Methodology and Data Sources
By Yasir Siraj · Last updated 21 October 2026
Every number on this site traces back to one source. This page explains what that source is, what is done to it, and where it falls short — because a data site that will not state its limitations is not worth trusting.
Source
All electricity price data comes from the
U.S. Energy
Information Administration Open Data API, specifically the retail sales of
electricity series (API route electricity/retail-sales). EIA is a federal
statistical agency and its output is in the public domain. The series covers revenue,
sales volume, customer counts and average price, broken down by state and by sector.
Everything this site stores is downloadable: each state page links its own CSV, and the whole dataset — every state, sector and month back to January 2001 — is one file.
Processing
- The feed is read daily and stored month by month, so a full price history builds up rather than a single snapshot.
- Prices are normalised to cents per kilowatt-hour. EIA has published this series in both cents and dollars, so the declared unit is read from the response rather than assumed.
- The national average is sales-weighted, not a plain mean of the 51 jurisdictions. Averaging states equally would give Vermont the same influence as Texas, which would be misleading.
- Year-over-year change compares the newest month against the same month twelve months earlier, which avoids seasonal distortion.
How the average bill and usage are calculated
The average electric bill on each state page is EIA's own method: total residential revenue divided by the number of residential customers, over the latest twelve months. Average usage is total residential sales divided by customers over the same months, and the twelve-month average rate is revenue divided by sales. All three use the same window, so usage multiplied by that rate gives the bill.
- Twelve consecutive months, never fewer. A single month is a season, not an average — a July bill is driven by air conditioning. If any month in the window is missing, the page does not quote a state average at all and says so, rather than averaging an unbalanced year.
- Missing sales are derived, not skipped. EIA occasionally publishes a month with revenue and price but no sales figure. Since EIA's price is revenue divided by sales, the missing sales figure is recovered from those two published numbers. Skipping the month instead would understate usage.
- The headline rate is the latest single month. It is what EIA publishes as that month's price, and it is what the national ranking is based on. Bill and usage figures use the twelve-month rate instead, and the page labels which is which.
- The national usage figure is the same calculation across all states, currently 875 kWh a month. It is used wherever a page compares states at identical usage.
Known limitations
- These are state-wide averages. Your utility's tariff may differ substantially, particularly in states with many small municipal utilities or co-operatives.
- There is a reporting lag. EIA publishes roughly two months behind, and revises recent months as utilities file corrected data. Pages here always show the newest published month, not the current calendar month.
- Averages hide tariff structure. Tiered and time-of-use rates are flattened into a single figure. If your utility charges more above a usage threshold, your marginal rate is higher than shown.
- An all-in average, not your tariff. EIA's price is total utility revenue divided by total sales, so delivery, transmission and the fixed monthly service charge are already inside it, spread across every kWh. The consequence is that a household using well below its state's average pays a higher effective rate than the figure shown, because the fixed charge divides over fewer kilowatt-hours — and a heavy user pays less. State and local sales taxes are generally not counted as utility revenue, so those do sit on top.
Update schedule
The data pipeline runs daily. Since EIA publishes monthly, most runs confirm the existing figures rather than change them; the page you are reading updates whenever a new month lands or an old one is revised.
Reuse
The underlying EIA data is public domain and you may use it freely. The presentation, rankings, calculators and written text on this site are ours — a link back is appreciated if you quote them.