Version 2, published October 12, 2026

Voxari Studio · Report

Ending the twice-yearly clock change

United States by state: Census 2020–2024 · Asked October 12, 2026 at 2:17 PM of 300 people

Synthetic simulation: these answers come from Jev's judgement of invented people, not from a survey.

What stands out

Hypotheses

  1. People with children at home say the clock change disrupts their sleep or routine more than those without.

    H₀ Net lean among Children at home is the same as net lean among No children at home.H₁ Net lean among Children at home is higher than net lean among No children at home. (what you expect)

    Not supported · Children at home: net lean +20 · No children at home: net lean +15 · gap +5 points · 97 and 203 people

    Each person's figure, in ten bands: how many people (weighted) fall in each. Triangles mark the means.

    −1000+100

    Children at home: mean +20No children at home: mean +15

    Welch's t-test, one-sided: t = 3.63, df = 290, p < 0.001. H₀ rejected at the 5% level.

    Gap, 95% confidence interval: +2 to +7 points. Effect size (Cohen's d): 0.40, small.

    GroupPeopleEffective sizeMeanSD (points)95% CI of the mean
    Children at home9797+207.9+18 – +21
    No children at home203203+1513.9+13 – +17

    SD is the spread of people's figures, in points (for a scale, points of its range). The effective size counts people by their weights.

Verdicts weigh size: supported takes a gap of 5 points or more the expected way; contradicted, 5 or more the other way; anything between is not supported.

With invented people, Jev's answers vary less than real people's, so p-values come out smaller than a real survey's would. How big a gap is matters more than its p-value.

Why this test

Each person's figure is an average of Jev's probabilities (or, for runs before October 2026, their single pick), so the test compares means rather than counts of 0 and 1.

Two groups: Welch's t-test, which does not assume the two groups vary equally; its degrees of freedom come from the data (Welch–Satterthwaite). Against everyone, the group is compared with everyone else.

One group on two questions: the same people answer both, so each person's difference is tested (a paired t-test), which removes what differs between people.

People count by their population weights, and each side's size is its effective size: (sum of weights)² ÷ sum of squared weights. "Higher" and "lower" are tested one-sided, "differs" two-sided, at the 5% level.

Groups are large, so this gives nearly the same p-value as a z-test; the t-test matters for small groups, such as two traits combined.

At a glance

Every question's answers, one bar each. Each question has its own answers and colours.

1What would you prefer the country do about the clock change?

Keep daylight saving time all year 31%Keep standard time all year 47%Keep changing the clocks twice a year 5%No preference 16%

2Does changing the clocks noticeably affect your sleep or your routine for a few days?

Yes 58%No 42%

How the questions relate

Question by question

Pick two questions: each row is the people giving one answer to the first, split by their answers to the second. People count by their probabilities, so the margins are the run's shares.

ShareYesNo
31%Keep daylight saving time all year57%43%
47%Keep standard time all year60%40%
5%Keep changing the clocks twice a year58%42%
16%No preference55%45%

300 people answered both. Each row adds up to 100%.

Segments

People grouped by how they answered every question, into 4 segments of people who answer alike (k-means on their answer probabilities). Each shows its size, the trait values most over-represented in it, and its answers (top bar) beside everyone's (faded).

Segment 1 · 30% of people (91)

  • Age “75 to 95”: 40%, against 12% of everyone
  • Marital status “Widowed”: 18%, against 6% of everyone
  • Age “65 to 74”: 36%, against 14% of everyone
  • Disability “Has a disability”: 56%, against 22% of everyone

1. What would you prefer the country do about the clock change?

2. Does changing the clocks noticeably affect your sleep or your routine for a few days?

Segment 2 · 27% of people (81)

  • Age “18 to 24”: 17%, against 11% of everyone
  • Children at home “Children at home”: 49%, against 32% of everyone
  • Party lean (modelled) “Democratic-leaning”: 64%, against 44% of everyone
  • Age “45 to 54”: 21%, against 15% of everyone

1. What would you prefer the country do about the clock change?

2. Does changing the clocks noticeably affect your sleep or your routine for a few days?

Segment 3 · 26% of people (79)

  • Age “25 to 34”: 35%, against 19% of everyone
  • Household income “$150,000 or more”: 48%, against 28% of everyone
  • Party lean (modelled) “Republican-leaning”: 70%, against 42% of everyone
  • Work “Working”: 95%, against 60% of everyone

1. What would you prefer the country do about the clock change?

2. Does changing the clocks noticeably affect your sleep or your routine for a few days?

Segment 4 · 16% of people (49)

  • Party lean (modelled) “Independent or no lean”: 63%, against 14% of everyone
  • Household income “Lives in group quarters”: 16%, against 4% of everyone
  • Home “Lives in group quarters”: 16%, against 4% of everyone
  • Work “Unemployed, looking for work”: 16%, against 4% of everyone

1. What would you prefer the country do about the clock change?

2. Does changing the clocks noticeably affect your sleep or your routine for a few days?

The number of segments is the one, from 2 to 5, whose segments are furthest apart for their own spread (the Calinski–Harabasz score; higher is better: 2: 304, 3: 393, 4: 406, 5: 356), leaving out any with a segment under 5% of people. The same run always gives the same segments.

Question 1 · Multiple choice

What would you prefer the country do about the clock change?

