Original research

Electricity consumption by state: kWh per household

Q: Does this figure include electricity from my own solar panels?

No. These are utility sales: the electricity a household bought through its meter. Self-consumed rooftop generation never passes through that meter and is not in this dataset, so a state with heavy rooftop solar records lower measured consumption than its homes actually use. California, with 14.7% of households on net-metered solar and the second-lowest recorded consumption in the country, is the clearest case where this matters. EIA-861 records no generation of any kind, so no correction is possible from it.

What 51 jurisdictions actually use, and why the cheapest power in the country sits under the largest meters.

Updated September 2026 · Data as of EIA-861 2024 final release

Written by HyreSolar Research team Research and analysis

Audited by HyreSolar Research team Data audit and fact check

10,359 kWh per household per year US average, 2024

14,422 kWh in Louisiana, the highest 5,938 kWh in Hawaii, the lowest

-0.736 correlation with price Across 51 jurisdictions, association, not cause

The headline

The average US household used 10,359 kWh of electricity in 2024, residential sales divided by residential customer accounts, across every utility that filed Form EIA-861. That single figure hides a range of roughly 2.4 to one: Louisiana averages 14,422 kWh per account and Hawaii 5,938 kWh. Consumption runs in the opposite direction to price (r = -0.736 across all 51 jurisdictions), which is the most repeated and least examined pattern in American electricity. It is an association, and this page sets out plainly why a cross-section of one year cannot tell you how much of it is price, how much is climate, and how much is the fuel a state heats with.

What this number is, before anything is done with it

Every figure on this page is one division: the residential electricity a state's utilities sold in 2024, divided by the number of residential accounts those same utilities were billing. Nothing is surveyed, modelled or weighted. Both inputs come from the same schedule of the same mandatory federal filing, so the numerator and the denominator cover exactly the same set of customers.

That makes it a mean across accounts, and a mean is not a typical household. A state where a tenth of homes run electric resistance heat through a cold winter is pulled upward by that tenth, and the median home there uses less than the average implies. Read every number here as the arithmetic average of a large and skewed population, which is what it is.

An account is also not a home. A master-metered apartment block (one meter, one bill, forty households) files as a single residential account carrying forty households' worth of consumption. Where that stock is large, consumption per account is overstated and the count of households understated. This is the single largest known bias in the table below, and it does not fall evenly across states.

The shape of the distribution

The national average is a poor summary of a distribution this wide. 28 of 51 jurisdictions sit above it and 23 below, and the states at the two ends are not marginally different: they are different by more than a factor of two.

Annual electricity consumption per household, 2024. Residential sales divided by residential customer accounts, by state. HyreSolar analysis of EIA-861 2024. Linear colour scale: unlike solar adoption, which spans four orders of magnitude, consumption spans under three to one and a linear ramp reads correctly.

10,359 kWh per household, US average Residential sales ÷ residential accounts, EIA-861 2024

10,582 kWh in the median jurisdiction The middle of the 51, not the household-weighted mean

14,422 kWh in Louisiana: the highest Ahead of Mississippi at 13,871 kWh

5,938 kWh in Hawaii: the lowest 2.4× below the top of the table

A geography that is not subtle

The top of the table is one region. Louisiana, Mississippi, Tennessee and Alabama take the first four places, and the states above 10,359 kWh are overwhelmingly southern and south-central. Louisiana's 14,422 kWh per account is 39% above the national average.

The bottom is equally concentrated. Hawaii at 5,938 kWh, California at 6,039 and Maine at 6,601 anchor a group that is almost entirely coastal: the Pacific coast and New England, plus a small number of dense, mild jurisdictions.

Two things are worth noticing before any explanation is offered. First, the low group contains Hawaii, which is warm the year round, alongside Maine and other northern states that are not, so temperature alone does not order the table. Second, the low group is also, almost exactly, the list of the most expensive electricity markets in the country. That coincidence is the subject of the next three sections, and it is worth being careful with.

