ONLY an economist would think to ask whether Christmas is efficient. In 1993 Joel Waldfogel, then a professor at Yale University, turned a lunchtime conversation with colleagues into a paper entitled “The deadweight loss of Christmas”, which argued that, no, it is not. That gift-giving might actually be bad is the kind of opinion which breeds a deep mistrust of economists—loathing is perhaps too strong—among those not schooled in the dismal science. It is also just the sort of analytical insight on which economists pride themselves: counterintuitive, irreverent and interesting. But they should perhaps be less pleased with themselves. The way they think about the most festive time of the year reveals something important about the shortcomings of the field’s approach to human behaviour.

Mr Waldfogel’s notion was a clever one. Massive amounts of money are spent on holiday presents; it makes sense to ask whether such spending leaves the world better off. In buying gifts, people do their best to find something the recipient will appreciate. But, economists assume, people know their own preferences better than others do. The best a gift-giver can hope to do, in terms of making another person better off, is to give the person what they would themselves choose to buy with the money to be spent. Because the giver inevitably understands the receiver’s preferences imperfectly, recipients usually value gifts by less than their purchase price, generating a substantial “deadweight loss” to the economy. Ho ho ho.

Somewhere between a tenth and a third of the value of the gifts given at Christmas is destroyed, Mr Waldfogel estimated, based on experiments conducted with his students at Yale. Because holiday spending accounts for a meaningful chunk of GDP, the loss from Christmas is about a tenth of that created by income taxation, he concluded. Not all gifts are equally poorly chosen. Close friends and family are likely to understand their loved ones better and therefore to give more efficiently. In general, however, people would be better off if they simply gave each other cash. This was an unusual and interesting way to think about Christmas (and indeed, about holidays in general). Practical, too: those considering giving gifts at holiday time should proceed if they know the tastes of their receiver well, and if they are prepared to think hard about what to get. Otherwise, it’s best to go with cash (or perhaps gift cards, which are less efficient than cash but which may be more acceptably gift-like).

It is not mere sentimentality to find something amiss in this analysis, however, as some economists have recognised. Gift-giving is not a meaningless transaction. The act of giving itself creates value. In response to Mr Waldfogel, John List, of the University of Chicago, and Jason Shogren, of the University of Wyoming, conducted their own experiment, using auctions rather than surveys to tease out the value people placed on gifts they had been given. They concluded that those receiving the gifts on average valued them by 21-35% more than the cost to the giver. Their experimental design, they noted, was based on the novel principle that “material value + sentimental value = total value”.

Similarly, a panel of economists convened by the University of Chicago and regularly polled on economic questions disagreed, when asked about the subject in 2013, that giving presents is inefficient. Gift-giving generates value by signalling to the recipient that the giver cares about the relationship, some noted; the signal is especially strong if the gift demonstrates the giver’s familiarity with the receiver’s tastes and preferences. Others reckoned that the pleasure the giver takes in giving ought also to be taken into consideration. “This is the sort of narrow view that rightly gives economics a bad name,” said Angus Deaton, a Nobel-prizewinning economist, of the efficiency question.

But it is the way that the question is posed, rather than the way economists choose to answer it, that is the real problem. Gift-giving does not occur in a vacuum; people do not randomly set out to raise the welfare of their loved ones with festively wrapped gifts. Rather, it occurs within a very specific social context: the holiday season. Why do people do Christmassy things at Christmas? Why do they place tinsel-strewn trees in their homes and let their children sit on the laps of men dressed as Santa? They do so because they are participating in a long-practised mass social ritual. Assessing gift-giving without taking account of this social context is a near-useless exercise.

How might that context be taken into account? Christmas is what in other circumstances an economist might refer to as an institution. Institutions are rules and norms that are developed to solve social problems; concepts of private property, for instance, help a society to manage a tendency to overuse common property. Some research assesses holidays in this way. A paper published in 2001 by Vijayendra Rao, of the World Bank, reckons that festivals in India provide an opportunity to cement family bonds, deepening social capital in ways that yield returns at other times of the year. Indians who spend more during festivals enjoy higher social status, which translates into tangible benefits, like getting better deals from shopkeepers on food purchases.

You’d better watch out

Even this is too bloodless an analysis. Many holidays are rooted in ancient religious or cultural practices. Whether and how you celebrate are matters of personal and group identity. As many clerics are keen to point out, Christmas is about more than presents. The exchange of gifts is an inseparable part of a communal time of celebration and goodwill. Economists would be more useful if they could recognise when and why maximising efficiency takes a back seat. They would also be more fun to have around at Christmas.