More on Circadian Disruption Day

[unixronin]

My basic contention:  While the rationale for DST is energy savings, I suggest that in today's lifestyles and home/work environments, any actual energy savings are negligible if present at all; further, that the disruption of schedules is not only frustrating, but dangerous, as people who have just gotten used to driving to work in daylight are suddenly not only doing it in the dark again, but doing so while their body's circadian rhythm says they should still be asleep.

Having done some followup checking on effects of DST, I find the following highlights dealing with adverse effects of DST.

Energy savings:

From Wikipedia (see the article for references):

• The U.S. Dept. of Transportation concluded in 1975 that DST might reduce the country's electricity usage by 1% during March and April, but the National Bureau of Standards reviewed the DOT study in 1976 and found no significant savings.
• In 2000 when parts of Australia began DST in late winter, overall electricity consumption did not decrease, but the morning peak load and prices increased.
• In Western Australia during summer 2006–07, DST increased electricity consumption during hotter days and decreased it during cooler days, with consumption rising 0.6% overall.
• Although a 2007 study estimated that introducing DST to Japan would reduce household lighting energy consumption, a 2007 simulation estimated that DST would increase overall energy use in Osaka residences by 0.13%, with a 0.02% decrease due to less lighting more than outweighed by a 0.15% increase due to extra cooling; neither study examined non-residential energy use. DST's effect on lighting energy use is noticeable mainly in residences.
• A 2007 study found that the earlier start to DST that year had little or no effect on electricity consumption in California.
• A 2007 study estimated that winter daylight saving would prevent a 2% increase in average daily electricity consumption in Great Britain.
• A 2008 study examined billing data in Indiana before and after it adopted DST in 2006, and concluded that DST increased residential electricity consumption by 1% to 4%, primarily due to extra afternoon cooling.
• The U.S. Dept. of Energy concluded in a 2008 report that the 2007 U.S. extension of DST saved 0.5% of electricity usage during the extended period.

General conclusion on energy saving:  Although done for reasons of energy savings, recent studies appear to show that at best there are little or no net energy savings as a result of DST, and at worst, increased energy usage for air conditioning outweighs energy savings from reduced usage of electric lighting.  I observe that this effect will become more pronounced as time goes by, because as we switch over to more energy-efficient light sources (CFL and LED vs. incandescent), light is becoming cheaper in watts per lumen by a factor of 4 or more, while the energy cost of residential cooling remains unchanged.

Traffic and other accident rates:

From INABIS 1998:

[...] some researchers have claimed that society is chronically sleep deprived, and even small additional reductions in sleep time may have consequences for safety (see Coren, 1996a for a review).  Coren (1996b) apparently confirmed this by showing that the shift to Daylight Savings Time (DST) had an impact on accident rates.  The spring shift to daylight savings time results in a loss of one hour of sleep while the fall shift provides an additional hour which can be used for sleep.  Using data from two years of Canadian traffic accident records, he found that on the Monday following the shift to DST in the spring, there was an increase in traffic accident rates of about 7 percent, while in the fall there was a decrease in accident rate of about the same magnitude.

Coren (1996c) attempted to confirm this finding using accidents other than those associated with vehicle operation.  He did this by looking at every accidental death in the US reported to the National Center for Health Statistics for the years 1986 through 1988.  Since over 80% of accident induced mortality occurs within 4 days of the accident, data for analysis was restricted to the first 4 workdays in the weeks preceding, immediately following, and one week after, the DST change.  While he found a significant increase in deaths following the spring shift (6.6%) he could not confirm the fall rebound, which only showed a nonsignificant (1.5%) decrease.  Other studies that have used traffic accidents have found the increase following the spring shift to DST but have also failed to find any decrease following the fall time change (e.g. and Hicks, Lindseth & Hawkins, 1983; Monk, 1980).

The rest of the paper then goes on to expand upon this and discuss methodology.  In short, it finds that not only do traffic fatalities increase by around 7% during the week following the change to DST, but that non-traffic-related accidental fatalities also increase by around 6.6% during that week, and that there is no corresponding reduction of accident rates during the week following the end of DST.  The paper's authors also found that the majority of the increase in accidents occurred later in the day, indicating that the primary cause of increased accident rate is tiredness due to sleep deprivation, not driving to work in the dark.

Taken together then, these data are consistent with the hypothesis that as a society [we] are sufficiently chronically sleep deprived so that a small decrease in sleep duration, such as that which occurs with the spring shift to DST, can significantly increase accident susceptibility.

While traffic accident rates to not appear to suffer a "rebound" drop at the end of DST, other studies indicate that DST is a major hazard to pedestrians.  This SpacingToronto article cites a CMU study that found that during the first two weeks after the end of DST, there is a 186% increase in pedestrian fatalities.

