2 of 7 Mixed Fact-check · 7 claims
Photo: Caitlinator / CC BY 2.0 · source
A viral post says nearly all research favors less homework, later school starts, more vacation and shorter shifts. Two hold up well, one partly, and homework runs the other way.
By The Crosscheck Desk · 2026-09-23
A widely shared post, circulating as a screenshot with about 68,000 likes
“love how basically every study says that students do better in school with less homework & later start times and workers do better with more vacation days & shorter shifts, but we do the exact opposite in this godforsaken country because Puritanical feelings apparently beat facts”
We could not locate the original post. The account it is attributed to is not viewable on X to logged-out readers, the public mirror services we tried are offline, archived copies of that account that we could read did not contain it, and exact-phrase web searches found no copy. This article checks the wording as it circulates. We read “this country” as the United States, because the complaint about vacation days fits no other rich country: every other one in the comparison we cite guarantees paid vacation by law.[17]
Students do better in school with less homework
For elementary pupils the main US meta-analysis finds homework time barely linked to achievement. For high schoolers the link is positive. The case for less holds only for very heavy loads and for stress.
Students do better in school with later start times
Later starts reliably add sleep, and the pediatricians’ academy endorses 8:30 a.m. Evidence on grades is thinner: a Cochrane review called it very low quality.
Workers do better with more vacation days
A meta-analysis finds a small boost to health and well-being from vacation that fades soon after work resumes. We found no strong evidence tying more days to productivity.
Workers do better with shorter shifts
Very long hours and 24-hour shifts carry measured harms to health and safety. Below about 49 hours a week, output tracks hours, and the four-day-week trials rely on volunteer firms.
The United States does the exact opposite on all four
True for vacation, which no federal law requires, and largely true for start times. US hours sit above the OECD average but far from the top. Homework loads were flat from 1984 to 2012, the last year Brookings analysed, with heavy loads a minority.
Implied: the research is settled and uniform, and basically every study agrees on all four
The reviews disagree with each other on homework, rate start-time evidence as low quality, and find vacation effects short-lived.
Implied: US policy on these four is driven by Puritanical feelings rather than cost, logistics or law
Some economics links Protestant values to longer hours. Other work credits taxes and labor regulation, and the school rollbacks we found cited buses and budgets.
The benchmark synthesis is Harris Cooper and colleagues’ 2006 review of US research since 1987. It found that all studies had design flaws, but that there was generally consistent evidence for a positive influence of homework on achievement, with a stronger correlation in grades 7–12 than in K–6.[1] Under the review’s random-error model, the correlation between time on homework and achievement was r = .20 for secondary students and r = .05 for elementary students, which was not significantly different from zero.[1] The same review adds a warning at the top end: even for the oldest students, too much homework may diminish its effectiveness, or even become counterproductive.[1]
A later meta-analysis of math and science homework from 1986 to 2015 found an overall small and positive relationship between homework and achievement. It found the relationship stronger for elementary and high school students than for middle school students, and strongest in studies of US students.[2] That contradicts Cooper on the elementary years, so even the two main reviews do not agree on which students benefit.
International data point to diminishing returns. The OECD’s PISA in Focus note on homework reports that evidence from PISA 2009 suggests that after around four hours of homework per week, the additional time invested in homework has a negligible impact on performance.[3] The same note puts the United States among systems in which students spend more than six hours per week, on average, doing homework.[3]
The strongest evidence for less homework concerns well-being, not grades. A survey of 4,317 students at 10 high-performing high schools in upper middle class communities found they averaged more than 3 hours of homework per night, and those who did more homework reported more academic stress, physical health problems, and lack of balance in their lives.[4] That sample is affluent and self-selected by school type, so it describes heavy loads, not the typical student.
Verdict: Misleading. “Do better in school” is an achievement claim. For high schoolers the reviews associate more homework with higher achievement, up to a point. The post’s version holds for young children’s test scores (per Cooper, not per Fan), for loads past the point of diminishing returns, and for stress.
