PEPI methodology

PEPI compares the pain of affording energy today to the pain of affording it during a baseline period. It tries to capture the full experience; not just the price, but the amount of energy you need to buy. Here’s how it works.

Baseline period

We compare each month with the same month in a baseline period: January 2026 is always compared with the Januarys in the baseline, July with the Julys, and so on. Prices use the average of 2015-2019 as a baseline. Weather uses 1990-2019. Both end in 2019, before the weirdness of Covid-19. We use a longer window for weather because weather varies a lot from year to year, so a longer record gives a steadier idea of what’s typical. Prices are different: they tend to drift upward over time, so a recent five-year window is a fairer comparison.
Because of this, a score tells you how far a month is from typical for that time of year. It doesn’t tell you whether a month cost more than the month before it. Winter and summer bills are naturally higher; PEPI shows how much higher or lower they were than usual.

PEPI equation

PEPI = (Natural Gas Index + Electricity Index + Gasoline Index) / 3
PEPI treats the three main energy sources American households use as equal. Here’s how those components are calculated.

Gasoline Index equation

Gasoline Index = gasoline price ÷ baseline gasoline price × 100
The most straightforward component, the gasoline pain index is a straightforward comparison between current gas prices and the baseline. It doesn’t take into account average miles driven in a month or location.

Natural Gas Index equation

Natural Gas Index = (gas price ÷ baseline gas price) × (weather factor ÷ baseline weather factor) × 100

Weather factor = (1 − heating share) + heating share × heating degree days ÷ average month’s heating degree days

(The average month is a full year of baseline degree days divided by 12.)
Natural gas use is closely tied to heating. The price part works like the gasoline index. The weather part asks how much more or less heating the month needed than a typical month of the same name, and how much that matters for natural gas where you live.
The heating share is the portion of a state’s household natural gas that goes to heating the home, as measured by the Energy Information Administration’s 2020 Residential Energy Consumption Survey. It runs from 20% in Florida to 82% in West Virginia. The rest, mostly water heating and cooking, doesn’t change with the weather.
Heating degree days measure how cold it was. The “average month” is a full year of baseline heating degree days divided by 12, so heating counts in proportion to how much of a year’s gas use it represents.
For example, about 73% of household gas in Arkansas goes to heating. April 2026 had 64 heating degree days, against a 1990-2019 average of 174 for April. That much milder weather lowered Arkansas’s natural gas index by about 40% from what price alone would have given.

Electricity Index equation

Electricity Index = (electricity price ÷ baseline electricity price) × (weather factor ÷ baseline weather factor) × 100

Weather factor = (1 − heating share − cooling share) + heating share × heating degree days ÷ average month’s heating degree days + cooling share × cooling degree days ÷ average month’s cooling degree days

(The average month is a full year of baseline degree days divided by 12.)
Electricity both heats and cools homes, so it responds to both kinds of weather. The shares come from the same survey and show how much of a state’s household electricity goes to heating and to air conditioning. Air conditioning is 36% of household electricity in Arizona and 4% in Washington, so a hot month moves Arizona’s electricity index far more than Washington’s. The rest, such as lighting, appliances, and water heating, doesn’t depend on the weather.
We compare all of that to the same calculation for baseline weather, which lets us know if the pain of affording electricity has gone up or down.

Where the numbers come from

Prices come from the Energy Information Administration: residential electricity and natural gas prices by state, and retail gasoline prices (regional averages where a state has no data of its own). Degree days come from NOAA’s state climate data. The heating and cooling shares come from the Energy Information Administration’s 2020 Residential Energy Consumption Survey. We’ll update the shares when the agency releases the 2024 survey’s end-use tables, which it expects in spring 2027.

What PEPI doesn’t account for

Differences within states: Not every utility charges the same rate. Your electricity or gas rate may have increased faster or slower than others in your state.
Individual use: Got a fully electric house? PEPI doesn’t fully represent your experience. Fully electric house with an EV? It’s even less representative.
Income levels: If your energy got 15% more expensive, but your income doubled, you probably don’t feel the increased cost as much as someone who’s income dropped.
Blackouts: Grid reliability is a real source of energy pain. Spoiled food or losing AC in a heat wave can be expensive and dangerous, but PEPI can’t take that into account yet.
States we can’t score yet: Alaska, Hawaii, and Washington, D.C. don’t have the weather data PEPI needs, and Florida and Louisiana are currently missing recent natural gas prices.
Estimated values: When a recent price hasn’t been published yet, PEPI fills it in with an estimate based on the same month a year earlier and the national trend. Estimated values are flagged in the article.