pages use customer-weighted means across all scored utilities in the relevant fuel dataset, not unweighted averages. This prevents a handful of tiny munis from dragging the reference figure in either direction. CCA profiles do not use a national comparison; they compare a CCA to the California CCA cohort average instead.
Each utility is ranked against same-type peers (IOU or Municipal) within its state, providing local context for the national score. CCAs are not state-ranked; they are ranked only within the 25-CCA cohort.