This is where the post’s claim is strongest. In 2014 the American Academy of Pediatrics said the evidence strongly implicates earlier school start times (before 8:30 am) as a key modifiable contributor to insufficient sleep, and urged middle and high schools to aim for start times that allow 8.5–9.5 hours of sleep.[6]
A meta-analysis of five longitudinal and 15 cross-sectional comparison group studies found that later starting school times are associated with longer sleep durations, less daytime sleepiness and less tardiness to school, while its authors cautioned that methodological considerations lead to a cautious interpretation.[10] When Seattle delayed its secondary school start time by nearly an hour, a pre-/post-study measured an increase in daily median sleep of 34 minutes, associated with a 4.5% increase in median grades and an improvement in attendance.[11]
The academic payoff is less certain than the sleep payoff. The Cochrane systematic review found later school starts associated with an increase in school-night sleep, but said later start times may provide academic benefits with mixed findings, that the quality of the evidence was very low, and that it could not determine the effects of later school start times with any confidence.[9] RAND modelled a statewide shift to 8:30 a.m. and projected an $83 billion contribution to the US economy after a decade; that is a projection built on the literature, not a measured outcome.[12]
Verdict: Mostly True. The sleep evidence is consistent and backed by a major medical body. The claim that students do better in school, meaning grades, rests on fewer and weaker studies.
The research here is thin. A meta-analysis of seven vacation studies found that vacation has positive effects on health and well-being (a small effect, d = +0.43), but that these effects soon fade out after work resumption.[19] Those studies track workers before, during and after a vacation; they do not test whether workers with 25 days do better than workers with 15, which is the policy question the post raises.
We found no meta-analysis or systematic review linking the number of paid vacation days to worker productivity. Cross-country comparisons that pair long statutory vacations with high output per hour exist in commentary, but they cannot separate vacation from the many other ways those economies differ.
Verdict: Mixed. Time off improves well-being briefly. “More days, better workers” is plausible and largely untested.
The World Health Organization and International Labour Organization concluded from their systematic reviews that there is sufficient evidence for higher risks of ischemic heart disease and stroke among people working long hours (55 or more hours a week) compared with standard hours of 35–40 hours a week, and estimated 745,194 deaths attributable to this exposure in 2016.[21]
On shift length, a randomized study of interns found nearly 36% more serious medical errors and 5.6 times more serious diagnostic errors on a traditional schedule with shifts of more than 24 hours in a row than on shorter shifts.[22] A cross-sectional survey of 31,627 registered nurses in 12 European countries found that nurses working shifts of 12 hours or more were more likely than those working 8 hours or less to experience burnout, job dissatisfaction and intention to leave.[24]
Not every trial points the same way. A national cluster-randomized trial of 117 US surgical residency programs found that flexible duty-hour policies, which waived rules on maximum shift lengths, produced noninferior patient outcomes and no significant difference in residents’ satisfaction with overall well-being.[23]
On output, John Pencavel’s study of wartime munition workers found the relationship nonlinear: below an hours threshold, output is proportional to hours, and above it output rises at a decreasing rate.[20] In his estimates, below 49 weekly hours output is proportional to hours, and output at 70 hours differs little from output at 56 hours.[20] That supports cutting very long weeks; it does not show that cutting a 40-hour week raises output.
The largest pilot, in the UK in 2022, included roughly 2,900 workers at 61 companies. Employees reported less work-related stress and lower burnout, revenue increased by an average of 1.4% among the 23 organizations that provided data, and 56 of the 61 companies said they would continue.[25] A 2025 analysis of 2,896 employees across 141 organizations in six countries found improvements in burnout, job satisfaction, mental health and physical health, a pattern not observed in 12 control companies.[28]
The limits are real. Michael Sanders of King’s College London pointed out that such trials are based on self-selection, so the companies agreeing to take part might be those where a four-day week would be taken up enthusiastically.[27] Iceland’s widely reported trials did not involve a four-day week: workers moved from a 40-hour to a 35- or 36-hour week, the maximum reduction was four hours, and the work units involved volunteers.[26]
Verdict: Mostly True. The evidence against very long hours and 24-hour shifts is strong. For modest cuts below a standard week, the well-being gains come from volunteer firms, and the output evidence is weaker.