Consumption and price run in opposite directions

Across all 51 jurisdictions the correlation between residential price and consumption per household is r = -0.736. On ranks, which are insensitive to the extreme prices at the top, Spearman's ρ = -0.712. Both are strong by the standards of state-level electricity data.

Price against consumption, all 51 jurisdictions, 2024. Higher-price states consume less per household; the relationship is strong, and it is not a causal estimate. HyreSolar analysis of EIA-861 2024. Both axes are derived from the same filings: price is residential revenue ÷ residential sales, consumption is residential sales ÷ residential accounts. Sales therefore appears in both, which is discussed in the methodology.

Setting aside Hawaii, whose price of 42.86¢ lies so far outside the range of the other 50 that a straight line extrapolated to it means nothing, the states furthest above the line (using more than their price alone would suggest) are Alabama, Louisiana, Mississippi and Texas. The states furthest below it are New Mexico, Colorado, Wisconsin and District of Columbia. Read those two lists side by side and the residual is not noise: the states above the line are hot and humid, the states below it arid, high-altitude, or cold and largely heated by gas. Whatever price is doing, climate is doing something too, and the same regression that produces r = -0.736 also produces a residual pattern that price cannot account for.

Setting aside Hawaii, whose price of 42.86¢ lies so far outside the range of the other 50 that a straight line extrapolated to it means nothing, the states furthest above the line, using more than their price alone would suggest, are Alabama, Louisiana, Mississippi and Texas. The states furthest below it are New Mexico, Colorado, Wisconsin and District of Columbia. Read those two lists side by side and the residual is not noise: the states above the line are hot and humid, the states below it arid, high-altitude, or cold and largely heated by gas. Whatever price is doing, climate is doing something too, and the same regression that produces r = -0.736 also produces a residual pattern that price cannot account for.

Hawaii deserves that specific note, because it breaks the functional form rather than the finding. At 42.86¢ against a national average of 16.48¢, a straight line extrapolated that far predicts a consumption figure well below the 5,938 kWh actually recorded: the fit fails at the edge, in the direction of predicting too little. A relationship that is roughly linear across the middle of a distribution need not stay linear at the edge of it, and we would not use this fit to predict any state's consumption from its price.

The same pattern, grouped

Sorting the 51 jurisdictions cheapest to dearest and cutting them into five groups shows the gradient without any model at all. It also shows something the correlation alone hides: the bill barely moves.

Price group	Jurisdictions	Mean price	Mean consumption	Mean monthly bill	Range of prices
Cheapest fifth	10	11.99¢	12,006 kWh	$119.88	11.51¢ – 12.47¢
Second fifth	10	13.45¢	11,940 kWh	$133.98	12.66¢ – 14.15¢
Middle fifth	10	14.71¢	11,186 kWh	$137.25	14.20¢ – 15.07¢
Fourth fifth	10	16.89¢	9,544 kWh	$133.40	15.18¢ – 19.30¢
Dearest fifth	11	27.25¢	6,964 kWh	$156.14	19.34¢ – 42.86¢

HyreSolar analysis of EIA-861 2024. Unweighted means across the jurisdictions in each group, so a small state counts as much as a large one.

Households in the cheapest fifth of states use 1.72× the electricity of those in the dearest fifth, while paying a monthly bill only 1.30× smaller, most of the price difference is absorbed by consumption rather than showing up on the bill. Figures labelled HyreSolar analysis are computed by us from the EIA source files named below. EIA publishes the inputs; it does not publish these ratios or rankings.

Why this cannot be read as "cheap power makes people use more"

The relationship is real and it is strong. What it is not is an estimate of how households respond to price, and nothing in this dataset can make it one. Four mechanisms are entangled in every one of those 51 points, and they push in the same direction.

Climate does not vary randomly across price

The states with the cheapest electricity are concentrated in the South and south-central US, which are also the states with the longest and most humid cooling seasons. Air conditioning is among the largest single loads an American home carries, and a household in a hot, humid climate runs it for months rather than weeks.

That is not a second explanation competing with price. It is a variable correlated with price, and any cross-section that regresses consumption on price alone assigns climate's share of the effect to price by default. EIA-861 contains no weather variable at all (no degree days, no temperature, nothing) so this page cannot separate the two and does not attempt to.

