The Law of Time

Most of us outside Saskatchewan put our clocks back an hour yesterday, and we’ve now returned to what some might call “God’s time”. Of course, when it comes to the o’clock, it’s actually the law that disposes, and the law’s been setting our watches backwards and forwards for just over a hundred years. At the beginning of the last century, the English builder, William Willet, found a champion in Parliament to get his scheme passed for recapturing “some of the hours of wasted sunlight in the spring, summer, and autumn.”

Perhaps fearing that a jump of a full hour would discombobulate people, Willet originally proposed that time be changed in four steps of twenty minutes each, on the first four Sunday’s in April—and then back again in the same four steps in the same way in September. Reason prevailed and the compromise of a one-hour change was reached. The proposed bill read as follows:

And, as we all know only too well, once the law has sat you down in the client chair, it’s not easy to get up and get away. So it has proved for time. That fickle hour has been dancing to legislatures’ tunes ever since — albeit sometimes reluctantly, as when the United States fairly recently decided to prolong Daylight Saving Time until the first Sunday in November and, after grumbling, Canada shuffled into step.

The relevant statutes are brief, yet interesting to look at.

Ontario, for instance, doesn’t shilly-shally but gets right to the point and calls the statute the Time Act. (There’s a whiff of hubris about this, it seems to me; but then I prefer that to an alternative such as “An Act to Improve the Mornings of the Ordinary People of this Province Thanks to the Beneficent Leadership of the Current Government,” a tendentious style that seems to have caught on here in the last decade or so.) Disappointing after such a brisk and bold title, section 1 evidences most of the sins of lawyerly English, though:

Where an expression of time occurs in any Act, proclamation, regulation, order in council, rule, order, by-law, agreement, deed or other instrument, heretofore or hereafter enacted, made or executed, or where any hour or other point in time is stated either orally or in writing, or any question as to time arises, the time referred to or intended shall, unless it is otherwise specifically stated, be held to be the time in effect as provided by this Act.

I found it interesting that Daylight Saving Time is set (in the only other section, i.e., 2) with reference to “Greenwich time.” But any system needs a zero-point, and since the English invented DST, it’s only fair they get to be it. Amusing, too, that in Quebec the very same zero point is, more scientifically perhaps, and certainly voided of geographic nuances, called “coordinated universal time” (as to which see Wikipedia). Also, Quebec, Pascal-like, hedges its bets and calls the act, modestly, the “Legal Time Act,” — also leaving room for “les bons temps” to “rouler” too.

Alberta’s act is very much like Ontario’s (or vice versa), referring to “Greenwich mean time” as the way to measure daylight saving time. Unlike Ontario, however, Alberta has clearly got some fractious regions to deal with, because it feels it necessary to say (quite clumsily) that no municipality or Metis settlement may “proclaim, enforce, use” etc. daylight saving time other than that set out in the act.

But when it comes to temporal complexity — as expressed in legislation, at least — nothing compares to the province that has as an unofficial but great motto: “hard to spell, easy to draw.” I’m talking of Saskatchewan, of course. It turns out that it’s only most of the province that doesn’t do the temporal two step, and so it takes a bit of verbiage to carve out the exceptions in its Time Act — 31 sections of it.

Oh, and you’ll observe, please, that it’s daylight saving time — no plurals anywhere — it’s the law.


  1. The legalese in s. 1 of Ontario’s Time Act dates from the 19th century’s establishment of standard time. It divides the province into the zones within Eastern time and within Central time.

    Daylight saving time was not regulated by provincial statute in Ontario until the extension in 1986. Before then it was governed sometimes by local by-law and more often by local practice.

    I recall listening to Windsor radio in the 1960s (CKLW), and Windsor kept on standard time all summer because Detroit did.

    I also recall the CNR – then still responsible for passenger trains as well as freight – serenely ignoring the time change for mere
    mortals and staying on Standard time all year round.

    There are a couple of remote (from us) communities in northern Ontario near the Eastern/Central boundary that are non-standard in their observance of daylight time (I think they don’t do it). I guess the police don’t go in to lay charges under the Act – and the Act does not create any offences anyway.

    There was some thought of referring to UTC in 1986 but it was a very rare term at the time (and IMHO remains so) so the Act stuck with Greenwich. (Romantic anglophilia, perhaps.)

    It makes no sense – and made no sense in 2006 – to stretch daylight time into November. If we (read the US) needed four weeks of extension of daylight time, all four weeks should have been added in the spring. We still would have more weeks of daylight time after the summer solstice than before.

    Some people argued that if it were still daylight time into November, then kids could go out trick-or-treating before dark, and thus be safer. This of course was complete rubbish, since the whole point of trick-or-treating is to be out after dark (and not just to have the cover of darkness for one’s tricks…) (For more comment on popular delusions about Hallowe’en risks, see Bruce Schneier’s blog.)

    When the US extended daylight time in 2006, there was talk of studying whether the measure actually saved any energy, and if not, perhaps the extension would be reversed. I have heard of no such study since then, though I would be interested to know if any was done.

    Recently the sleep scientists have been warning of the deleterious effects on our alertness caused by the extra hour of sleep lost or gained at the transition to and from standard to daylight time. This seems just about as well founded as the Hallowe’en motive mentioned earlier. I expect very large numbers of people go to bed and get up later on weekends than during the working week, by more of a margin than the single hour of the time change. So the measurable, much less socially significant, effect of the time change would be infinitesimal, while the added pleasures of long light summer evenings are substantial, if not quantifiable.

  2. The “God” in “God’s time” was of course Sir Sandford Fleming, who was responsible for the implementation of Standard time. Before that, every community estimated its own time according to its perception (usually) of when the sun got to its highest point. Before speedy longish-distance travel, i.e. before the railways, this didn’t matter, because one spent so long getting from one place to another that one expected to have to check the time when one got there. (The accuracy of one’s portable timepiece no doubt had something to do with that too.)

    So there is nothing particularly natural, much less God-given, about what time it would be if we did not do daylight time. Standard time is made for people’s convenience, just like Daylight time, though the convenience measure is different.

  3. It’s not clear that all 4 weeks added should have been added in the spring. It’s not desired that DST be symmetric about the summer solstice. The constraint is that we don’t want the sun to rise too late in the morning. Due to the elliptical orbit of the earth, the sun rises about 25 minutes earlier in October as compared with March (see Thus, there are more months of DST before the solstice than after the solstice.

  4. Aaron’s note shows that it’s more complicated than it may look!

    However, under the present rules (daylight saving time starts on the second Sunday in March and ends on the first Sunday in November), there are at least 14 weeks of daylight time before the solstice and at least 19 weeks after the solstice. (It can be up to a week more depending on what date the relevant Sunday falls on, reading the solstice always as June 21.)

    If all four additional weeks allocated to daylight time in 2007 had been added before the solstice, the balance would have been 15 weeks before and 18 weeks after.

    Days lengthen or shorten on average about 16 minutes a week, so the balance would still have given lighter mornings in March than October. This graph shows that the sunrise in early March is earlier than the sunrise at the end of October, with or without daylight time.

    (I think Aaron’s last line is intended to be ‘therefore it is appropriate that there are more months of DST after the solstice than after the solstice.’ If not, I don’t understand it.