WOULD a feudal king swap places with a present-day insurance clerk? The king has many palaces and armies of servants. But he is also a martyr to toothache. News reaches him slowly. His food is often stale. And he is bored of his courtiers. He might willingly trade his kingdom for life as an office drone in the 21st-century, with its dentistry, refrigeration, Google and YouTube.

The poser underlines the great material advances made in recent centuries. It also shows how tricky it is to compare living standards over time and across societies. Comparisons over recent decades are routinely made using GDP, but they are troublesome. Why is GDP a poor measure of progress?

Gross domestic product is a measure of output, income and spending all at the same time. In post-war Europe and America, the growth in living standards and in GDP were synonymous. GDP growth became a target for politicians and a scorecard by which they were judged by voters. Even so, it has always had critics. Environmentalists have long lamented that GDP treats the plunder of the planet as something that adds to income, rather than being treated as an expense. A repeated charge is that GDP is divorced from notions of spiritual well-being. Robert Kennedy once famously took aim at GDP which, he said, counted cigarette advertising and jails but did not include “the beauty of our poetry or the strength of our marriages”. Still, GDP growth was a decent, if rough, guide to material progress. The more output and income was generated (after adjusting for inflation), the better off we were.

That equation worked pretty well when the economy was still mostly farms and factories, producing things of similar quality that could easily be counted.  But GDP is less suited to the task of measuring modern, service-led economies that are geared towards the quality of consumer experience, rather than consumption of greater quantities. It is far harder to identify the extent to which changes in price reflect a better service. When medical charges rise, it will generally count as inflation, even if the quality of health care is improving faster than prices are rising. And where consumers pay nothing, as is often now the case with digital services, they do not register in GDP. The consumer benefits from Google and Facebook are thus excluded. Previously paid-for things, such as maps, encyclopedias and music recordings are now free, so have dropped out of GDP. Online shopping, banking and travel-arranging is more convenient for consumers. To the extent that all this saves on buildings, it detracts from GDP. For the most part, the trickiness of measuring the output of services leads real GDP to be understated. But mis-measurement works the other way, too. For instance, if an airline squeezes more (and thus cheaper) seats on to a plane, it counts as extra output, even thought the quality of service falls. Perversely, the more risks bank take, the more they contribute to GDP, even as the quality of lending falls.

GDP is a creature of the industrial era of mass-produced, homogenous goods. It is far less useful guide to affluent economies where the quality of services is prized over simply having more stuff. It is badly attuned to digital economies, where activities that were once paid for, such as contacting friends or finding information, no longer attract a charge. GDP is thus increasingly failing to fully capture gains in average living standards. It is tricky to compare the life of a medieval king to that of a modern-day worker. But it is almost as difficult to put a number on how much better is a consumer basket that includes smartphones and music streaming to one filled with fax machines and audio-cassettes.