Heating fuel is a step change, not a gradient

A home heated by electric resistance or a heat pump draws its winter heat through the meter; an identical home heated by piped gas, oil or propane does not, and the difference does not appear in this dataset as heating at all, only as a larger annual total. Electric heating is common through the South and parts of the Northwest, and comparatively rare in the gas-served Midwest and Northeast.

This matters more than any behavioural response to price, because it is a difference in what the meter is being asked to do rather than in how carefully a household uses it. EIA-861 records electricity sales; it does not record what the house burns for heat, so the share of each state's consumption that is simply "heating that happens to be electric" is unobservable here. The summer and winter peak section below is the closest this dataset gets to seeing it.

The housing stock is different

Average floor area, the share of detached single-family homes against flats, the vintage of the stock and the insulation standards in force when it was built all vary systematically between the regions at the top and bottom of this table. Larger, newer, detached houses in the South; smaller, older, denser stock in the Northeast and on the Pacific coast.

Household size varies too, and consumption per account is not consumption per person. A state with more people per home will show higher consumption per account at identical per-person use. None of these variables is in EIA-861; all of them are correlated with the geography of price.

Price is partly an outcome, not only an input

The direction of the arrow is not obvious even in principle. Fixed network costs recovered over a larger volume of sales produce a lower cents-per-kWh price, so heavy consumption mechanically lowers the average price a state records: the same denominator that appears in this page's consumption figure appears in its price figure.

Where consumption is high because the climate demands it, the state's average price is lower for that reason alone, before any household decides anything. A correlation between two ratios that share a term is not evidence about behaviour, and treating it as one is the most common mistake made with this dataset.

What would be needed to answer the question properly

Separating these would take variation in price that is not itself caused by consumption or climate (a tariff change, a rate case, a policy discontinuity) observed within the same state over time, with weather controlled. That is a different study on different data, and there is a large economics literature doing exactly that.

What we can say from this dataset, and all we say: cheap-power states consume much more, expensive-power states consume much less, the association is strong and stable across 51 jurisdictions, and the causal share of price within it is not identified here.

Summer peaks, winter peaks, and what they reveal

EIA-861 asks every utility for its summer and its winter system peak. The ratio between them is the one instrument in this dataset that says something about what the electricity is doing rather than only how much of it there is.

Summer and winter peak demand, the twelve largest state systems. Gold bars run from the winter peak up to a higher summer peak; grey bars run down to a lower one. HyreSolar analysis of EIA-861 2024 Operational Data. These are sums of utility-reported system peaks and not coincident state peaks: two utilities peaking on different days are added as though they peaked together, so the levels overstate. The summer-to-winter ratio is far more robust than either figure alone, because both halves are inflated the same way.

A system whose demand peaks in summer is cooling-driven. A system whose demand peaks in winter is being asked to heat, and in a country where most heat comes from gas, a winter peak is a strong indication that a meaningful share of the local housing stock heats with electricity.

Of the 51 jurisdictions, 18 record a winter peak above their summer peak and 33 peak in summer. The most summer-weighted is Nevada, at 2.08× its winter peak; the most winter-weighted is West Virginia, whose winter peak is 1.30× its summer one.

The consumption difference between the two groups is large. Winter-peaking jurisdictions average 11,713 kWh per household; summer-peaking ones average 9,471 kWh. That is a gap of 2,242 kWh a year, and it is consistent with electric heating adding a load that cooling-dominated states do not carry for as long.

It is also, immediately, another confounded comparison, which is the point of showing it. The winter-peaking group averages 14.28¢ per kWh and the summer-peaking group 18.58¢, so the two groups differ in price as well as in what they use electricity for. Restricting the comparison to the cheaper half of the country only (states at or below 14.91¢) narrows it: within that half, the 14 winter-peaking states average 12,212 kWh against 11,302 kWh for the 12 summer-peaking ones. A gap of 910 kWh survives a crude price control. That is suggestive, and with 26 observations and no weather data it is not more than suggestive.

