How much do heating fuel prices actually move between seasons?
The short version
- Across 760 state-season pairs from 2010-11 to 2025-26, the median season-over-season change in the electricity to gas price ratio is 7.1%. The largest was 62%.
- 37% of moves exceeded 10%. That is what earns this site's ten percent tie band, and it is not what this site previously said earned it.
- 14% of consecutive seasons changed the answer in that state, and 28 of 45 states flipped at least once in sixteen years. 17 never did.
- So the honest answer to "will gas stay cheap" is that in a third of states it has not stayed anything for long. The durable question is not what prices will do, but whether your state's current margin is bigger than its own history of moving.
This site has been asserting something about price volatility since the first build. Here is what sixteen years of it actually look like.
The claim this page was built to test
Every verdict on this site carries a tie band. When a heat pump and the incumbent system land within 10% of each other, the site refuses to call a winner and says the comparison is too close. That band is doing a lot of work: it is the difference between telling somebody their answer is clear and telling them it is not.
The justification given for it, from the first build until now, was one sentence: fuel prices move more than 10% between seasons, so a narrower margin than that is noise.
What sixteen seasons actually show
The number that decides a verdict is not the price of electricity or the price of gas on its own. It is the ratio between them per unit of delivered energy. So that is the quantity whose movement matters, and this is how much it has moved.
| Season-over-season change in the electricity to gas ratio | Value |
|---|---|
| Observations, 2010-11 to 2025-26 | 760 state-season pairs |
| Median move | 7.1% |
| 75th percentile | 13.4% |
| 90th percentile | 21.8% |
| Largest single move observed | 62% |
| Share of moves exceeding 10% | 37% |
The threshold is right. The reason given for it was wrong.
The typical move is 7.1%, which is less than ten. So the old sentence, that prices move more than 10% between seasons, is not true of a typical season and should not have been written as though it were.
The band survives anyway, on a better argument. 37% of season-to-season moves exceeded 10%, and the tail runs out to 62%. A margin of 8% is not safe just because the median move is 7.1%: roughly one season in three moves further than that, and you do not get to know in advance which one. The band is justified by the frequency of large moves, not by the size of an average one.
This correction is logged with the old and new wording, because a justification that turns out to be wrong is worth the same treatment as a number that turns out to be wrong.
How often the answer actually changed
Percentages of a ratio are abstract. The concrete version is how often a state's verdict, under the same equipment assumptions this site uses everywhere, was different from one heating season to the next.
The states that changed answer most often, with their own typical move and where they sit today:
| State | Times flipped | Typical move | Margin today |
|---|---|---|---|
| Arizona | 7 | 5.3% | 2% |
| Virginia | 6 | 3.4% | 6% |
| Missouri | 6 | 8.6% | 9% |
| Washington | 6 | 6.9% | 10% |
| Arkansas | 6 | 12.4% | 30% |
| Georgia | 5 | 6.2% | 1% |
| North Carolina | 5 | 4.8% | 14% |
| West Virginia | 5 | 8.9% | 29% |
| South Carolina | 4 | 8.6% | 1% |
| Hawaii | 4 | 5.2% | 15% |
Flips counted between consecutive complete December to February seasons, using a cold climate heat pump at HSPF2 11.0 with 10% resistance backup against a 95% AFUE gas furnace, which is the same reference equipment used on every other page here.
Notice that a high flip count does not require a volatile state. It requires a state sitting near the line. Somewhere with a large typical move and a large margin is stable in practice; somewhere with a small typical move and a margin of 2% will flip repeatedly on movements nobody would call dramatic.
Which answers are fragile right now
This is the part that is actually useful for a decision. For each state, compare the margin today against how much that state's own ratio typically moves in a year. Where the margin is smaller than the typical move, the current answer is not something to rely on.
| State | Margin today | Its own typical move | Read |
|---|---|---|---|
| South Carolina | 1% | 8.6% | Current answer is smaller than this state's own typical move |
| Georgia | 1% | 6.2% | Current answer is smaller than this state's own typical move |
| Oregon | 2% | 9.0% | Current answer is smaller than this state's own typical move |
| Arizona | 2% | 5.3% | Current answer is smaller than this state's own typical move |
| Texas | 6% | 7.9% | Current answer is smaller than this state's own typical move |
In these 5 states the honest position is that running cost does not decide it, and it may well read the other way next winter. That is more useful than a confident number, because it points the decision at the things that do not move every year: whether the house needs air conditioning anyway, what the ductwork will take, what rebates apply, and how long the existing equipment has left.