Adjusted with Claude Beta

Keep daylight saving time all year37%
Keep standard time all year27%
Keep changing the clocks twice a year18%
No preference18%
By age, adjusted
  • 18 to 24: Keep daylight saving time all year 41% · Keep standard time all year 17% · Keep changing the clocks twice a year 18% · No preference 25%
  • 25 to 34: Keep daylight saving time all year 43% · Keep standard time all year 20% · Keep changing the clocks twice a year 17% · No preference 20%
  • 35 to 44: Keep daylight saving time all year 39% · Keep standard time all year 26% · Keep changing the clocks twice a year 16% · No preference 18%
  • 45 to 54: Keep daylight saving time all year 38% · Keep standard time all year 33% · Keep changing the clocks twice a year 16% · No preference 13%
  • 55 to 64: Keep daylight saving time all year 32% · Keep standard time all year 38% · Keep changing the clocks twice a year 16% · No preference 14%
  • 65 to 74: Keep daylight saving time all year 35% · Keep standard time all year 29% · Keep changing the clocks twice a year 20% · No preference 16%
  • 75 to 95: Keep daylight saving time all year 32% · Keep standard time all year 24% · Keep changing the clocks twice a year 24% · No preference 20%

Jev's answers moved toward Claude's for 50 of the same people, group by group of age. On 17 held-out questions with published results, this landed about 8 to 18 points from the real result, against Jev's 21 to 34. Jev's own answers are below; hypotheses, drivers and downloads use them.

Jev's own answers

Keep daylight saving time all year31% ±3
Keep standard time all year47% ±3
Keep changing the clocks twice a year5% ±<1
No preference16% ±2

Most likely answer: Keep daylight saving time all year 36% · Keep standard time all year 54% · Keep changing the clocks twice a year 0% · No preference 10%

Jev's average confidence 47%

± is the 95% sampling interval for these people (an effective 300 after weights): how far another draw of the same population could move a figure.

Model error is separate and wider: on 20 published questions, four answers in five landed within ±22 points of the real survey's share (±25 within groups).

Claude, 50 people: Keep daylight saving time all year 38% · Keep standard time all year 25% · Keep changing the clocks twice a year 19% · No preference 17%

Jev and Claude are 22 points apart: far apart.

By age: the answer where they differ most
  • 18 to 24 (7 asked): Keep changing the clocks twice a year Jev 5%, Claude 19%
  • 25 to 34 (11 asked): Keep changing the clocks twice a year Jev 5%, Claude 18%
  • 35 to 44 (9 asked): Keep changing the clocks twice a year Jev 6%, Claude 20%
  • 45 to 54 (7 asked): Keep standard time all year Jev 41%, Claude 25%
  • 55 to 64 (5 asked): Keep standard time all year Jev 52%, Claude 28%
  • 65 to 74 (5 asked): Keep standard time all year Jev 70%, Claude 30%
  • 75 to 95 (6 asked): Keep standard time all year Jev 80%, Claude 30%

With 50 people, gaps of a few points can be chance. A large gap means the result depends on the model: compare with real results where you can.

What drives the answers

How much of the variation in people's answers lines up with each trait. Traits overlap: pay travels with role, so both can rank high for the same reason.

Age41%
Work28%
Disability22%
Household income22%
Party lean (modelled)18%
Marital status16%
Division / State15%
Education10%

With the other traits held constant

Each group's difference from its trait's largest group in the share choosing “Keep standard time all year”, in points, from one regression with every trait in at once, so a trait no longer gets credit for the traits that travel with it (pay with role). The dot is that effect with its 95% interval; the ring is the same difference one trait at a time, as the ranking above reads it. Where the two part, the trait's apparent effect came from others.

Age against 25 to 34

18 to 24-2 ±4
35 to 44+5 ±3
45 to 54+9 ±4
55 to 64+19 ±5
65 to 74+30 ±5
75 to 95+33 ±6

Party lean (modelled) against Democratic-leaning

Republican-leaning-19 ±3
Independent or no lean-14 ±3

Disability against No disability

Has a disability+14 ±3

Household income against $150,000 or more

Under $25,000+14 ±5
$25,000–49,999+13 ±4
$50,000–74,999+11 ±4
$75,000–99,999+9 ±4
$100,000–149,999+1 ±3
Lives in group quarters0 ±3

Division against South Atlantic

East South Central+1 ±6
Pacific+4 ±4
Mountain+8 ±5
West South Central-1 ±4
New England+6 ±5
East North Central+6 ±4
West North Central+4 ±4
Middle Atlantic+10 ±4
12 more traits

Marital status against Married

Widowed+7 ±7
Divorced-1 ±5
Separated+5 ±5
Never married-2 ±3

Education against High school diploma or GED

No high school diploma-5 ±5
Some college or associate's degree-3 ±3
Bachelor's degree-3 ±4
Graduate degree+2 ±4

Race and ethnicity against White

Black+2 ±3
Hispanic-3 ±3
Asian-2 ±5
Other or more than one race-4 ±8

Sex against Woman

Man-5 ±2

Voted in 2024 (modelled) against Voted

Did not vote+2 ±3
Not eligible (not a citizen)-3 ±2

Work against Working

Unemployed, looking for work+2 ±6
Not in the labor force+5 ±3

Health coverage against Has health coverage

No health coverage+4 ±5

Citizenship against US citizen

Not a US citizen-3 ±2

Children at home against No children at home

Children at home+3 ±3

Veteran against Not a veteran

Veteran-3 ±5

Born against Born in the US

Born abroad+1 ±5

Home against Lives in an owned home

Lives in a rented home-1 ±3
Lives in group quarters0 ±3

Together the traits account for 88% of the variation in people's answers (300 people).

