Electricity myths against the data
Each claim gets a verdict, True, False or Partly, with the number that decides it on the first screen of its page. Verdicts follow written rules (below) and every figure is computed from public data.
The 12 claims
- False
Turning off lights saves the most
18 kWh
kWh a year saved by one lamp off
- Partly
EVs will overload the grid
0.71%
share of U.S. electricity sold used by today's EVs
- Partly
Rooftop solar always pays back
9 of 51
states with solar payback under 10 years at $3.00 a watt
- Partly
Heat pumps don't work in cold states
0 of 10
of the 10 coldest states where a heat pump beats gas on fuel cost
- False
Higher prices mean more reliable power
r = +0.03
correlation of home price and outage minutes
- False
Renewables make electricity more expensive
r = -0.07
correlation of renewable share and home price
- Partly
The state with the highest rate has the highest bill
#2
bill rank of the state with the highest price
- False
Electricity prices only go up
3 of 51
states where the home price fell in 2025
- Partly
Deregulated states always pay more
15 of 15
choice states above the median of the others
- Partly
Older homes always mean a higher energy burden
r = +0.34
county correlation of old housing and energy burden
- Partly
Big storms cause most outage minutes
77%
share of outage minutes on major event days
- Partly
Hot states have the biggest bills
r = +0.34
correlation of cooling degree-days and the typical bill
Verdict summary with CSV
| # | Myth | Verdict | Key number | What the number is |
|---|---|---|---|---|
| 1 | Turning off lights saves the most | False | 18 kWh | kWh a year saved by one lamp off |
| 2 | EVs will overload the grid | Partly | 0.71% | share of U.S. electricity sold used by today's EVs |
| 3 | Rooftop solar always pays back | Partly | 9 of 51 | states with solar payback under 10 years at $3.00 a watt |
| 4 | Heat pumps don't work in cold states | Partly | 0 of 10 | of the 10 coldest states where a heat pump beats gas on fuel cost |
| 5 | Higher prices mean more reliable power | False | r = +0.03 | correlation of home price and outage minutes |
| 6 | Renewables make electricity more expensive | False | r = -0.07 | correlation of renewable share and home price |
| 7 | The state with the highest rate has the highest bill | Partly | #2 | bill rank of the state with the highest price |
| 8 | Electricity prices only go up | False | 3 of 51 | states where the home price fell in 2025 |
| 9 | Deregulated states always pay more | Partly | 15 of 15 | choice states above the median of the others |
| 10 | Older homes always mean a higher energy burden | Partly | r = +0.34 | county correlation of old housing and energy burden |
| 11 | Big storms cause most outage minutes | Partly | 77% | share of outage minutes on major event days |
| 12 | Hot states have the biggest bills | Partly | r = +0.34 | correlation of cooling degree-days and the typical bill |
4 False, 8 Partly and 0 True. Every number is computed from the data listed under Sources; each page repeats its own sources and retrieval dates.
CSVHow we check a myth
Each claim is turned into something the data can test: a count of states, a correlation across states or counties, or a comparison of two measured quantities. We compute it from the open datasets listed below, write the rule before reading the result, and print the number next to the verdict.
Where our data cannot test the claim, the page says so: no hourly load data, no cause of an outage, no inflation series, no supplier plan prices. Scenarios, such as a 10 W lamp or a price per watt of solar, are labelled our scenario and are inputs you can change.
The rule behind each verdict
| Myth | Verdict rule |
|---|---|
| 1 | True if one lamp's yearly saving is at least the largest appliance in the comparison; otherwise False. |
| 2 | True if today's EV energy is 10% or more of retail sales; False if under 1% and the whole-fleet scenario is under 5%; otherwise Partly. |
| 3 | True if 90% or more of states pay back in under 10 years at $3.00 a watt; False if 10% or fewer; otherwise Partly. |
| 4 | Partly by design: the model tests fuel cost at a constant COP, not performance in cold weather. |
| 5 | True if the correlation of price and outage minutes is -0.3 or lower; otherwise False. |
| 6 | True if the correlation of renewable share and price is 0.5 or higher; Partly from 0.3 to 0.5; otherwise False. |
| 7 | True if the highest-price state also has the highest typical bill; Partly if it is in the top 3 on bills; otherwise False. |
| 8 | True if no state's price fell in 2025 and the U.S. annual series never dipped since 2010; False if both happened; otherwise Partly. |
| 9 | True if every choice state is above the other-state median and no other state is above the cheapest choice state; Partly if the choice median is higher; otherwise False. |
| 10 | True if the county correlation is 0.5 or higher; Partly from 0.2 to 0.5; otherwise False. |
| 11 | True if the national share of minutes on major event days is over 50% and 75% of states agree; Partly if only the national share is over 50%; otherwise False. |
| 12 | True if the correlation of cooling degree-days and the bill is 0.5 or higher; Partly from 0.3 to 0.5; otherwise False. |
What the verdicts mean
True means the data supports the claim as worded. False means the numbers contradict it as a general rule. Partly means the claim holds in some places, for some measures or only as an average. A correlation is a pattern across states and is never read as a cause, and a state average does not describe any one home.