The 18 winter-peaking jurisdictions

State	Winter ÷ summer	kWh per household	Consumption rank	Price	Monthly bill
West Virginia	1.30×	12,320	#15	15.07¢	$154.76
Wyoming	1.25×	10,361	#28	12.47¢	$107.65
Washington	1.25×	11,465	#20	11.90¢	$113.68
Alaska	1.25×	6,939	#44	24.82¢	$143.54
Alabama	1.17×	13,718	#4	15.18¢	$173.50
Kentucky	1.14×	12,558	#12	12.79¢	$133.81
Tennessee	1.11×	13,843	#3	12.42¢	$143.32
North Dakota	1.09×	12,344	#14	11.51¢	$118.38
Montana	1.05×	10,226	#29	12.66¢	$107.91
South Carolina	1.03×	12,605	#10	14.23¢	$149.51
Vermont	1.03×	6,884	#45	21.90¢	$125.66
South Dakota	1.02×	11,925	#18	12.86¢	$127.81
North Carolina	1.02×	12,184	#16	14.13¢	$143.50
Oregon	1.02×	10,582	#26	14.70¢	$129.62
Arkansas	1.01×	12,580	#11	12.32¢	$129.13
Louisiana	1.00×	14,422	#1	11.73¢	$140.96
Mississippi	1.00×	13,871	#2	13.39¢	$154.83
Missouri	1.00×	12,007	#17	12.91¢	$129.18

HyreSolar analysis of EIA-861 2024. Jurisdictions whose reported winter system peak exceeds their summer peak, ordered by how strongly.

Of the 18, 9 sit in the top fifteen of the consumption table. The exceptions matter as much as the rule: a cold state with gas heat and no air conditioning can peak in winter at a low level of annual use, which is why this column is an indicator of what the load is, not a second measure of how large it is.

Using a lot is not the same as paying a lot

The consumption table and the bill table are close to mirror images at the extremes, and almost unrelated in the middle. Prices vary by roughly 3.7 to one across the country, consumption by 2.4 to one, and the monthly bill (the product of the two) by only 2.3 to one.

Where a state ranks on consumption against where it ranks on its bill. The sixteen jurisdictions with the widest divergence between the two. HyreSolar analysis of EIA-861 2024. Ranks run 1 to 51 with 1 the highest in each case, and are precomputed in the dataset so that no two pages on this site can rank a state differently.

The two ratios are cancelling. Households in cheap-power states buy far more electricity and are billed at a fraction of the rate; households in expensive states buy much less at a much higher rate. What arrives in the post is far more similar than either input suggests, and that is why a national bill average is a more stable and less informative number than a national consumption average.

Idaho is the sharpest example in the direction of volume: 11,326 kWh a year puts it #21 on consumption, while its bill of $108.73 ranks only #46. Hawaii is the sharpest in the other direction: #51 on consumption, #1 on the bill.

For anyone reading this to work out what solar would do for them, that gap is the important part of the page. A rooftop array replaces kilowatt-hours, and what those kilowatt-hours are worth depends entirely on the rate they would otherwise have been bought at. Two households consuming the same 10,359 kWh (one in Louisiana at 11.73¢ and one in Hawaii at 42.86¢) are looking at completely different economics for physically identical output. The bill analysis takes that decomposition further.

All 51 jurisdictions, ranked by consumption

Every jurisdiction, ordered by annual electricity consumption per residential account. Price, bill and the two system peaks are shown alongside, so the relationships described above can be checked row by row.