Age and Work together

Share choosing “Keep standard time all year”, the most chosen answer, in each combination of the two traits that drive answers most. Combinations of fewer than 30 people are greyed. Everyone: 47%

Age \ WorkWorkingUnemployed, looking for workNot in the labor force
18 to 2430%34%27%
25 to 3430%35%29%
35 to 4434%44%48%
45 to 5437%21%53%
55 to 6447%36%67%
65 to 7464%36%74%
75 to 9583%–80%

On the map

Share answering “Keep daylight saving time all year”.

Delaware: 1 person, too few to readDistrict of Columbia: 1 person, too few to readFlorida: 21 people, too few to readGeorgia: 10 people, too few to readHawaii: 1 person, too few to readIdaho: 2 people, too few to readIllinois: 11 people, too few to readIndiana: 6 people, too few to readIowa: 3 people, too few to readKansas: 3 people, too few to readKentucky: 4 people, too few to readLouisiana: 4 people, too few to readMaine: 1 person, too few to readMaryland: 5 people, too few to readMassachusetts: 6 people, too few to readMichigan: 9 people, too few to readMinnesota: 5 people, too few to readMississippi: 3 people, too few to readMissouri: 5 people, too few to readMontana: 1 person, too few to readNebraska: 2 people, too few to readNevada: 3 people, too few to readNew Hampshire: 1 person, too few to readNew Jersey: 8 people, too few to readNew Mexico: 2 people, too few to readNew York: 18 people, too few to readNorth Carolina: 10 people, too few to readNorth Dakota: 1 person, too few to readOhio: 11 people, too few to readOklahoma: 3 people, too few to readOregon: 4 people, too few to readPennsylvania: 12 people, too few to readRhode Island: 1 person, too few to readSouth Carolina: 5 people, too few to readSouth Dakota: 1 person, too few to readTennessee: 6 people, too few to readTexas: 26 people, too few to readUtah: 3 people, too few to readVermont: 1 person, too few to readVirginia: 8 people, too few to readWashington: 7 people, too few to readWest Virginia: 2 people, too few to readWisconsin: 5 people, too few to read56: not in this populationAlabama: 4 people, too few to readAlaska: 1 person, too few to readArizona: 7 people, too few to readArkansas: 3 people, too few to readCalifornia: 28% (n=35)Colorado: 5 people, too few to readConnecticut: 3 people, too few to read

Most places here have fewer than 30 people. To fill the map, build the population with “At least this many in each subgroup” set to 30: about 1,500 people for these places, weighted back to their true sizes.

0%28%Fewer than 30 people

By division

Keep daylight saving time all yearKeep standard time all yearKeep changing the clocks twice a yearNo preference
Everyone30047% ±3 Keep standard time all year
East South Central1743% ±10 Keep standard time all year
Pacific4847% ±6 Keep standard time all year
Mountain2344% ±9 Keep standard time all year
West South Central3647% ±9 Keep standard time all year
New England1345% ±13 Keep standard time all year
South Atlantic6348% ±6 Keep standard time all year
East North Central4250% ±7 Keep standard time all year
West North Central2049% ±10 Keep standard time all year
Middle Atlantic3849% ±8 Keep standard time all year
Every place (50)
Keep daylight saving time all yearKeep standard time all yearKeep changing the clocks twice a yearNo preference
Everyone30047% ±3 Keep standard time all year
East South Central / Alabama444% ±21 Keep standard time all year
Pacific / Alaska143% Keep daylight saving time all year
Mountain / Arizona747% ±26 Keep standard time all year
West South Central / Arkansas371% ±40 Keep standard time all year
Pacific / California3550% ±8 Keep standard time all year
Mountain / Colorado550% ±28 Keep standard time all year
New England / Connecticut348% ±55 Keep standard time all year
South Atlantic / Delaware144% No preference
South Atlantic / District of Columbia157% Keep daylight saving time all year
South Atlantic / Florida2147% ±12 Keep standard time all year
South Atlantic / Georgia1052% ±16 Keep standard time all year
Pacific / Hawaii154% Keep standard time all year
Mountain / Idaho248% ±286 Keep daylight saving time all year
East North Central / Illinois1148% ±14 Keep standard time all year
East North Central / Indiana641% ±22 Keep daylight saving time all year
West North Central / Iowa345% ±30 Keep standard time all year
West North Central / Kansas339% ±52 Keep standard time all year
East South Central / Kentucky445% ±32 Keep standard time all year
West South Central / Louisiana451% ±43 Keep daylight saving time all year
New England / Maine164% Keep standard time all year
South Atlantic / Maryland563% ±28 Keep standard time all year
New England / Massachusetts642% ±24 Keep standard time all year
East North Central / Michigan951% ±22 Keep standard time all year
West North Central / Minnesota554% ±32 Keep standard time all year
East South Central / Mississippi339% ±87 No preference
West North Central / Missouri548% ±31 Keep standard time all year
Mountain / Montana152% Keep daylight saving time all year
West North Central / Nebraska254% ±248 Keep standard time all year
Mountain / Nevada349% ±28 Keep daylight saving time all year
New England / New Hampshire164% Keep daylight saving time all year
Middle Atlantic / New Jersey845% ±25 Keep standard time all year
Mountain / New Mexico253% ±210 Keep standard time all year
Middle Atlantic / New York1846% ±11 Keep standard time all year
South Atlantic / North Carolina1058% ±19 Keep standard time all year
West North Central / North Dakota147% Keep daylight saving time all year
East North Central / Ohio1154% ±18 Keep standard time all year
West South Central / Oklahoma355% ±55 Keep standard time all year
Pacific / Oregon454% ±40 Keep daylight saving time all year
Middle Atlantic / Pennsylvania1255% ±18 Keep standard time all year
New England / Rhode Island172% Keep standard time all year
South Atlantic / South Carolina549% ±14 Keep daylight saving time all year
West North Central / South Dakota179% Keep standard time all year
East South Central / Tennessee654% ±17 Keep standard time all year
West South Central / Texas2646% ±11 Keep standard time all year
Mountain / Utah342% ±56 Keep daylight saving time all year
New England / Vermont147% No preference
South Atlantic / Virginia836% ±24 Keep daylight saving time all year
Pacific / Washington743% ±17 Keep standard time all year
South Atlantic / West Virginia266% ±286 Keep standard time all year
East North Central / Wisconsin552% ±27 Keep standard time all year