“The change that’s going to occur on Sunday is going to have some pronounced effects on your risks of walking between 5 p.m. and 7 p.m.,” Dr. Gerard said last night.  “Basically, these are the hours when it’s just getting dark.  Next week at this time, it will be pitch black.  But people walking and people driving won’t have adjusted.  The baseline risk for getting killed is almost tripled.”

Their study of pedestrian fatalities from 1999-2005 shows that there is an average of 37 more U.S. pedestrian deaths around 6 p.m. in November compared to October.  That amounts to an increase of 186 percent.

Their report indicates that after the initial spike in November, pedestrian fatality rates decline each month until DST begins again, at which point there is a 78% drop.  The risk of a pedestrian being killed in a traffic accident at 6pm just after the end of DST in November is 11 times higher than at the same time just after DST ends.  Overall, Fischbeck and Gerard found that between 1999 and 2005, the end of DST was associated with 250 extra pedestrian fatalities, vs. 139 fewer associated with the beginning of DST.  A 2001 University of Michigan study reported similar results.

A paper in Psychiatric Times by Stanley Coren, head of the Human Neurology and Perception laboratory at UBC (presumably the same Coren referenced above), goes beyond that to say (at the top of page 2):

A microsleep is a short period of time, usually between 10 seconds to a minute in length, in which the brain actually enters a sleep state, regardless of what the person is doing at the time.  The affected individual often does not know that this momentary blackout has occurred.  The effects of these microsleeps combined with attentional lapses, however, can be dramatic.

There is now evidence that many major disasters have been due to sleep-debt related effects.  The evidence shows that these include the oil spill of the Exxon Valdez, the nuclear accidents at Chernobyl and Three Mile Island, and the loss of the space shuttle Challenger (Coren, 1996a).

General conclusion on safety:  The end of DST is dangerous to pedestrians.  The beginning of DST is dangerous to everyone else.  The mechanism doesn't work exactly as I'd supposed, but the eventual result is the same; the extra accidents just occur in the evenings, not the mornings.

(Additional note:  According to hrrunka, some UK studies have shown increases in traffic fatalities after both the spring and fall DST transitions.)

Economic cost:

Wikipedia says many commercial sectors claim increased business during the extended evening daylight of DST.  This can only go so far, though, as people only have just so much money to spend.  Further, it seems only reasonable to assume that increased business for vendors on summer evenings entails increased fuel consumption for people to get there.

In any case, this economic benefit has a cost as well.  Referring back to Wikipedia again:

[...]  Conversely, DST can adversely affect farmers and others whose hours are set by the sun.  For example, grain harvesting is best done after dew evaporates, so when field hands arrive and leave earlier in summer their labor is less valuable.  DST also hurts prime-time broadcast ratings and drive-in and other theaters.

Clock shifts and DST rule changes have a direct economic cost, entailing extra work to support remote meetings, computer applications and the like.  For example, a 2007 North American rule change cost an estimated $500 million to $1 billion.

(It is noted that the estimated numbers depend on the methodology, and the results have been disputed.)

[...]  People who work across time zone boundaries need to keep track of multiple DST rules, as not all locations observe DST or observe it the same way.  The length of the day becomes variable.  Disruption to meetings, travel, broadcasts, billing systems, and records management is common, and can be expensive.  During an autumn transition from 02:00 to 01:00, a clock reads times from 01:00 to 02:00 twice, possibly leading to confusion.

Some computer-based systems require downtime or restarting when clocks shift; ignoring this requirement damaged a German steel facility in 1993.  Medical devices may generate adverse events that could harm patients, without being obvious to clinicians responsible for care.  These problems are compounded when the DST rules themselves change, as in the Year 2007 problem.  Software developers must test and perhaps modify many programs, and users must install updates and restart applications.  Consumers must update devices such as programmable thermostats with the correct DST rules, or manually adjust the devices' clocks.

From Wikifun (in addition to much of what is cited above):

DST's twice-annual shifts in recorded time cause legal and business-operational complications, as shown in the following examples.  During a North American time change, a fall night during which clocks are reset from 2 a.m. DST to 1 a.m. Standard Time, times between 1 a.m. and 2 a.m. will occur twice, causing confusion in transport schedules, payment systems, etc.

[...]

People who work nights often have an extra hassle logging how many hours they worked, since it will be either one hour more or one hour less than the simple difference in start/stop times.

DST is particularly unpopular among people working in agriculture because they must rise with the sun regardless of what the clock says, and thus the people are placed out of synchronization with the rest of the community, including school times, broadcast schedules, and the like.