#	State	kWh/year	vs US avg	Price	Monthly bill	Summer peak (MW)	Winter peak (MW)	Peaks in
1	Louisiana	14,422	+39%	11.73¢	$140.96	17,125	17,179	Winter
2	Mississippi	13,871	+34%	13.39¢	$154.83	12,010	12,035	Winter
3	Tennessee	13,843	+34%	12.42¢	$143.32	51,750	57,639	Winter
4	Alabama	13,718	+32%	15.18¢	$173.50	22,378	26,199	Winter
5	Florida	13,250	+28%	14.14¢	$156.09	60,570	47,603	Summer
6	Texas	13,154	+27%	14.94¢	$163.72	102,324	95,486	Summer
7	Oklahoma	12,949	+25%	12.24¢	$132.05	32,774	28,075	Summer
8	Arizona	12,899	+25%	14.91¢	$160.24	22,777	14,156	Summer
9	Georgia	12,888	+24%	14.08¢	$151.25	53,252	52,922	Summer
10	South Carolina	12,605	+22%	14.23¢	$149.51	20,283	20,902	Winter
11	Arkansas	12,580	+21%	12.32¢	$129.13	13,329	13,398	Winter
12	Kentucky	12,558	+21%	12.79¢	$133.81	16,610	18,997	Winter
13	Virginia	12,388	+20%	14.41¢	$148.77	26,819	26,146	Summer
14	North Dakota	12,344	+19%	11.51¢	$118.38	9,843	10,772	Winter
15	West Virginia	12,320	+19%	15.07¢	$154.76	10	13	Winter
16	North Carolina	12,184	+18%	14.13¢	$143.50	44,949	45,785	Winter
17	Missouri	12,007	+16%	12.91¢	$129.18	34,889	34,895	Winter
18	South Dakota	11,925	+15%	12.86¢	$127.81	4,222	4,301	Winter
19	Nebraska	11,477	+11%	11.53¢	$110.28	10,740	7,725	Summer
20	Washington	11,465	+11%	11.90¢	$113.68	16,114	20,130	Winter
21	Idaho	11,326	+9%	11.52¢	$108.73	4,376	3,360	Summer
22	Nevada	11,154	+8%	15.00¢	$139.39	9,479	4,549	Summer
23	Maryland	11,146	+8%	17.86¢	$165.87	8,036	6,922	Summer
24	Delaware	10,928	+5%	16.57¢	$150.87	4,965	4,159	Summer
25	Indiana	10,814	+4%	14.77¢	$133.06	25,568	24,182	Summer
26	Oregon	10,582	+2%	14.70¢	$129.62	29,114	29,600	Winter
27	Kansas	10,508	+1%	14.15¢	$123.90	9,547	6,645	Summer
28	Wyoming	10,361	+0%	12.47¢	$107.65	1,009	1,266	Winter
29	Montana	10,226	-1%	12.66¢	$107.91	4,672	4,908	Winter
30	Ohio	10,146	-2%	15.99¢	$135.16	72,208	63,200	Summer
31	Iowa	9,987	-4%	13.40¢	$111.54	11,107	9,490	Summer
32	Pennsylvania	9,806	-5%	17.77¢	$145.17	26,580	23,277	Summer
33	Utah	9,290	-10%	12.22¢	$94.57	2,526	2,214	Summer
34	Minnesota	8,548	-17%	15.45¢	$110.06	19,443	15,520	Summer
35	Connecticut	8,335	-20%	28.75¢	$199.66	8,347	6,302	Summer
36	Illinois	8,317	-20%	15.87¢	$109.99	30,387	23,308	Summer
37	Colorado	8,091	-22%	14.92¢	$100.57	21,407	17,977	Summer
38	New Jersey	7,949	-23%	19.34¢	$128.13	13,438	8,562	Summer
39	New Mexico	7,849	-24%	14.20¢	$92.88	3,182	2,530	Summer
40	Wisconsin	7,742	-25%	17.18¢	$110.87	28,738	20,663	Summer
41	District of Columbia	7,672	-26%	17.71¢	$113.23	5,314	4,332	Summer
42	New Hampshire	7,428	-28%	23.40¢	$144.87	2,464	1,886	Summer
43	Michigan	7,419	-28%	19.30¢	$119.31	45,139	30,937	Summer
44	Alaska	6,939	-33%	24.82¢	$143.54	841	1,050	Winter
45	Vermont	6,884	-34%	21.90¢	$125.66	916	942	Winter
46	New York	6,854	-34%	24.43¢	$139.53	43,260	32,826	Summer
47	Massachusetts	6,836	-34%	29.35¢	$167.20	6,569	4,869	Summer
48	Rhode Island	6,802	-34%	28.65¢	$162.40	2,719	1,974	Summer
49	Maine	6,601	-36%	24.29¢	$133.60	2,056	1,958	Summer
50	California	6,039	-42%	31.97¢	$160.86	85,239	52,852	Summer
51	Hawaii	5,938	-43%	42.86¢	$212.12	1,501	1,465	Summer