At the moment 4 claims are False, 8 Partly and 0 True. The data change each year, so the verdicts can change; each page names its data period and retrieval date.
Sources and licences
| Source | Data period | Retrieved | Licence | Myth pages |
|---|---|---|---|---|
| EIA electricity retail sales, annual and monthly (API v2) | annual 2010 to 2025, monthly to Jul 2026 | 1 Oct 2026 | U.S. government data (public domain); cite the agency. Our table: CC BY 4.0 | 1, 2, 3, 5 to 9, 12 |
| EIA-861 annual electric power industry report, 2024 final | data year 2024 (2015 to 2023 files for history) | 1 Oct 2026 | U.S. government data (public domain); cite the agency. Our table: CC BY 4.0 | 5, 9, 11 |
| EIA-923 generation (2025 final) and EIA-860 generator file (2025) | data year 2025 | 2 Oct 2026 | U.S. government data (public domain); cite the agency. Our table: CC BY 4.0 | 6 |
| EIA natural gas prices, residential, annual and monthly | 2025 | 1 Oct 2026 | U.S. government data (public domain); cite the agency. Our table: CC BY 4.0 | 4 |
| NOAA NCEI nClimDiv statewide heating and cooling degree-days | 2015 to 2025, normals 1991 to 2020 (files dated 4 Sep 2026) | 2 Oct 2026 | U.S. government data (public domain); cite the agency. Our table: CC BY 4.0 | 4, 12 |
| NLR (formerly NREL) PVWatts v8 API | typical weather year, 1 kW fixed roof system | 2 Oct 2026 | Computed by us from the free API; results only | 3 |
| ENERGY STAR certified products dataset | certified-product listings as of the download | 2 Oct 2026 | U.S. government data (public domain); cite the agency. Our table: CC BY 4.0 | 1 |
| AFDC vehicle registration counts by state (NLR, Experian) | light-duty vehicles, 2021 to 2025, rounded to 100 | 2 Oct 2026 | U.S. government data (public domain); cite the agency. Our table: CC BY 4.0 | 2 |
| EPA and DOE fueleconomy.gov vehicles.csv | model years 2024 to 2026, 913 electric models | 2 Oct 2026 | U.S. government data (public domain); cite the agency. Our table: CC BY 4.0 | 2 |
| FHWA Highway Statistics: annual miles per driver | average miles per driver | 2 Oct 2026 | U.S. government data (public domain); cite the agency. Our table: CC BY 4.0 | 2 |
| IRS: Residential Clean Energy Credit | property placed in service 2022 to 2025 | 2 Oct 2026 | U.S. government data (public domain); cite the agency. Our table: CC BY 4.0 | 3 |
| DOE LEAD (Low-Income Energy Affordability Data) tool | 2022 model year, by county | 2 Oct 2026 | U.S. government data (public domain); cite the agency. Our table: CC BY 4.0 | 10 |
| Census ACS 5-year 2023, table B25034 (year built), by county | 2019 to 2023 | 2 Oct 2026 | U.S. government data (public domain); cite the agency. Our table: CC BY 4.0 | 10 |
Keep reading
Questions
What is a myth page on this site?
A claim about electricity, a verdict and a test. Today 12 claims are covered: 4 False, 8 Partly and 0 True.
Which claims come out False?
Turning off lights saves the most; Higher prices mean more reliable power; Renewables make electricity more expensive; Electricity prices only go up.
Which are Partly true?
EVs will overload the grid; Rooftop solar always pays back; Heat pumps don't work in cold states; The state with the highest rate has the highest bill; Deregulated states always pay more; Older homes always mean a higher energy burden; Big storms cause most outage minutes; Hot states have the biggest bills.
Which are True?
None at the moment: every claim fails or holds only in part on our data.
How are verdicts decided?
By a written rule for each claim, such as a correlation threshold or a count of states, listed in the rule table on this page.
Do you claim causes?
No. A correlation across states shows a pattern. Where a page names a possible reason it says our data cannot test it.
Where do the numbers come from?
EIA, EPA, DOE, Census, NOAA, IRS, FHWA and AFDC data, The sources table gives each period, retrieval date and licence.
Are the scenarios real data?
No. A 10 W lamp, a price per watt of solar or 13,476 miles a year are labelled our scenario or taken from a named average, and every one can be changed in the page's calculator.
Can I download the numbers?
Each page has a table of 51 states with a CSV link, and this page has the verdict summary as a CSV.