By race and ethnicity

Keep daylight saving time all yearKeep standard time all yearKeep changing the clocks twice a yearNo preference
Everyone30047% ±3 Keep standard time all year
White18749% ±4 Keep standard time all year
Black3351% ±8 Keep standard time all year
Hispanic5345% ±6 Keep standard time all year
Asian1637% ±4 Keep standard time all year
Other or more than one race1142% ±12 Keep standard time all year

By age

Keep daylight saving time all yearKeep standard time all yearKeep changing the clocks twice a yearNo preference
Everyone30047% ±3 Keep standard time all year
18 to 243339% ±6 Keep daylight saving time all year
25 to 345645% ±5 Keep daylight saving time all year
35 to 445239% ±6 Keep daylight saving time all year
45 to 544541% ±5 Keep standard time all year
55 to 643552% ±7 Keep standard time all year
65 to 744270% ±6 Keep standard time all year
75 to 953780% ±3 Keep standard time all year

By education

Keep daylight saving time all yearKeep standard time all yearKeep changing the clocks twice a yearNo preference
Everyone30047% ±3 Keep standard time all year
No high school diploma3246% ±8 Keep standard time all year
High school diploma or GED8856% ±5 Keep standard time all year
Some college or associate's degree8141% ±5 Keep standard time all year
Bachelor's degree6542% ±5 Keep standard time all year
Graduate degree3449% ±8 Keep standard time all year

By household income

Keep daylight saving time all yearKeep standard time all yearKeep changing the clocks twice a yearNo preference
Everyone30047% ±3 Keep standard time all year
Under $25,0002663% ±9 Keep standard time all year
$25,000–49,9993960% ±7 Keep standard time all year
$50,000–74,9994455% ±7 Keep standard time all year
$75,000–99,9994048% ±8 Keep standard time all year
$100,000–149,9995641% ±6 Keep daylight saving time all year
$150,000 or more8344% ±4 Keep daylight saving time all year
Lives in group quarters1244% ±13 No preference

By party lean (modelled)

Keep daylight saving time all yearKeep standard time all yearKeep changing the clocks twice a yearNo preference
Everyone30047% ±3 Keep standard time all year
Democratic-leaning13256% ±4 Keep standard time all year
Republican-leaning12641% ±4 Keep standard time all year
Independent or no lean4240% ±4 No preference

By sex

Keep daylight saving time all yearKeep standard time all yearKeep changing the clocks twice a yearNo preference
Everyone30047% ±3 Keep standard time all year
Man14243% ±4 Keep standard time all year
Woman15852% ±4 Keep standard time all year

By work

Keep daylight saving time all yearKeep standard time all yearKeep changing the clocks twice a yearNo preference
Everyone30047% ±3 Keep standard time all year
Working18142% ±3 Keep daylight saving time all year
Unemployed, looking for work1335% ±9 No preference
Not in the labor force10665% ±4 Keep standard time all year

By segment

Keep daylight saving time all yearKeep standard time all yearKeep changing the clocks twice a yearNo preference
Everyone30047% ±3 Keep standard time all year
Segment 19178% ±2 Keep standard time all year
Segment 28144% ±2 Keep standard time all year
Segment 37961% ±2 Keep daylight saving time all year
Segment 44943% ±4 No preference

Who answered what

For each answer, the people who gave it most surely, as Jev read them. They show who answered, not why.

