On a slightly more hackerish note: I worked on a power trading system, and it would break like hell twice a year, on DST switch days. Because electricity trades on an hourly basis (except in Texas, which uses 15-minute intervals), everyone who deals with aggregating power flow or power over the course of a day, ends up writing a loop like "for (hour = 0; hour < 24; hour++) { ... }", which is totally wrong. The spring forward day has 23 hours, and the fall back day has 25. That code therefore breaks twice a year and loses money because (a) it needs to be fixed, and (b) it did its accounting wrong. I hope studies examining the economic efficiency of DST take this sort of thing into account. Probably not.