The opposite case matters too. A state where gas wins by 60% would need a price move several times larger than anything in this sixteen year record to change answer. Those verdicts are about as durable as this kind of analysis gets.
So will natural gas stay cheap?
Nobody knows, this site included, and any page that answers that question is selling something. But the question can be replaced with a better one that the record does answer.
The question you can actually answer
How far would the price have to move to change my answer, and how does that compare with how much it has moved before?
That turns a forecast into arithmetic. Your state page gives the current ratio and the breakeven ratio; the gap between them is the margin. This page gives you what a normal year's movement looks like. If the margin is several times the typical move, the answer is durable. If it is smaller, it is not.
What the record does and does not license you to say
- Defensible: "In this state the two are within a few percent, and over the last sixteen years that has changed answer 7 times. I would not decide on running cost here."
- Defensible: "Gas wins here by a wide margin and has done consistently. It would take an unprecedented move to change that."
- Not defensible: "Gas will stay cheap." Nothing in this data supports a claim about the future, and 28 of 45 states have changed answer at least once.
If you advise people on this for a living, there is a page for that which puts the same material in front of the conversation you actually have.
Common questions
How much do heating fuel prices change year to year?
Measured as the electricity to gas price ratio, which is what decides whether a heat pump is cheaper to run, the median change between consecutive heating seasons is 7.1%. The 90th percentile is 21.8% and the largest single move in sixteen years of EIA data was 62%. Roughly 37% of moves exceed 10%.
Can I rely on a heat pump staying cheaper to run than gas?
It depends on the size of the margin in your state relative to how much that state moves. Over sixteen seasons, 28 of 45 states changed answer at least once and 17 never did. Where the current margin is several times larger than the state's typical annual move, the answer is durable. Where it is smaller, it is not.
Why does this site use a 10% tie band?
Because roughly 37% of season-to-season moves in the price ratio exceed 10%, so a margin narrower than that will not reliably survive the next winter. Note that this is a different justification from the one the site gave previously: the typical move is 7.1%, not more than 10%, and the band is earned by the frequency of large moves rather than the size of an average one.
Does this account for propane and heating oil?
No. EIA publishes those by PADD region rather than by state and the regional series do not go back as consistently, so this analysis covers electricity and natural gas only. Propane and heating oil are more volatile than either, not less, so treat a narrow margin against those fuels with at least as much caution.
Is past volatility a good guide to future volatility?
Only loosely, and this page is not a forecast. What sixteen years of record can tell you is what counts as a normal move and what counts as unusual, which is enough to judge whether a given margin is comfortable. It cannot tell you what happens next.
How this was checked
Sixteen complete December to February windows pulled from the EIA Open Data API v2 by scripts/fetch-history.js, using the same two series and the same conversion the live pipeline uses: electricity/retail-sales for residential electricity by state and natural-gas/pri/sum for residential gas, converted from $/Mcf at 1,037 BTU per cubic foot.
A season is counted only if all three of its months are present for that state, the same rule the current-season pipeline applies. A partial window is not comparable to a full one and averaging it in would understate the movement.
As a check that the history and the live figures agree, the most recent season computed here reproduces the ratios published on the state pages exactly: New York 5.374, Texas 2.618.
What this page does not cover
Any forecast. This is a description of what has happened, not a claim about what will.
Propane and heating oil, which EIA publishes regionally and which are more volatile than gas rather than less.
Within-season movement. Everything here compares one three month mean with the next. Prices also move inside a winter, and a household buying oil or propane by the tank feels that directly.
Your utility. These are state means. A single utility's rate case can move a bill much more than a state average moves in a year.
Equipment and installation cost, which is excluded everywhere on this site and is not what this page is about.
References
- U.S. Energy Information Administration. Electricity Data Browser, retail sales and average price by state and sector, monthly. Series accessed through the EIA Open Data API v2.
- U.S. Energy Information Administration. Natural Gas Prices, residential price by state, monthly, dollars per thousand cubic feet. Series accessed through the EIA Open Data API v2.
- U.S. Energy Information Administration. Units and calculators explained, energy conversion factors and fuel heat content.
- U.S. Energy Information Administration. Short-Term Energy Outlook, heating degree days by census division, prior ten year average series. Accessed through the EIA Open Data API v2.
Related reading
Check your own state's margin
The other half of the question: how much room does your answer actually have?
Open the state index