Time changing image licensed from iStockphoto.
 

Today, March 13, most of Canada switches from standard time to daylight saving time. On November 6, 2016 we will switch back. It is an annual ritual that repeats every second Sunday of March and every first Sunday of November. Alone in Canada, only the province of Saskatchewan remains on Central Time year round. Is Daylight Saving Time (DST) worth the trouble? Or should the rest of Canada follow Saskatchewan's lead and abandon DST switchovers?

The history of daylight saving time is checkered. The idea of DST has been credited to Benjamin Franklin (1784), who had argued that being asleep during sunlit summer morning hours would lead to a waste of tallow and wax during summer evenings. Some Canadian cities used DST as early as 1908 (Thunder Bay) and 1914 (Regina). Germany introduced DST nation-wide in 1916, ostentatiously to save fuel for the country's war efforts. The United States followed two years later in 1918, but the experiment ended the same year. Between 1942 and 1945, the Untied States and Canada adopted year-round DST, which was known as "war time". Modern DST came to the United States in 1966 through its Uniform Time Act, and Canada—where time zones remain a provincial matter—closely coordinated its own time zones with the United States. The most recent change came in 2007 following the U.S. Energy Policy Act of 2005, which extended DST to roughly nine months. In 1988, Newfoundland also experimented with "double DST", but abandoned this policy after a public backlash. My wife, who was a teenager in Newfoundland at the time, recounts her unhappiness about fireworks on Canada Day not starting until midnight! DST is used in over 70 countries around the world, but beginning and ending are poorly coordinated internationally. Fortunately, our smartphones have saved us from "temporal confusion" when we travel internationally during DST-switching season.

Why bother with daylight saving time? Even though most of my clocks and computers adjust their time automatically, many simpler devices still have to be adjusted manually. Most importantly, our own bodies do not adjust instantaneously. Today we have to get up an hour earlier—or sleep in because it is Sunday and shift the difficult adjustment to Monday morning. We are "borrowing" one of hour sleep when time switches from 2am to 3am today, only to get it "paid back" in November. Does it leave us better off as individuals? Probably not. Perhaps it just leaves us a little more grumpy in March and a little more rested in November. But are we better off as a society, reducing our energy consumption as a result of DST? Do societal benefits outweigh societal (and personal) costs? The answer may surprise you. Rigorous empirical research has found few if any benefits, and some very significant costs. Perhaps it is time to reconsider abandoning DST and stick to standard time or year-round daylight saving time.

Whether you like it or not, our Canadian choices are intimately tied to what happens south of the border. And there, a movement to abandon daylight saving time is gathering momentum. As the Washington Post reported on March 12, 2015, a number of states are on their way to abolish daylight saving time. And as the Globe and Mail reported in February 2015, Washington State mulls abolishing daylight savings, British Columbia's neighbour was considering to stay on standard time year round. The proposal in the Washington State House died without a vote, however. United States law allows states to opt out of daylight saving time, and Arizona and Hawaii already do. Bills in other states are still moving forward. Back home, B.C.'s Ministry of Justice is not eager to move alone on this issue and pointed to the benefits of coordination with other provinces and U.S. states.

The raison d'être for DST is energy conservation. So what exactly is the evidence that DST is actually delivering energy savings? When the United States passed the Energy Policy Act of 2005, it was argued that each extra day of DST would save 100,000 barrels of oil per day. But where was the evidence? Empirically, one can observe energy consumption with and without DST, but the counterfactual observation for comparison is not available except in rare circumstances. Kotchen and Grant (2011) found such a natural experiment in the state of Indiana between 2004 and 2006. Most counties in Indiana did not observe DST until the new DST policies were adopted in 2007. The evidence showed that DST resulted in a 1% increase(!) in residential electricity demand, not a reduction. The effect was particularly pronounced in the fall. DST saved electricity on illumination, but increased electricity used for hearing. Kotchen and Grant found that DST costs Indiana households an average of US$3.29 per year in increased electricity costs. Kellog and Wolff (2008) used a quasi-experiment in Australia, where DST was extended in 2000 during the Sydney Olympics. They found that the extension of DST did not reduce overall electricity consumption but led to intra-day shifts in the pattern of electricity consumption. Aries and Newsham (2008) surveyed the literature on DST's effect on lighting energy use, which is only one part of the overall effect of DST on energy use. Some of the earlier, slightly positive results that were found in the literature may be waning due to the much superior efficiency of lighting today, as well as better home insulation. There are some studies that find small benefits from maintaining DST year-round, such as Hill et al. (2010) for Great Britain. The research that is conducted at the macroeconomic and household level is increasingly backed up also with evidence from behavioural data—how individuals change their time use in response to DST shifts. Sexton and Beatty (2014) find evidence that individuals are shifting potentially energy-intensive activities earlier into the day as a result of DST changes, which points towards increased energy usage overall.