An editor's footnote to the Psychiatric Times article quoted above also notes that a 1995 report by the US Dept of Transportation found that 100,000 sleep-related traffic accidents result in 1,500 deaths per year in the US, and that a National Commission on Sleep Disorders Research report estimated that sleep-related accidents and sleep disorders which impact work productivity cost the US economy between $100 billion and $150 billion per year.  I was unable to find data on those to determine whether there is a spike in that economic cost at the beginning of DST comparable to the already-documented spike in traffic accidents, but it seems a reasonable expectation.

Health:

Most of the studies above limit themselves to a week or so after DST transitions.  This Finnish paper on Bio-Med Central, though, written by researchers at the University of helsinki, the Helsinki University Central Hospital, and the Finnish National Public Health Institute, found that clock shifts disrupt sleep and reduce its efficiency; that effects on seasonal adaptation of the circadian rhythm can be severe and last for weeks; and (perhaps unsurprisingly) that the effects are more detrimental to persons already suffering from mood disorders or circadian-rhythm sleep disorders.  Interestingly, the Finns found a larger disruption in the fall transition than in spring:

Our main finding was that transitions into and out of daylight saving time disrupted night-time sleep. The movement and fragmentation index was increased significantly after both transitions, on average by 54% in fall and 37% in spring. There is a small reservation while using movement and fragmentation index as it still is slightly unclear parameter [3]. However our experience is that movement and fragmentation index reflects well the quality of sleep. Unexpectedly, sleep efficiency was reduced after the transition out of DST only. The durations of slow-wave sleep stages tends to decrease and those of rapid-eye-movement sleep to increase during winter [4]. Transitions out of DST may affect sleep efficiency more than those into DST due to these underlying changes in sleep stages.

[...]

We found that the transition out of DST was more detrimental to individuals with the preference to morning activities.  Also Kantermann et al. noticed in their recent study that the timing of activity does not adjust to the DST imposition in spring, especially in late chronotypes.  This was unexpected, since the fall transition brings one more hour light to the mornings.  However, this option to have earlier light exposure may not materialize at all or not be enough to maintain the daily rest-activity cycle and night-time sleep undisturbed.  To visualize, the sunrise to sunset times were 8:35 to 17:32 prior to the transition, and 7:38 to 16:29 one day and 7:40 to 16:26 two days after the transition.

In contrast, the transition into DST affected more those with the preference to evening activities.  This may have been due to the longer exposure to light in the evening which is known to delay the phase position of the circadian rhythms and their subsidiary rest-activity cycles.  To visualize, the sunrise to sunset times were 6:07 to 18:46 prior the transition, and 7:04 to 19:49 one day and 7:01 to 19:51 two days after the transition.

The latter finding — that the transition into DST affected evening activities more than morning activities — is consistent with Coren's thesis that there is a sleep debt issue involved, and that while people cope relatively well earlier in the day, by evening sleep deprivation is taking a heavy toll.

Coren's paper in general says that we, as a society, already operate continually in a state of sleep debt, and that the invention of effective artificial lighting cost us around 500 hours per year of sleep.

Confirmation of these natural sleep durations comes from Palinkas, Suedfeld and Steel (1995).  These researchers spent a summer above the arctic circle where there is continuous light 24 hours a day.  All watches, clocks and other timekeeping devices were removed, and only the station's computers tracked the times that the team went to sleep and awakened.  Individual researchers did their work, and chose when to sleep or wake according to their "body time."  At the end of the experiment, they found that their overall average sleep daily time was 10.3 hours.  Every member of the team showed an increase in sleep time, with the shortest logging in at 8.8 hours a day, and the longest almost 12 hours a day.  This study, like many others, seems to suggest that our biological need for sleep might be closer to the 10 hours per day that is typical of monkeys and apes living in the wild, than the 7 to 7.5 hours typical of humans in today's high-tech, clock-driven lifestyle.

Psychological researchers have tended to minimize the effects of sleep insufficiency, acknowledging that society may be getting too little sleep, but treating the effects of this sleep deprivation as nothing more significant than an inconvenience which makes people feel a bit tired now and then.

Or, to sum it up:  Our industrialized, artificially-lit lifestyle is bad for us, and deliberately disrupting everyone's circadian cycles twice a year only makes it worse.

Overall, we appear to have the following to work with:

• Currently, and increasingly as we shift to more energy-efficient lighting technologies, there appears to be little or no net reduction of power consumption associated with DST, and some evidence indicates there is fact a net increase in consumption several times greater than the Department of Energy's claimed energy savings.
• There appears to be a well-documented spike in both traffic fatalities and other accidental fatalities associated with the disruption of sleep patterns at the beginning of DST.  This spike does not appear to be balanced by any corresponding dip at the end of DST, indicating that it is related to sleep deprivation, not to sudden loss of morning daylight.
• The sudden loss of evening daylight at the end of DST, however, does appear to be related to a significant spike in pedestrian fatalities, which is only partially countered by a dip in fatalities associated with the beginning of DST.
• Implementation of DST, and adaptation of business operations and industrial systems to DST, causes significant additional complexity and economic cost, and even more so when DST schedule are changed.  This effect is considerably greater when coordination across multiple time zones with different DST schedules is involved.