HyreSolar analysis of EIA-861 2024. Consumption is residential sales ÷ residential customer accounts; price is residential revenue ÷ residential sales; the bill is residential revenue ÷ residential accounts ÷ 12.

Peak columns are sums of utility-reported system peaks within each state and are not coincident state peaks. One row shows what that does: West Virginia reports a summer peak of 10 MW across 866,456 residential accounts (12 watts per account against a median of 7,533) which cannot be a description of that state's actual system. Where the load serving a state's accounts is reported by a filer attributed to a different state, the sum here understates it, and we have not attempted to repair the row. Its summer-to-winter ratio survives; its level does not. Read the peak columns for direction, never for magnitude. Figures labelled HyreSolar analysis are computed by us from the EIA source files named below. EIA publishes the inputs; it does not publish these ratios or rankings.

Two readings of the same table

What the table supports

That American households differ enormously in how much electricity they buy (2.4 to one between the extremes) and that the difference is systematic and regional rather than random.

That the low-consumption end of the table and the high-price end are close to the same set of states, at r = -0.736.

That 18 jurisdictions are asked to heat as well as cool, that they consume more on average, and that some of that gap survives a crude control for price.

That the monthly bill is a far narrower distribution than either of the quantities that produce it, which is why it makes a poor proxy for either.

What it does not support

Any statement that cheap electricity causes higher consumption, or that raising a price by a stated amount would reduce use by a stated amount. Nothing here identifies an elasticity.

Any inference about an individual household. This is a mean across accounts in a skewed distribution, and includes master-metered buildings counted as single accounts.

Any comparison with a figure from another source without checking the denominator. Estimates built on Census occupied housing units rather than utility accounts will differ, and neither is wrong.

Any conversion into what a solar array would offset. EIA-861 records sales, not generation, and nothing on this page can be turned into output on a specific roof.

Methodology

Source and vintage

US Energy Information Administration, Form EIA-861, 2024 final release, with the 2014–2024 archive used for context elsewhere in this library. Downloaded as the published ZIP archives on 2 September 2026. Filing is a legal obligation for US electric utilities, so this is a census rather than a sample and carries no sampling error.

The consumption, price, bill and account figures come from the Sales to Ultimate Customers schedule. The peak columns come from the Operational Data schedule of the same filing and the same year.

The formula, stated exactly

Consumption per household = total residential megawatthours sold in the state × 1,000 ÷ total residential customer accounts, both summed across every utility filing for that state in 2024, then rounded to the nearest kWh.

Price = total residential revenue ÷ total residential megawatthours, expressed in cents per kWh. Monthly bill = total residential revenue ÷ total residential accounts ÷ 12. The three are internally consistent by construction: price × consumption ÷ 12 reproduces the monthly bill for every row in the table.

Peak ratio = summer peak MW ÷ winter peak MW, both as reported. No adjustment, no normalisation by customer count.

The double-counting trap, and how it is handled

EIA files service in restructured markets in two halves: a competitive supplier's energy sale and the incumbent utility's delivery of that same electricity to the same household. The two carry identical megawatthours. Summing sales across every Part therefore counts the same kilowatt-hour twice in states such as Texas, Ohio and much of the Northeast, and inflates consumption per household badly.

We sum revenue across all Parts, and sales and customer accounts across Parts A, B and D only. The evidence that this handling is correct is external: the resulting price series reproduces EIA's own published residential price across all 357 overlapping state-years to within 0.005¢/kWh, which is rounding. Consumption is built from the same handled quantities, so it inherits that validation.

