Daylight Saving in Aotearoa:
Three Clock Futures

Is New Zealand really scrapping daylight saving? No — but the question is live. This map lets you see what each option actually does to your mornings and evenings, region by region, with the health, safety, and energy evidence weighed honestly.

✓ 14 citations verified, none retracted NOAA solar algorithm Stats NZ regional data computed 2026-09-16

First: is DST being scrapped?

No. The official schedule (govt.nz, updated Aug 2026) runs through 2029: clocks forward at 2am on the last Sunday of September, back at 3am on the first Sunday of April. What exists is two live petition streams at Parliament — one to abolish DST (PET_125604), one from the lobby group Take Back the Clocks for permanent DST — and neither has been adopted. There has been no select committee inquiry, referendum, or government review, unlike the US (Sunshine Protection Act, passed the House, stalled in the Senate) or the EU (2018 vote to end clock changes, never implemented).

⚠️ Misinformation flag: blog posts claiming "the government has cancelled daylight saving from 2026" are false and circulating. The official schedule stands. If you've seen that claim, this is your debunking.
Why a map? Because geography does most of the talking. All of New Zealand sits west of the 180° meridian that anchors NZST (UTC+12) — the country's populated east coast is effectively "west-of-zone," the mirror image of the US, where the worst clock-change effects cluster. Latitude compounds it: at 46°S, Southland's winter days are short enough that either permanent option bites hard.

The map: three futures, your region

Pick a clock policy and a measure. Darker red means more of the thing you probably don't want; green means less of it. Click any region for its full profile.

Click a region

or hover to preview

Chatham Islands (not shown — a territory, not a region) run 45 minutes ahead of the mainland: UTC+12:45 standard, UTC+13:45 daylight.

What it means nationally

Population-weighted days per year, across all 16 regions (Stats NZ subnational estimates, ~2024 vintage):

days with sunrise ≥ 7am — status quo
days with sunrise ≥ 7am — permanent NZST
days with sunrise ≥ 7am — permanent NZDT
days with sunset < 6pm — status quo
days with sunset < 6pm — permanent NZST
days with sunset < 6pm — permanent NZDT

The trade in one sentence: permanent NZDT buys light evenings (sunset stays past 6pm nearly all year) at the price of ~57 extra pre-7am-dark mornings per person per year; permanent NZST does the reverse; the status quo splits the difference by paying for it with two clock-change jolts a year. Whether the jolts or the dark mornings cost more is exactly what the evidence below is about.

Regional extremes (computed)

Under the status quo, days with sunrise ≥ 7am range from 116 (Gisborne) to 222 (Southland). Under permanent NZDT, every region gets more of them: from 195 (Gisborne) to 238 (Southland), whose winter-solstice sunrise slides to 09:31. The country's latest midsummer sunset under any scenario is Southland's 21:39 (20:39 under permanent NZST) — Auckland's is 20:40 under the status quo and 19:40 under permanent NZST, trading away an evening hour to protect its winter mornings.

The evidence — spikes are real, so are the counterweights

Measured in New Zealand

Road accidents rise at spring clock-onset. The first southern-hemisphere study — 12.6 million ACC accident claims, 2005–2016 — found the spring start of DST associated with +16% road accidents on the first day (Sunday) and +12% on Monday. Falls and home/community accidents shifted before the transitions, suggesting people anticipate the change. Work accidents showed limited impact.

Robb & Barnes (2018). Accident Analysis & Prevention 111:193–201. doi:10.1016/j.aap.2017.11.029 ✓ verified

Health & safety spikes (international)

• Heart attacks: meta-analysis of 7 studies (>115,000 subjects): AMI risk +5% (OR 1.05, 1.02–1.07) in the two weeks after spring onset; no significant change after autumn. Manfredini et al. (2019). J Clin Med 8(3):404. doi:10.3390/jcm8030404 ✓

• Stroke: Finnish nationwide registry: ischemic-stroke hospitalizations RR 1.08 (1.01–1.15) in the first two days after either transition. Sipilä, Ruuskanen, Rautava, Kytö (2016). Sleep Medicine 27–28:20–24. doi:10.1016/j.sleep.2016.10.009 ✓

• Fatal crashes: 732,835 US fatal crashes 1996–2017: +6% in the week after spring onset — worse in the west of each time zone, the same "west-of-zone" geometry NZ's populated east coast has. Fritz, VoPham, Wright, Vetter (2020). Current Biology 30(4):729–735.e2. doi:10.1016/j.cub.2019.12.045 ✓

• Sleep → injuries: on the Monday after spring onset, US workers sleep 40 min less and sustain more, and more severe, workplace injuries. Barnes & Wagner (2009). J Applied Psychology 94(5):1305–1317. doi:10.1037/a0015320 ✓

• The one clear benefit: the autumn clock-back extends sleep; German hospital data (160M admissions) show cardiovascular admissions fall by ~10 per day per million for about four days after fall-back. Jin & Ziebarth (2020). J Econ Behavior & Organization 170:174–192. doi:10.1016/j.jebo.2019.12.003 ✓

The counterweights (why this isn't settled)

• The largest review to date — 157 studies, 36 countries (Steponenaite et al. 2026, Eur J Epidemiol) — concludes the "uniformly detrimental" framing is not supported: onset transitions associate with more acute MI and fatal accidents but fewer crimes of physical harm; offset transitions with lower all-cause mortality and fewer workplace accidents. doi:10.1007/s10654-026-01372-8 ✓

• Spring onset cuts robberies ~7% (ambient-light mechanism; the 2007 US extension ≈ $59M/yr in avoided social cost). Doleac & Sanders (2015). Rev Econ & Statistics 97(5):1093–1103. doi:10.1162/REST_a_00547 ✓

• A Dutch PCI registry found no association between DST transitions and heart-attack interventions at all — the null result the headlines skip. Derks et al. (2021). Netherlands Heart Journal 29:427–432. doi:10.1007/s12471-021-01566-7 ✓

• The sleep-medicine position (AASM: permanent standard time) and the recreation/economics case (NZ's 2007 extension was justified on evening use) are a values question, not a settled fact — which is precisely what a toggleable map should let you explore.

