Free exchange | Working hours

A plea for more data

One in four Americans works between the hours of 10pm and 6am. In France the figure is one in 14

By C.W. | LONDON

IN AMERICA, one in three workers does some work on the weekend. Europeans are more likely to treat Saturday and Sunday as sacred: only one in five workers in France, Germany and the Netherlands buck the trend. American workers are rather nocturnal, too. One in four works between 10pm and 6am. In France and the Netherlands, one in every fourteen does so.

Why? The obvious explanation is longer working hours. Average hours are higher in Anglo-Saxon countries than in other advanced economies (see chart). A goodly chunk of Americans and Brits works over 65 hours a week. If you are working longer, the argument goes, it is inevitable that working time will spill over into evenings and weekends.

A new paper, though, argues that long working hours have little to do with it. In a series of regressions, they show that only a small part of the high incidence of work during unsociable hours is due to Americans’ long workweeks.

So what explains Americans’ tendency to work at weird times? The authors offer the unhelpful suggestion that it’s to do with “differences in the way that work is structured in America.” Or possibly it’s because of “some unique characteristics of American and European culture.” In other words, no one knows.

This paper exemplifies the problem facing economists who research time-use. Beyond some highly aggregated OECD numbers, there is a pathetic amount of data available. In this paper, the economists were forced to use French data from 1999 and British data from 2001. The Economist faced similar data problems when researching a Free exchange column on working hours, published in April. Making comparisons between countries—which economists love doing—is sometimes impossible. According to OECD data, the only country that gives information about the hours of the self-employed is the Czech Republic.

Some countries have bucked the trend. American time-use data is a model for the rest of the world. The "American Time Use Survey", began in 2003. Data are released annually and offer some lovely little nuggets of information. (Did you know that each day in 2013 the average American spent 17% longer on the phone than they had done in 2003?). The Harmonised European Time Use Survey (HETUS) has some advantages over the American. For instance, it measures the time use of individual members of a household, not the household as a whole (as the American survey does). But HETUS is released once a decade.

Hard-nosed economists complain that measuring time is basically impossible: if I am reading while eating, is my primary activity eating or reading? But many important economic statistics suffer from measurement problems (see, for example, Buttonwood's excellent discussion of productivity measurement).

A minority of economists, notably Gary Becker, have for a long time been interested in time-use. But now the mainstream is gradually coming round to the idea that time-use is important, says Oriel Sullivan of Oxford University. In recent years some of the big-hitting economics journals have publishedpapers that look at time allocation. That belated recognition should be welcomed. Time-use allocation influences quality of life. It also exerts strong effects on health. More data would help us answer some of the big economic questions. Why did Keynes' prediction of a 15-hour working week not come true? Do college graduates work longer than high-school dropouts across the world, as they do in America?

Discover more

Religious competition was to blame for Europe’s witch hunts

Many children are still persecuted as alleged witches in Africa for similar reasons

Has BRICS lived up to expectations?

The bloc of big emerging economies is surprisingly good at keeping its promises


How to interpret a market plunge

Whether a sudden sharp decline in asset prices amounts to a meaningless blip or something more depends on mass psychology