The correlation, and what was computed

r is the Pearson product-moment correlation between residential price and consumption per household across all 51 jurisdictions, unweighted, each jurisdiction is one observation regardless of population, which is the right choice for a statement about states and the wrong one for a statement about households. ρ is Spearman's rank correlation on the same pairs. Both are computed in the page module from the same dataset the tables read.

The dashed line on the scatter plot is an ordinary least squares fit of consumption on price, drawn only across the observed range of prices. It is there to show the strength and the scatter, not to be used for prediction. No standard errors or p-values are reported: with 51 non-independent, non-randomly-sampled jurisdictions, the assumptions behind an inferential test are not met, and a p-value would imply a precision the design does not have.

Limitations

Accounts, not households. Master-metered buildings appear as one account carrying many homes' consumption. This inflates consumption per account and deflates household counts, unevenly across states, and it is the largest known bias in the table.

A mean in a skewed distribution. EIA-861 reports totals, so a median household is not computable from it. Every figure here is an arithmetic mean, and in states with a long tail of very large or electrically heated homes the mean sits above the typical home.

One year, no weather. 2024 is a single cross-section. A hot summer or a cold winter moves a state's figure, and this dataset contains no degree-day or temperature variable with which to adjust for it. Year-to-year comparisons of a single state should be made cautiously for that reason.

No fuel mix, no floor area, no household size. The three variables most likely to explain the residual variation are absent from the source. That is a limitation of the data, not an omission from the analysis, and it is why the causal question is left open above rather than answered badly.

Shared terms. Residential sales appears in the numerator of consumption and the denominator of price. Two ratios that share a term are mechanically related even before any behaviour is involved, which weakens what the correlation between them can carry.

Peaks are not coincident. The summer and winter peak columns sum utility-reported system peaks that did not necessarily occur on the same day, or within the state's borders. They support the direction of the peak, not its magnitude.

Reproducing or correcting this

The input is a public EIA download, the arithmetic is stated in full above, and every number on this page (in the prose, the tables and the charts alike) is generated from one machine-written dataset rather than typed. A figure in a sentence and the same figure in a table cannot drift apart.

If you find an error, tell us. Corrections are made on the page with a dated note rather than silently, and the page is rebuilt against the same URL when EIA publishes the next data year.

Questions

How much electricity does the average American household use per year?

10,359 kWh a year in 2024, or about 863 kWh a month, residential electricity sales divided by residential customer accounts across every US utility that filed Form EIA-861 (HyreSolar analysis). It is a mean across accounts rather than a typical household: a master-metered apartment building counts as one account, and the distribution is skewed upward by homes with electric heating.

Which state uses the most electricity per household?

Louisiana, at 14,422 kWh per residential account in 2024, 39% above the national average. Mississippi (13,871 kWh), Tennessee (13,843 kWh) and Alabama (13,718 kWh) follow. The lowest is Hawaii at 5,938 kWh, a spread of about 2.4 to one.

Do people use more electricity because it is cheap?

This data cannot answer that, and it is worth being precise about why. Cheap-power states do consume much more (r = -0.736 across 51 jurisdictions) but the cheap states are also the hottest and most humid, are more likely to heat with electricity, have larger detached housing, and mechanically record a lower average price precisely because they sell more kilowatt-hours over the same fixed network costs. A single-year cross-section with no weather, fuel or housing variables cannot separate those. The association is solid; the causal share of price within it is not identified here.

Why do southern states use so much more electricity?

The most likely contributors are a long cooling season, a housing stock more often heated by electricity than by gas, and larger detached homes, but EIA-861 contains no weather, fuel or floor-area data, so their relative sizes cannot be measured from it. The one internal indicator is the peak ratio: 18 jurisdictions report a winter system peak above their summer peak, and those jurisdictions average 11,713 kWh per household against 9,471 kWh for summer-peaking ones.

My state has low consumption but a high bill. How?

Because the bill is price multiplied by quantity, and the two move in opposite directions across the country. Prices span roughly 3.7 to one between states and consumption 2.4 to one, but monthly bills span only 2.3 to one. Hawaii ranks #51 of 51 on consumption and #1 on the monthly bill. The bill page separates the two components state by state.