Energy — the original rationale, dead or dying

The Indiana natural experiment (counties switching to DST for the first time): residential electricity rose ~1% overall, 2–4% in autumn — heating/lighting in dark mornings outweighed evening savings. Kotchen & Grant (2011). Rev Econ & Statistics 93(4):1172–1185. doi:10.1162/REST_a_00131 ✓

Australia's Sydney-Olympics DST extension: no conservation, and a prominent simulation model overstated savings. Kellogg & Wolff (2008). J Env Econ & Management 56(3):207–220. doi:10.1016/j.jeem.2008.02.003 ✓

Meta-analysis of 44 studies (162 estimates): mean 0.34% savings, only meaningful at high latitudes — NZ qualifies geographically, but no NZ-specific estimate exists. Havranek, Herman, Irsova (2018). The Energy Journal 39(2):35–61. doi:10.5547/01956574.39.2.thav ✓

What we could not do (and won't fake): public Waka Kotahi CAS crash data carries year-level granularity only (no day-level dates), so we could not re-derive the NZ accident spike from open data — the +16%/+12% figures are Robb & Barnes' published results, quoted from their abstract. And no NZ-specific myocardial-infarction or stroke study exists; cardiovascular claims ride on international evidence and are labeled as such.

A brief history (it's more interesting than you'd think)

1895 — Entomologist George Hudson proposes shifting clocks to get more after-work daylight for insect collecting. The idea is New Zealand's own.
1927 — Sir Thomas Sidey's Summer Time Act passes: clocks forward one hour. Public reaction is so hostile it's cut to half an hour in 1928.
1941–45 — Wartime regulations keep "summer time" year-round; the Standard Time Act 1945 locks UTC+12 permanently — meaning NZ's current standard time is its former summer time.
1974 — Time Act re-establishes the modern DST hour, after decades of half-hour experiments.
2007 — Extended to today's 27 weeks after a petition and public consultation; a 2008 survey found 82% approval. The stated rationale: evening recreation and retail — not energy.
2023–26 — Two live Parliament petitions (abolish vs. permanent-DST). Neither adopted; the 2008 approval level suggests abolition is the harder sell.

Methods, data & provenance

Sun times: NOAA solar-calculator spreadsheet algorithm (implemented in Python, validated against timeanddate.com within ±3 min), computed per day over calendar 2026 for each region's largest city. Scenarios: keep (NZST/NZDT per the Time Act 1974 calendar — last Sunday September to first Sunday April), abolish (UTC+12 year-round), permanent (UTC+13 year-round).

Boundaries: Stats NZ Regional Council 2023 (Generalised), CC-BY 4.0, simplified (Douglas-Peucker, ~0.005° tolerance) for the web. Populations: Stats NZ subnational estimates (~2024 vintage, as retrieved Oct 2025) — used for the population-weighted national means only. Representative-city caveat: a region's largest city doesn't capture intra-regional spread (e.g. Queenstown vs Dunedin); the map is indicative, not a per-address forecast.

Citations: all 14 references verified via Crossref/OpenAlex structured checks on 2026-09-16; none retracted. Effect sizes quoted only from primary abstracts or authors' own pages, never from memory.

Full citation ledger (verified 2026-09-16)
1. Robb & Barnes (2018) Accid Anal Prev — 10.1016/j.aap.2017.11.029 ✓
2. Sipilä, Ruuskanen, Rautava, Kytö (2016) Sleep Med — 10.1016/j.sleep.2016.10.009 ✓
3. Barnes & Wagner (2009) J Appl Psychol — 10.1037/a0015320 ✓
4. Smith (2016) AEJ: Applied — 10.1257/app.20140100 ✓
5. Fritz, VoPham, Wright, Vetter (2020) Curr Biol — 10.1016/j.cub.2019.12.045 ✓
6. Manfredini et al. (2019) J Clin Med — 10.3390/jcm8030404 ✓
7. Jin & Ziebarth (2020) J Econ Behav Organ — 10.1016/j.jebo.2019.12.003 ✓
8. Kotchen & Grant (2011) Rev Econ Stat — 10.1162/REST_a_00131 ✓
9. Kellogg & Wolff (2008) JEEM — 10.1016/j.jeem.2008.02.003 ✓
10. Havranek, Herman, Irsova (2018) Energy J — 10.5547/01956574.39.2.thav ✓
11. Doleac & Sanders (2015) Rev Econ Stat — 10.1162/REST_a_00547 ✓
12. Derks et al. (2021) Neth Heart J — 10.1007/s12471-021-01566-7 ✓
13. Steponenaite, Wallraff, Wild, Brown (2026) Eur J Epidemiol — 10.1007/s10654-026-01372-8 ✓
14. Hurst, Morfeld, Lewis, Erren (2024) Dtsch Arztebl Int — 10.3238/arztebl.m2024.0078 ✓
Smith (2016) is cited for the fatal-crash sleep mechanism (+30 deaths/yr, $275M social cost); the +6% figure on the page is Fritz et al.'s, not Smith's.