The story does not end with a look at whether or not energy savings were realized. There is a real and measurable cost to DST switching. There is a significant body of empirical research of the effect of DST on motor vehicle accidents, going back decades. Insufficient sleep and disrupted circadian rhythms can cause accidents. Therefore, the spring shift could contribute to accidents. Researchers have looked at the Monday after the DST switch and compared it to the Monday a week later. There is consistent evidence that your risk of being involved in a motor vehicle accident the Monday after the DST switch is about 6-8% higher than normal (Varughese and Allen, 2001). Other researchers have also looked at pedestrian fatalities and motor vehicle occupant fatalities; Lambe and Cummings (2000) find a 11% increased incidence in Sweden. On the other hand, there may be benefits from scrapping DST shifts. Coate and Markowitz (2004) found that full-year DST would reduce pedestrian fatalities by 13% in morning and afternoon time windows, while motor vehicle occupant fatalities would drop by 3%. There may well be other adverse health effects. For example, Sandhu et al. (2014) report a 24% increase in acute myocardial infarctions (AMIs) on the Monday after DST, closely followed by a 21% reduction on Tuesday. While this does not seem to raise the overall incidence of AMIs, hospitals should be aware that the Monday after DST in spring will be busier than usual. Psychologists also point to a loss of productivity as a result of the spring shift: Wagner et al. (2012) find evidence of increased "cyberloafing"— you know, surfing the net when you should be focusing on your work.

The quirks with DST go beyond driving and health. Even in the world of finance, DST shifts leads to peculiar side effects. In 2000, my Sauder School of Business colleague Maurice Levi, together with Mark Kamstra and Lisa Kramer, found that financial markets may be impacted by DST shifts, magnifying the usual "weekend effect" by 200 to 500 percent. Their calculation suggested that in the United States alone, the DST effect implied a one-day loss of $31 billion on the major stock exchanges. Unsurprisingly, their suggestion has led to an excited academic dispute about whether this effect may have other explanations (Steigerwald and Conte 2007; Berument, Dogan and Onar, 2010). Rebuttals and rejoinders (Kamstra, Kramer, and Levi, 2010 and 2013) confirm the continued existence of the DST-shift stock market effect.

Lastly, let us look at some statistics about how much daylight saving time is out of sync with astronomical time. A good measure for this is to compare the time when the sun reaches its peak in the sky ("solar noon") with the time when your clock says its 12 noon ("local noon"). The difference in minutes is not fixed over the course of the year because of the equation of time—the effects of Earth's elliptical orbit around the sun and the tilt of Earth's rotational axis. You can use the NOAA Solar Calculator to compute solar noon and sunrise and sunset for any given day and any location on the planet. The table below shows the time deviation of solar noon and local noon for major Canadian cities for four days of the year: the two solstice days (June, December) and the two equinox days (March, September). It also shows the time zones and whether the city uses DST. Saskatoon has the largest deviation because it is, effectively, on year-round DST. Time in St. John's, Newfoundland is pretty much on target thanks to its half-hour special time zone. As they say in an iconic TV ad, Newfoundland is always half an hour ahead of the rest of the country.

Notes: Time zones are P(acific), M(ountain), C(entral), E(astern), A(tlantic), and N(ewfoundland). Saskatchewan does not switch to daylight saving time. Negative numbers indicate that local noon occurs after solar noon.

Stefano Maggiolo has prepared a beautiful map that shows how much solar time and standard time are out of sync around the world; you can find more discussions on his blog How much is the time wrong around the world?

So where does this analysis leave us? Considering the mounting scientific evidence, daylight saving time does not survive a rigorous cost-benefit analysis. If we are to abandon DST—perhaps along with our US neighbours—where should we go? Adopt year-round standard time or adopt year-round DST? Our workdays typically run 9am to 5pm, centered around 1pm. So perhaps we should aim to have local noon around 1pm, which means year-round daylight saving time. The available statistical evidence also suggests that if any energy savings are to be gained, it is with year-round DST. I, for one, would be happy to avoid the DST switches every year. As I often travel across Canada's time zones, I should perhaps be used to time zone changes, but why have two more than absolutely necessary? Perhaps it's time for us to consider a policy change, especially if the United States is moving in the same direction. And drive carefully on Monday after today's DST switch!

Perhaps the strongest opposition to DST switching comes from the television industry. As Brian Handwerk reported in National Geographic in 2013, "the Nielsen ratings during the first week of daylight saving, no matter when it is, even the most popular shows go down by 10 to 15 percent in viewership." Now here is a reason to get rid of DST: don't let DST interfere with your TV viewing habits. You may miss a crucial episode of your favorite TV show. Oh well, that must have been in the era before Netflix and iTunes.