Keep daylight saving time all year

Mia Allen

24 years old

a woman

lives in a home their household owns

probably leans Republican (estimated from people like them and their county's 2024 vote, not stated by them)

White

Texas

probably voted in the 2024 election (estimated)

household earns $150,000 or more a year

a US citizen

never married

not a veteran

West South Central

no children at home

has a bachelor's degree

has health coverage

no disability

has a job

born in the US

82% sure

Minh Chen

31 years old

a man

lives in a rented home

probably leans Republican (estimated from people like them and their county's 2024 vote, not stated by them)

White

Florida

probably voted in the 2024 election (estimated)

household earns $100,000–149,999 a year

a US citizen

never married

not a veteran

South Atlantic

no children at home

went to some college or has an associate's degree

has health coverage

no disability

has a job

born in the US

81% sure

Jordan Thompson

65 years old

a man

lives in a home their household owns

probably leans Republican (estimated from people like them and their county's 2024 vote, not stated by them)

White

Oregon

probably voted in the 2024 election (estimated)

household earns $150,000 or more a year

a US citizen

married

not a veteran

Pacific

has children at home

has a bachelor's degree

has health coverage

no disability

has a job

born in the US

79% sure

Keep standard time all year

Leila Thomas

79 years old

a woman

lives in a home their household owns

probably leans Democratic (estimated from people like them and their county's 2024 vote, not stated by them)

White

California

probably voted in the 2024 election (estimated)

household earns $50,000–74,999 a year

a US citizen

widowed

not a veteran

Pacific

no children at home

has a graduate degree

has health coverage

has a disability

is not working or looking for work (retired, student, caring for family, or other)

born in the US

95% sure

Jordan Harris

69 years old

a woman

lives in a home their household owns

probably leans Democratic (estimated from people like them and their county's 2024 vote, not stated by them)

Black

Texas

probably did not vote in the 2024 election (estimated)

household earns $75,000–99,999 a year

a US citizen

divorced

not a veteran

West South Central

no children at home

finished high school (or a GED)

has health coverage

has a disability

is not working or looking for work (retired, student, caring for family, or other)

born in the US

93% sure

Camila Rivera

83 years old

a woman

lives in a home their household owns

probably leans Democratic (estimated from people like them and their county's 2024 vote, not stated by them)

White

Georgia

probably voted in the 2024 election (estimated)

household earns $50,000–74,999 a year

a US citizen

married

not a veteran

South Atlantic

no children at home

has a bachelor's degree

has health coverage

has a disability

is not working or looking for work (retired, student, caring for family, or other)

born in the US

92% sure

Keep changing the clocks twice a year

No one chose this.

No preference

Ava Hill

25 years old

a woman

lives in a home their household owns

probably leans to neither party (estimated from people like them and their county's 2024 vote, not stated by them)

Black

Mississippi

probably did not vote in the 2024 election (estimated)

household earns $100,000–149,999 a year

a US citizen

never married

not a veteran

East South Central

no children at home

did not finish high school

has health coverage

no disability

is unemployed and looking for work

born in the US

74% sure

Robert Martin

29 years old

a man

lives in group quarters (a dorm, barracks, nursing home, or prison)

probably leans Republican (estimated from people like them and their county's 2024 vote, not stated by them)

Black

Virginia

probably did not vote in the 2024 election (estimated)

no household income of their own

a US citizen

never married

not a veteran

South Atlantic

no children at home

went to some college or has an associate's degree

has no health coverage

no disability

is not working or looking for work (retired, student, caring for family, or other)

born in the US

72% sure

Thomas Baker

39 years old

a man

lives in group quarters (a dorm, barracks, nursing home, or prison)

probably leans Democratic (estimated from people like them and their county's 2024 vote, not stated by them)

Hispanic

Missouri

cannot vote, as they are not a US citizen

no household income of their own

not a US citizen

never married

not a veteran

West North Central

no children at home

did not finish high school

has health coverage

no disability

is not working or looking for work (retired, student, caring for family, or other)

born outside the US

66% sure

Question 2 · Yes or no

Does changing the clocks noticeably affect your sleep or your routine for a few days?

Adjusted with Claude Beta

Yes48%
No52%
By age, adjusted
  • 18 to 24: Yes 51% · No 49%
  • 25 to 34: Yes 53% · No 47%
  • 35 to 44: Yes 52% · No 48%
  • 45 to 54: Yes 51% · No 49%
  • 55 to 64: Yes 48% · No 52%
  • 65 to 74: Yes 39% · No 61%
  • 75 to 95: Yes 42% · No 58%

Jev's answers moved toward Claude's for 50 of the same people, group by group of age. On 17 held-out questions with published results, this landed about 8 to 18 points from the real result, against Jev's 21 to 34. Jev's own answers are below; hypotheses, drivers and downloads use them.

Jev's own answers

Yes58% ±1
No42% ±1

Most likely answer: Yes 95% · No 5%

Jev's average confidence 59%

± is the 95% sampling interval for these people (an effective 300 after weights): how far another draw of the same population could move a figure.

Model error is separate and wider: on 20 published questions, four answers in five landed within ±22 points of the real survey's share (±25 within groups).

Claude, 50 people: Yes 49% · No 51%

Jev and Claude are 9 points apart: some way apart.

By age: the answer where they differ most
  • 18 to 24 (7 asked): No Jev 45%, Claude 49%
  • 25 to 34 (11 asked): No Jev 44%, Claude 47%
  • 35 to 44 (9 asked): No Jev 43%, Claude 49%
  • 45 to 54 (7 asked): No Jev 43%, Claude 47%
  • 55 to 64 (5 asked): Yes Jev 57%, Claude 45%
  • 65 to 74 (5 asked): No Jev 39%, Claude 59%
  • 75 to 95 (6 asked): No Jev 33%, Claude 56%

With 50 people, gaps of a few points can be chance. A large gap means the result depends on the model: compare with real results where you can.

How firm

Each person's lean from “No” to “Yes”: the gap between Jev's probabilities for the two, in ten bands. An even result can be two camps or a crowd in the middle; this shows which.

