Four statistics we will not use
- 300,000 strokesA 2025 PNAS paper modelled permanent standard time preventing roughly 300,000 strokes and 2.6 million obesity cases. It is the most quotable number in our favour and we will not touch it. Two independent groups published critical letters, PNAS issued a correction, and the fuller methodological critique identifies a sign reversal in the model's source code that inverted the US east-west axis — meaning county health data was correlated with circadian estimates belonging to counties on the opposite side of the time zone. If you see this figure on a permanent-standard-time site, that site has not done its reading.
- +24% heart attacksReal, from a Michigan registry, and routinely misrepresented. The same paper reports a 21% drop after the fall transition and states plainly that the total weekly number of heart attacks did not change. It is a shift in timing among already-vulnerable patients, not additional disease. The defensible pooled figure is about 4%.
- Sleepy judgesThe widely-cited finding that judges hand down 5% longer sentences after the spring change was reanalysed in the same journal, and the effect did not survive. We leave it out.
- Public supportWe cannot claim the public wants permanent standard time. Two of the three best recent polls say the opposite. What the polling supports is that about two thirds want the switching to stop, and that the answer to the second question flips depending on how you ask it.
Circadian-informed modeling predicts regional variation in obesity and stroke outcomes under different permanent US time policies. PNAS. 2025;122(38):e2508293122. The source of the widely-quoted 300,000-stroke figure. Subject to a published correction and three critical letters.
pubmed.ncbi.nlm.nih.gov/40953265/
The sum of absolute circadian shifts: Questioning the metric linking daylight saving time policy to stroke and obesity. PNAS. 2026;123(17):e2532075123. Argues the summed-absolute-shift metric cannot distinguish a stable circadian phase from one that drifts monotonically.
pubmed.ncbi.nlm.nih.gov/41996167/
The full methodological critique, and the document that actually identifies the code-level defect: the longitudinal offset was divided by Ω = −15°/hour instead of +15°/hour, which 'effectively inverted the US East-West axis within the model,' cross-correlating each county's health data against a geographically flipped counterpart.
arxiv.org/html/2606.19541
Challenges of county-level circadian modeling for time-policy debates. PNAS. 2026;123(17):e2527389123. The second of the two critical letters, raising ecological-inference objections.
pubmed.ncbi.nlm.nih.gov/41996169/
Correction for Weed & Zeitzer. PNAS. 2026;123(18):e2612351123.
pubmed.ncbi.nlm.nih.gov/42044348/
Sandhu A, Seth M, Gurm HS. Daylight savings time and myocardial infarction. Open Heart. 2014;1(1):e000019. +24% Monday after spring, −21% Tuesday after fall, and NO difference in the total weekly count — a shift in timing, not incidence.
pmc.ncbi.nlm.nih.gov/articles/PMC4189320/
Are Sleepy Punishers Really Harsh Punishers? Comment on Cho, Barnes, and Guanara (2017). Psychol Sci. 2018;29(6):1006–1009. Reanalysis finds no significant 'sleepy Monday' sentencing effect.
pubmed.ncbi.nlm.nih.gov/29683778/
The three papers that hurt this argument most
The 2026 systematic review. 157 studies across 36 countries in the European Journal of Epidemiology, the newest and largest synthesis in the field. Its conclusions cut against us directly: "the messaging of transitions and DST during summer months being uniformly detrimental is not supported," and living with daylight time appears associated with decreased all-cause mortality and traffic accidents in summer. It calls for a balanced approach rather than abolition.
Our answer, and you should weigh it yourself: the review largely compares summer daylight time against summer standard time. The policy on the table is year-round daylight time with January mornings, which is not the comparison it makes. That is a real distinction. It is not a refutation.
The 2025 BMJ cohort. 683,809 people in England over eleven years found little evidence of change after the spring clock change across cardiovascular disease, road injuries, anxiety, depression or self-harm — and the autumn change was associated with fewer events.
The 2025 JAMA Network Open registry. 168,870 heart-attack patients across 1,124 hospitals: no significant difference in the daylight-time week, spring or fall. The German MONICA/KORA registry found the same null.
And a structural problem with the whole literature
Across thirteen well-known transition studies, the largest effect sizes come from the smallest samples — the classic signature of small-study bias. The author concludes that a roughly 5% spring increase explains all of them, and that anything larger is an artefact. We have kept our transition numbers close to that ceiling rather than under it: the pooled heart-attack figure is 4%, and Fritz's crash estimate is 6%, which sits just above it. We have quoted no transition figure larger than that.
A systematic review of epidemiological studies into daylight-saving time & health identifying beneficial & adverse effects. Eur J Epidemiol. 2026. 157 studies, 36 countries. Concludes the messaging of transitions and summer DST being uniformly detrimental 'is not supported' and calls for a balanced approach.
pubmed.ncbi.nlm.nih.gov/41721993/
de Lange MA, Birnie K, Richmond RC, et al. Acute effects of daylight saving time clock changes on mental and physical health in England: population based retrospective cohort study. BMJ. 2025. 683,809 people, 2008–2019: little evidence of change after the spring clock change; the autumn change was associated with fewer events.
pubmed.ncbi.nlm.nih.gov/41412603/
Rymer JA, Li S, Chiswell K, Kansal A, Nanna MG. Daylight Savings Time and Acute Myocardial Infarction. JAMA Netw Open. 2025;8(9):e2530442. 168,870 patients across 1,124 hospitals, 2013–2022: no significant difference in AMI incidence in the DST week, spring or fall.
pubmed.ncbi.nlm.nih.gov/40924425/
Are daylight saving time transitions associated with changes in myocardial infarction incidence? BMC Public Health. 2015;15:778. German MONICA/KORA registry, n=25,499: no significant overall change after either transition.
pmc.ncbi.nlm.nih.gov/articles/PMC4535383/
Sample size bias in the empirical assessment of the acute risks associated with daylight saving time transitions. Chronobiol Int. 2023;40(2). Across 13 well-known studies the largest effect sizes come from the smallest samples; a ~5% spring increase suffices to explain all of them.
pubmed.ncbi.nlm.nih.gov/36594375/
What is actually solid
| Claim | Confidence | Why |
|---|---|---|
| Clock changes have real acute costs | High | Two meta-analyses agree on ~4–5%; large registries dispute even that |
| Daylight time saves no meaningful energy | High | Meta-analysis, Indiana, Australia, DOE all converge |
| Standard time matches solar time better | Certain | Definitional |
| No scientific body backs permanent daylight time | High | We searched adversarially and found none |
| Permanent daylight time gets reversed | High | Four for four at high latitude |
| Chronic daylight time causes population harm | Moderate | Border studies are observational; schedules would adapt |
| The public wants permanent standard time | Not supported | Polls split by framing; two of three go the other way |