“Exact opposite” implies US policy pushes each lever the wrong way. Lever by lever:
| Lever | What the evidence supports | US practice | “Opposite”? |
|---|---|---|---|
| Homework | Little benefit in early grades; diminishing returns at heavy loads | Loads flat 1984–2012; heavy loads a minority[5] | No |
| Start times | 8:30 a.m. or later for teens | Most high schools start before 8:30[8] | Mostly |
| Vacation | Well-being gains, short-lived | No federal requirement[16] | Yes |
| Hours | Avoid 55+ hour weeks and 24-hour shifts | Above OECD average, below the top[29] | Partly |
Tom Loveless’s analysis of NAEP long-term trend data for Brookings found that, with one exception, the homework load remained stable from 1984 to 2012. The exception is 9-year-olds, whose share with no homework fell.[5] For 1984–2012, the share reporting two hours or more ranged from 5–6% for age 9, 6–10% for age 13, and 10–13% for age 17.[5] In national polls cited in the same report, parents were more likely to say their children have too little homework than too much.[5] Against that, US 15-year-olds reported more homework than the OECD average in PISA 2012, above the four-hour mark past which the OECD saw negligible gains.[3]
In 2011–12, the average start time across public middle, high and combined schools was 8:03 a.m., and only 17.7% started at 8:30 a.m. or later.[7] In 2020–21 the average start time for public secondary/high schools was 8:07, with 28.2% starting at 8:30 a.m. or later, which leaves 71.8% starting earlier.[8] The two surveys cover different school sets, so the change between them is not a clean trend.
California’s SB 328 requires the schoolday for high schools to begin no earlier than 8:30 a.m. and for middle schools no earlier than 8:00 a.m., by July 1, 2022.[13] Florida passed a similar requirement in 2023, then rolled it back: SB 296 repeals the 2023 mandate that middle schools start no earlier than 8 a.m. and high schools not before 8:30 a.m., after local officials warned of logistical nightmares.[14] The governor approved SB 296 on 21 May 2025, and it deems a district in compliance once it submits a report to the Department of Education.[15] As of September 2026 that remains the law we could find; we found no later change.
The Fair Labor Standards Act does not require payment for time not worked, such as vacations.[16] CEPR’s comparison of 21 rich countries calls the United States the only advanced economy that does not guarantee its workers paid vacation and holidays, while EU workers are legally guaranteed at least 20 paid vacation days a year.[17] Employers fill part of the gap: in March 2025, 80% of private industry workers had access to paid vacation, according to the Bureau of Labor Statistics.[18]
OECD data put average annual hours actually worked per US worker at 1,800 in 2025, against an OECD average of 1,736, about 3.7% higher.[29] Germans worked 1,332 hours, about 468 fewer than Americans, while Mexico, Korea and Greece all worked more than the United States.[29] The United States is a long-hours country among rich democracies, but not the longest.
Verdict: Mixed. Vacation and start times fit the post. Hours fit in part. Homework does not: US loads have not risen, and most students report under two hours a night.
“Basically every study says” is a claim about a whole literature, so we tested it against the reviews and major bodies for each lever, not single studies.
Compared against: homework, Cooper, Robinson & Patall’s 2006 synthesis of US research, 1987–2003 primary · meta-analysis; Fan et al.’s 2017 meta-analysis of math and science homework, 1986–2015 primary · meta-analysis; OECD PISA in Focus No. 46 primary · official data. Start times: the AAP’s 2014 policy statement primary · professional body; Marx et al.’s 2017 Cochrane systematic review primary · systematic review; Bowers & Moyer’s 2017 meta-analysis primary · meta-analysis. Vacation: de Bloom et al.’s 2009 meta-analysis primary · meta-analysis. Hours and shifts: the WHO/ILO 2021 joint estimates built on their systematic reviews primary · systematic review; Pencavel’s munitions study primary · peer-reviewed; the 2025 Nature Human Behaviour four-day-week analysis primary · peer-reviewed. Single studies (Seattle, the intern and surgical-resident trials, the nurse survey, the high-performing-schools survey) are cited for illustration, not counted toward consensus. Think-tank and campaign reports (RAND, Brookings, CEPR, the UK pilot organizers) are secondary; news and commentary are popular. Not checked: reviews of homework in non-math subjects after 2006, and reviews of statutory leave and productivity, which we could not find.