← No0Yes →

0% firmly toward “Yes” and 0% firmly toward “No” (a gap of 60 points or more); 67% within 20 points of even.

What drives the answers

How much of the variation in people's answers lines up with each trait. Traits overlap: pay travels with role, so both can rank high for the same reason.

Age32%
Disability31%
Division / State17%
Voted in 2024 (modelled)12%
Marital status11%
Household income8%
Race and ethnicity5%
Home5%

With the other traits held constant

Each group's difference from its trait's largest group in net lean, in points, from one regression with every trait in at once, so a trait no longer gets credit for the traits that travel with it (pay with role). The dot is that effect with its 95% interval; the ring is the same difference one trait at a time, as the ranking above reads it. Where the two part, the trait's apparent effect came from others.

Age against 25 to 34

18 to 240 ±3
35 to 44-2 ±2
45 to 54-3 ±2
55 to 64-1 ±3
65 to 74+8 ±4
75 to 95+15 ±4

Disability against No disability

Has a disability+17 ±2

Children at home against No children at home

Children at home+12 ±2

Household income against $150,000 or more

Under $25,000+5 ±3
$25,000–49,999+6 ±2
$50,000–74,999+6 ±2
$75,000–99,999+7 ±2
$100,000–149,999+2 ±2
Lives in group quarters-3 ±3

Work against Working

Unemployed, looking for work-4 ±3
Not in the labor force-7 ±2
12 more traits

Division against South Atlantic

East South Central-3 ±4
Pacific+2 ±2
Mountain+2 ±3
West South Central0 ±3
New England+4 ±3
East North Central0 ±2
West North Central+2 ±3
Middle Atlantic+5 ±2

Education against High school diploma or GED

No high school diploma-3 ±3
Some college or associate's degree+1 ±2
Bachelor's degree0 ±2
Graduate degree+2 ±3

Citizenship against US citizen

Not a US citizen-4 ±2

Voted in 2024 (modelled) against Voted

Did not vote-2 ±2
Not eligible (not a citizen)-4 ±2

Sex against Woman

Man-4 ±1

Race and ethnicity against White

Black-2 ±3
Hispanic-3 ±2
Asian0 ±3
Other or more than one race-2 ±3

Party lean (modelled) against Democratic-leaning

Republican-leaning-3 ±2
Independent or no lean-2 ±3

Home against Lives in an owned home

Lives in a rented home-1 ±2
Lives in group quarters-3 ±3

Marital status against Married

Widowed0 ±4
Divorced-1 ±3
Separated+1 ±4
Never married-2 ±2

Born against Born in the US

Born abroad+1 ±3

Veteran against Not a veteran

Veteran+1 ±3

Health coverage against Has health coverage

No health coverage0 ±3

Together the traits account for 81% of the variation in people's answers (300 people).

Age and Disability together

Net lean in each combination of the two traits that drive answers most: support minus oppose, in points of average probability. Combinations of fewer than 30 people are greyed. Everyone: +17

Age \ DisabilityNo disabilityHas a disability
18 to 24+11+18
25 to 34+12+15
35 to 44+16+12
45 to 54+13+17
55 to 64+7+26
65 to 74+15+34
75 to 95+20+39

On the map

Support minus oppose (or yes minus no), from Jev's probabilities: green leans for, orange against.

Delaware: 1 person, too few to readDistrict of Columbia: 1 person, too few to readFlorida: 21 people, too few to readGeorgia: 10 people, too few to readHawaii: 1 person, too few to readIdaho: 2 people, too few to readIllinois: 11 people, too few to readIndiana: 6 people, too few to readIowa: 3 people, too few to readKansas: 3 people, too few to readKentucky: 4 people, too few to readLouisiana: 4 people, too few to readMaine: 1 person, too few to readMaryland: 5 people, too few to readMassachusetts: 6 people, too few to readMichigan: 9 people, too few to readMinnesota: 5 people, too few to readMississippi: 3 people, too few to readMissouri: 5 people, too few to readMontana: 1 person, too few to readNebraska: 2 people, too few to readNevada: 3 people, too few to readNew Hampshire: 1 person, too few to readNew Jersey: 8 people, too few to readNew Mexico: 2 people, too few to readNew York: 18 people, too few to readNorth Carolina: 10 people, too few to readNorth Dakota: 1 person, too few to readOhio: 11 people, too few to readOklahoma: 3 people, too few to readOregon: 4 people, too few to readPennsylvania: 12 people, too few to readRhode Island: 1 person, too few to readSouth Carolina: 5 people, too few to readSouth Dakota: 1 person, too few to readTennessee: 6 people, too few to readTexas: 26 people, too few to readUtah: 3 people, too few to readVermont: 1 person, too few to readVirginia: 8 people, too few to readWashington: 7 people, too few to readWest Virginia: 2 people, too few to readWisconsin: 5 people, too few to read56: not in this populationAlabama: 4 people, too few to readAlaska: 1 person, too few to readArizona: 7 people, too few to readArkansas: 3 people, too few to readCalifornia: +19 (n=35)Colorado: 5 people, too few to readConnecticut: 3 people, too few to read

Most places here have fewer than 30 people. To fill the map, build the population with “At least this many in each subgroup” set to 30: about 1,500 people for these places, weighted back to their true sizes.

