The clocks go back tomorrow, which means an extra, delicious, guiltless hour in bed. Every year I dream of a world where the clocks go back every day. Of course, that would create trouble, notably that in a fortnight I’d be waking up after sunset. I would also be getting out of bed about the time all the good restaurants were closing. So I toe the line and put the clocks back only once a year, and - sigh - put them forward again in the spring.

My annual inner monologue suggests two reasons to get up in the morning and go to bed at night: first, to enjoy the sunshine; and second, because that is what everybody else does. But what if the two imperatives collide? What matters more, waking up at the same time as everyone else, or waking up with the sun? It might sound like a daft question, but not if you’re a Hong Kong-based journalist writing for a London-based newspaper, or a financial analyst in Silicon Valley who needs to be awake when the market opens on Wall Street at 6.30am Pacific Time.

Co-ordination may be important not just for the global elite but for those who deal with them. If enough Californian financiers are going to work at 5am, the gas stations and coffee shops will start to open early too: the baristas don’t need to co-ordinate with the baristas in Manhattan, but the financiers provide a link that is powerful enough to make sure that they do anyway.

Any resident of Queensland, Australia can assure you of this. Queensland doesn’t switch on to Daylight Saving Time when the other east coast states do, but some people simply get out of bed earlier in the winter to stay in sync. Or consider the states in northern Mexico, which align their time zones with the neighbouring US states rather than the rest of the country.

Three economists, Daniel Hamermesh, Caitlin Knowles Myers, and Mark Pocock, have devised a way to find out how important these co-ordination effects really are. One of their tools is based on a historical relic: the fact that television schedules in the US vary by time zone. This practice originated in the 1920s, when the Eastern and Central time zones received simultaneous live radio broadcasts, the Central time zone broadcast being an hour earlier on the clock. The sparsely populated Mountain and Pacific time zones had to listen to repeats instead. For no other reason than pure tradition, prime time evening shows screen at 10pm Eastern and Pacific, 9pm Central and Mountain.

But then take a state such as Arizona, which does not observe Daylight Saving Time. In summer it is in the Mountain time zone; in winter it is in the Pacific time zone but with the evening shows still at Mountain timings. That sort of quirky set up - there are others - allows Hamermesh and his colleagues to show the impact of the television schedules on people’s daily routines.

Rather depressingly, late-night talkshow host David Letterman outshines the sun in his effect on what people are doing. Push the television schedules an hour later and 5 per cent of people will be watching television later - nearly a third of those actually watching the television. But if sunset is an hour later (because the individual is at the western end of a time zone) only half of 1 per cent of people will watch later television.

That itself might not be surprising, but the effect spills over on to sleeping patterns: the television, more than the sunrise, determines when people get up in the morning. It even governs the behaviour of people who don’t watch much television, since they need to be co-ordinated with everyone else. It seems that co-ordination with other human beings is more important than official time, or even than sunlight itself.