Scored against that set:
The consensus claim holds for one lever (start times, on sleep), partly for a second (hours, at the long end), is untested for a third, and runs backwards for the fourth.
False. “Basically every study” tells readers the debate is over. The two leading homework meta-analyses disagree about elementary students.[1][2] The Cochrane review of start times rated its evidence very low quality.[9] A randomized trial of surgical residents found no patient harm from relaxing shift-length limits.[23] A reader who takes the post at its word would expect a unanimity the literature does not show.
Unproven. No study we found tests whether Puritan attitudes set US homework, school-bell or leave policy. Some evidence points toward culture. Using a sixteenth-century treaty to separate religion from other factors in Germany, Jörg Spenkuch found that Protestants appear to work longer hours than Catholics, and that religious differences in individuals’ values can account for most of the effect.[32]
Other evidence points elsewhere. Edward Prescott found that Americans worked about 50 percent more than Germans, French and Italians, a reversal from the early 1970s when Western Europeans worked more, and attributed most of the gap to marginal tax rates on labor income.[30] Alberto Alesina, Edward Glaeser and Bruce Sacerdote disputed the tax explanation and the culture one: they wrote that tax rates can explain only a small amount of the difference in hours, that Europeans worked more than Americans as late as the 1960s, and that European labor market regulations advocated by unions explain the bulk of the difference.[31] A gap that opened within living memory is hard to pin on settlers who arrived in the 1600s.
The school examples we found cite money and logistics. Florida’s rollback followed warnings of logistical nightmares for larger districts, which were already facing a bus driver shortage and would have needed to hire even more drivers.[14] RAND notes that a major argument against later starts is the cost of rescheduling bus routes.[12] On homework, most parents in national polls call their children’s load about right.[5] Vacation policy runs through labor law: the FLSA leaves vacation to agreements between employers and employees.[16] Whether those choices reflect an underlying work ethic is a fair question, but the post states the answer as fact.
The evidence supports two of the post’s policy asks on firm ground. Later secondary start times would give teenagers more sleep, a point on which the AAP and the reviews agree, and policies that curb 55-plus-hour weeks and 24-hour shifts address harms the WHO and ILO have quantified.[6][21] The obstacle to later starts, on the record, is logistics and cost: Florida’s rollback came after warnings that districts already facing a bus driver shortage would need even more drivers.[14] An argument aimed at bus budgets would engage that obstacle more directly than one aimed at Puritanism.
The homework and vacation parts of the argument lean on the overstated consensus. A campaign to cut high school homework on the premise that “every study” backs it would meet reviews linking homework to higher secondary achievement.[1] The defensible version is narrower: little homework for young children, caps on very heavy loads, and attention to stress in high-pressure schools.[3][4] On vacation, the case for a statutory floor rests more on the international comparison, where the US is an outlier, than on productivity research, which is thin.[17][19]
What would change these verdicts: a randomized or well-controlled study showing that cutting secondary homework raises achievement; higher-quality trials showing grade gains from later starts, which would lift that rating to True; or a four-day-week trial with random assignment of firms, an element the 4 Day Week Foundation says it would like to add in future trials.[27]
New York’s DoorDash settlement is real and, by our count, the largest by any US city labor office. The investigation behind it began under Mayor Adams, and no court ruled.
Three of the four claims in the viral McConnell graphic hold up on the roll-call record. The fourth fails on one word: the Senate has no paid leave for him to have taken.
Trump said America has too many holidays, costing billions. US law mandates zero paid holidays and zero paid vacation, alone among 21 rich countries. Japan guarantees 15.
New Crosscheck fact-checks, every other Monday. Sourced, never spun.