−20+20Fewer than 30 people

By division

YesNo

Lean: support minus oppose, in points of average probability

Everyone30058% ±1 Yes+17±1
East South Central1756% ±3 Yes+12±6
Pacific4859% ±2 Yes+19±3
Mountain2359% ±2 Yes+19±4
West South Central3658% ±2 Yes+16±4
New England1357% ±3 Yes+14±6
South Atlantic6359% ±2 Yes+18±3
East North Central4257% ±2 Yes+14±4
West North Central2057% ±4 Yes+14±7
Middle Atlantic3860% ±2 Yes+19±4
Every place (50)
YesNo

Lean: support minus oppose, in points of average probability

Everyone30058% ±1 Yes+17±1
East South Central / Alabama458% ±8 Yes+15±16
Pacific / Alaska165% Yes+30
Mountain / Arizona758% ±4 Yes+15±7
West South Central / Arkansas357% ±10 Yes+15±20
Pacific / California3559% ±2 Yes+19±4
Mountain / Colorado563% ±6 Yes+26±13
New England / Connecticut356% ±15 Yes+13±29
South Atlantic / Delaware154% Yes+8
South Atlantic / District of Columbia154% Yes+8
South Atlantic / Florida2158% ±3 Yes+16±6
South Atlantic / Georgia1061% ±4 Yes+22±9
Pacific / Hawaii168% Yes+36
Mountain / Idaho256% ±57 Yes+13±114
East North Central / Illinois1157% ±5 Yes+13±10
East North Central / Indiana652% ±5 Yes+3±10
West North Central / Iowa358% ±20 Yes+15±40
West North Central / Kansas353% ±5 Yes+6±10
East South Central / Kentucky458% ±12 Yes+16±24
West South Central / Louisiana459% ±11 Yes+18±21
New England / Maine155% Yes+10
South Atlantic / Maryland561% ±10 Yes+22±19
New England / Massachusetts659% ±6 Yes+17±11
East North Central / Michigan959% ±5 Yes+18±9
West North Central / Minnesota561% ±6 Yes+23±12
East South Central / Mississippi351% ±7 Yes+1±14
West North Central / Missouri554% ±15 Yes+9±30
Mountain / Montana161% Yes+22
West North Central / Nebraska256% ±6 Yes+11±13
Mountain / Nevada358% ±18 Yes+15±35
New England / New Hampshire160% Yes+20
Middle Atlantic / New Jersey859% ±4 Yes+18±8
Mountain / New Mexico260% ±83 Yes+19±165
Middle Atlantic / New York1859% ±4 Yes+19±7
South Atlantic / North Carolina1060% ±4 Yes+20±8
West North Central / North Dakota151% Yes+2
East North Central / Ohio1159% ±4 Yes+19±8
West South Central / Oklahoma363% ±4 Yes+27±8
Pacific / Oregon452% ±5 Yes+4±11
Middle Atlantic / Pennsylvania1260% ±4 Yes+20±8
New England / Rhode Island157% Yes+14
South Atlantic / South Carolina559% ±4 Yes+17±8
West North Central / South Dakota163% Yes+26
East South Central / Tennessee656% ±6 Yes+13±12
West South Central / Texas2658% ±3 Yes+15±6
Mountain / Utah359% ±10 Yes+18±20
New England / Vermont150% Yes0
South Atlantic / Virginia856% ±6 Yes+11±12
Pacific / Washington761% ±2 Yes+23±4
South Atlantic / West Virginia263% ±32 Yes+25±64
East North Central / Wisconsin555% ±5 Yes+10±10

By race and ethnicity

YesNo

Lean: support minus oppose, in points of average probability

Everyone30058% ±1 Yes+17±1
White18759% ±1 Yes+18±2
Black3356% ±2 Yes+13±4
Hispanic5357% ±2 Yes+14±3
Asian1655% ±3 Yes+11±5
Other or more than one race1161% ±4 Yes+22±8

By age

YesNo

Lean: support minus oppose, in points of average probability

Everyone30058% ±1 Yes+17±1
18 to 243355% ±1 Yes+11±3
25 to 345656% ±1 Yes+12±3
35 to 445257% ±1 Yes+15±3
45 to 544557% ±2 Yes+13±3
55 to 643557% ±2 Yes+13±4
65 to 744261% ±2 Yes+21±4
75 to 953767% ±2 Yes+34±4

By education

YesNo

Lean: support minus oppose, in points of average probability

Everyone30058% ±1 Yes+17±1
No high school diploma3256% ±3 Yes+13±6
High school diploma or GED8859% ±1 Yes+19±3
Some college or associate's degree8158% ±1 Yes+15±3
Bachelor's degree6558% ±1 Yes+16±3
Graduate degree3460% ±2 Yes+20±4

By household income

YesNo

Lean: support minus oppose, in points of average probability

Everyone30058% ±1 Yes+17±1
Under $25,0002659% ±3 Yes+18±7
$25,000–49,9993960% ±2 Yes+20±5
$50,000–74,9994460% ±2 Yes+20±4
$75,000–99,9994059% ±2 Yes+18±4
$100,000–149,9995657% ±1 Yes+13±2
$150,000 or more8358% ±1 Yes+16±2
Lives in group quarters1252% ±6 Yes+4±11

By party lean (modelled)

YesNo

Lean: support minus oppose, in points of average probability

Everyone30058% ±1 Yes+17±1
Democratic-leaning13259% ±1 Yes+18±2
Republican-leaning12658% ±1 Yes+17±2
Independent or no lean4256% ±2 Yes+12±3