Does this figure include electricity from my own solar panels?

No. These are utility sales: the electricity a household bought through its meter. Self-consumed rooftop generation never passes through that meter and is not in this dataset, so a state with heavy rooftop solar records lower measured consumption than its homes actually use. California, with 14.7% of households on net-metered solar and the second-lowest recorded consumption in the country, is the clearest case where this matters. EIA-861 records no generation of any kind, so no correction is possible from it.

How does this compare with figures published elsewhere?

Most published US household consumption figures divide the same EIA sales data by a different denominator (occupied housing units from the Census, or households from the Residential Energy Consumption Survey) and land a few hundred kWh away as a result. Ours uses utility customer accounts, from the same filing as the sales, so the numerator and the denominator cover identical customers. Neither approach is wrong; they answer slightly different questions, and comparing figures across the two without checking is how contradictory numbers end up in circulation.

How current is this, and when does it update?

It is the 2024 data year from EIA's final release, retrieved 2 September 2026. EIA publishes final Form EIA-861 data for a year in approximately October of the following year. An early release for the next year exists before then and is deliberately excluded here, because it covers only part of the filer population and would show a false collapse in every total. This page is rebuilt against the same URL when the next final release lands.

Written and audited by

HyreSolar Research

Primary-source research, data analysis and fact checking

We are a research desk, not a sales floor. We read the statute, the tariff, the code section, the federal filing or the manufacturer data sheet ourselves, and we publish the figure with the document it came from and the date we retrieved it. Where a number cannot be traced to a primary source, we publish the shorter page and say what we could not verify. That rule has cost us whole sections, and it is the reason the rest can be trusted.

160: primary sources read and cited
220: figures with a retrieval date
115: federal and state government sources
66: researched pages published

How this desk works

Primary sources only. Statutes from the legislature’s own publishing system, federal data from the agency that collects it, code text from the adopted edition, manufacturer claims from the data sheet. We do not cite an article that cites a source; we go and read the source.
Every figure carries its provenance. A named document and the date we retrieved it, so you can check it and so you know how old it is. Retrieval dates are not decoration: an EIA rate from May is a different fact from an EIA rate from August.
We publish what we could not verify. Every research page carries a section naming the things we tried to establish and could not, and why. A paywalled standard, a state website that refused the request, a manufacturer that publishes no figure at all.
We separate measurement from modelling from our own reasoning, and label which is which on the page. A laboratory measurement, an assumption inside a modelling tool and our own inference are three different kinds of claim and they are never presented as one.
We do not sell solar, and we take no payment for placement, ranking or a favourable mention. Nobody buys a position on this site.

Data as of EIA-861 2024 final release. Authorship on this site is organisational: the analysis belongs to the desk rather than to a named individual, and we do not publish credentials we do not hold. Our editorial policy sets out how we source, date and correct what we publish.

Sources & retrieval dates

US EIA, Form EIA-861 — Net Metering (annual files, 2014–2024) — Utility-level net-metered installations, capacity and PV-paired batteries by state and customer sector. Retrieved 2 September 2026.
US EIA, Form EIA-861 — Sales to Ultimate Customers (annual files, 2014–2024) — Residential revenue, sales and customer counts by utility and state. The basis for price, bill and consumption figures. Retrieved 2 September 2026.
US EIA, Average Price of Electricity to Ultimate Customers by End-Use Sector — EIA’s published price series, used to validate our derivation. Agreement across all 357 overlapping state-years is within 0.005¢/kWh. Retrieved 2 September 2026.

Your own consumption, not your state's average

The last twelve months of your bills carry a number no state average can give you. Put it into the model with your tariff and your roof, and it prices a system rather than a market.

Price it against your own usage All calculators

HyreSolar is an independent analysis and matching service. We are not an installer, lender or utility. When a reader asks to be introduced, installers may pay us a referral fee. That fee never buys ranking, scores or placement in research. Our editorial policy sets out the rules.