Conclusion

"Your mileage may vary", but personally I find the evidence strongly supportive of a conclusion that the costs of implementing daylight savings time outweigh its claimed benefits.  Kazakhstan, for one, apparently came to the same conclusion, and ceased observing daylight savings time in 2005.  Maybe the rest of the world should consider following their lead.

Comments

[rbos.livejournal.com]

Here's a map of countries that do and don't use DST. Note that Saskatchewan and Manitoba don't.

http://www.worldtimezone.com/daylight.html

[rbos.livejournal.com]

pardon; Saskatchewan and part of Alberta. I notice that part of the USA also ignores it.

[dr-nebula.livejournal.com]

I HATE DST with a super-nova-esque intensitiy. It totally borks with my astro-imaging, to the point it is hardly worth the trouble to do during the work week.

[otherbill.livejournal.com]

Yep, you should move to Arizona.

Or parts of Indiana, although I'm not sure they still ignore it.

[ithildae.livejournal.com]

The area's near Chicago observe DST. The other side of the state goes to a different time zone. The center of the state remains on CST year round. It gets confusing if you have multiple stores in Indiana.

[unixronin.livejournal.com]

Nope, my understanding is all of Indiana now obeys DST. The holdout counties switching over was what gave the opportunity for one of the studies — it enables them to compare power consumption vs. the surrounding counties both before and after adopting DST, and found net power consumption actually went up.

.....Or, reading the comment above, that was what I THOUGHT the change was, but if that's correct it's much more complicated than I was aware.

[ithildae.livejournal.com]

I hadn't heard of the change. My last information is some years old. All I knew is that I was the IT guy responsible for keeping the times right on the system for all stores. It was not straightforward, nor automatic.

[ithildae.livejournal.com]

Last I counted, there are 37 time zones in the world. Some locals are offset by 30 minutes, and there are one or two that are offset by 15 minutes. States like Indiana can have up to three active time zones at once. (These things all cost lots of money for business to keep up with.)

The striking thing for me is, cows don't care what time it is, they just want to be milked at the same time every day. Field work on farms is dependent on sunrise. Astronomy only cares about the leap seconds and leap years to keep the sky in sync. With all due respect to Ben Franklin, he blew it with this DST idea. (He also blew it with the direction of current flow, causing EE's and Physicists to insert a minus between each other's papers.)

I vote we follow the airlines, everything is UTC. Just figure it out based on where you live. You have no idea the confusion and expense you will save in international businesses. (And a few interstate ones as well.)

[unixronin.livejournal.com]

I vote we follow the airlines, everything is UTC. Just figure it out based on where you live. You have no idea the confusion and expense you will save in international businesses. (And a few interstate ones as well.)

Exactly what I've been saying for some years now. So that means local noon is at 18:50? So what? It's just a number. Who cares, as long as it's consistent?

[rbos.livejournal.com]

I'd even like to take that one step further, and go for the "seconds since 1970" measure of time. Then, we can get rid of leap seconds, hours, and days, in addition.

100 kiloseconds is about 27 hours. 1 megasecond is about a third of a month. 1 gigasecond is about 30 years.

Those are not inconvenient units to work with. I'd love to have a 27 hour day.

I think that won't happen until about 200 years after the solar day becomes irrelevant, though. :P

[unixronin.livejournal.com]

We should probably switch to a 64-bit counter at the same time, lest the 32-bit counter become institutionalized and everyone panics as 2038 comes up.

I personally would get along fine with a 27-28 hour day, but I think you're right about the likelihood of it happening while we're a planetary civilization.

[ithildae.livejournal.com]

Sounds like the French and their metric obsession. They had ten day weeks, ten hour days, ten minute hours... They still payed attention to solar realities, but it never caught on. I suspect that measuring seconds will likewise have adoption problems. (The second is just too fine a measure for most human activities. When you need 10^5 to measure a earth revolution, your granularity is too fine.) {Of course, if the Good Lord had intended us to use the metric system, He would have had ten apostles.}

At first we thought that the solar day would be irrelevant in about five billion years, when the sun runs out of fuel. Now it looks like we won't need to worry about that, as we are colliding with Andromeda Galaxy in two to three billion years. We can try your way then. (How many places of precision when we are measuring 10^17 seconds from now?)

[wyrdling.livejournal.com]

i was extremely angry about it this year. i was running ragged as was and that just knocked me cold for weeks