By sex

YesNo

Lean: support minus oppose, in points of average probability

Everyone30058% ±1 Yes+17±1
Man14257% ±1 Yes+14±2
Woman15859% ±1 Yes+19±2

By work

YesNo

Lean: support minus oppose, in points of average probability

Everyone30058% ±1 Yes+17±1
Working18158% ±1 Yes+16±1
Unemployed, looking for work1354% ±2 Yes+8±4
Not in the labor force10660% ±2 Yes+20±3

By segment

YesNo

Lean: support minus oppose, in points of average probability

Everyone30058% ±1 Yes+17±1
Segment 19163% ±1 Yes+27±2
Segment 28158% ±1 Yes+17±2
Segment 37956% ±1 Yes+13±2
Segment 44952% ±2 Yes+5±3

Who answered what

For each answer, the people who gave it most surely, as Jev read them. They show who answered, not why.

Yes

Linh Clark

80 years old

a woman

lives in a home their household owns

probably leans Republican (estimated from people like them and their county's 2024 vote, not stated by them)

White

New York

probably did not vote in the 2024 election (estimated)

household earns under $25,000 a year

a US citizen

married

not a veteran

Middle Atlantic

no children at home

finished high school (or a GED)

has health coverage

has a disability

is not working or looking for work (retired, student, caring for family, or other)

born in the US

75% sure

Betty Adams

75 years old

a woman

lives in a rented home

probably leans Democratic (estimated from people like them and their county's 2024 vote, not stated by them)

of another race or more than one race

California

probably voted in the 2024 election (estimated)

household earns under $25,000 a year

a US citizen

never married

not a veteran

Pacific

has children at home

finished high school (or a GED)

has health coverage

has a disability

is not working or looking for work (retired, student, caring for family, or other)

born in the US

73% sure

Aaliyah Rivera

81 years old

a woman

lives in a home their household owns

probably leans Republican (estimated from people like them and their county's 2024 vote, not stated by them)

White

New York

probably voted in the 2024 election (estimated)

household earns $75,000–99,999 a year

a US citizen

married

not a veteran

Middle Atlantic

no children at home

finished high school (or a GED)

has health coverage

has a disability

is not working or looking for work (retired, student, caring for family, or other)

born in the US

73% sure

No

Thomas Baker

39 years old

a man

lives in group quarters (a dorm, barracks, nursing home, or prison)

probably leans Democratic (estimated from people like them and their county's 2024 vote, not stated by them)

Hispanic

Missouri

cannot vote, as they are not a US citizen

no household income of their own

not a US citizen

never married

not a veteran

West North Central

no children at home

did not finish high school

has health coverage

no disability

is not working or looking for work (retired, student, caring for family, or other)

born outside the US

64% sure

Robert Martin

29 years old

a man

lives in group quarters (a dorm, barracks, nursing home, or prison)

probably leans Republican (estimated from people like them and their county's 2024 vote, not stated by them)

Black

Virginia

probably did not vote in the 2024 election (estimated)

no household income of their own

a US citizen

never married

not a veteran

South Atlantic

no children at home

went to some college or has an associate's degree

has no health coverage

no disability

is not working or looking for work (retired, student, caring for family, or other)

born in the US

60% sure

Nancy Flores

56 years old

a woman

lives in a home their household owns

probably leans Republican (estimated from people like them and their county's 2024 vote, not stated by them)

Asian

Illinois

cannot vote, as they are not a US citizen

household earns $150,000 or more a year

not a US citizen

married

not a veteran

East North Central

no children at home

has a bachelor's degree

has health coverage

no disability

is not working or looking for work (retired, student, caring for family, or other)

born outside the US

57% sure

How this was made

Who they are: Adults living in the United States in 2024.

What they were told before answering: Most of the country moves its clocks forward an hour in March and back in November. Congress has considered keeping one time all year: either permanent daylight saving time (lighter evenings, darker winter mornings) or permanent standard time (lighter mornings, earlier sunsets).

Population built from the “United States by state: Census 2020–2024” template: 300 invented people, seed 2f13515f. Each person's traits come from the population's stated proportions; Jev (jev-1.13.0) judged how each would answer from those traits and from the context shown here, nothing else. Shares average each person's chance of each answer (Jev's full distribution), which lands closer to real surveys than counting each person's most likely answer; those counts are shown beside them.

Two kinds of error

Sampling: the ± beside each figure is a 95% interval for the people in this population, using their effective number after weights (Kish). It says how much the figure could move with another draw of the same population, nothing more. Rows too small to read (under 30 people, or an interval wider than ±15 points) are greyed out and left out of what stands out.

Model: how far real surveys land from Studio, measured on published questions with known results. Four answers in five landed within ±22 points of the real share on 20 questions that are not scales, and within ±23 on 24 scales (the average level within ±15 points of 100). This error is not chance and does not shrink with more people; it is why a contrast between groups is safer to read than a single percentage.

Traits marked (modelled) are estimated, not recorded: no Census record says how anyone votes. Each person's party lean and 2024 turnout come from Cooperative Election Study respondents like them, calibrated so every county matches its certified 2024 presidential result and turnout. Fine for including or comparing groups; not anyone's stated party, and not a poll.

Synthetic simulation: these answers come from Jev's judgement of invented people, not from a survey.

Ending the twice-yearly clock change · Voxari Studio