Others are more cautious. Non-sponsors assume that, if they advertise during the games, their message will get lost in the hubbub, says Ian Whittaker of Liberum Capital, a bank. He predicts that many will wait until September, and that the Olympic boost will be less than expected.

Governments of countries that host the Olympics usually boast that the games will generate vast economic returns for the nation. David Cameron, Britain's prime minister, promised to "turn these games into gold for Britain", to the tune of £13 billion over four years (see article). Several studies suggest that is unlikely. Victor Matheson of the College of the Holy Cross in Massachusetts finds that organisers of big sporting events tend to overestimate the benefits and underestimate the costs.

They are particularly bad on opportunity costs, counting every penny that sports-mad tourists spend while forgetting that others will cancel trips to avoid the crush. South Korea attracted lots of football fans during the 2002 World Cup, but because so many non-fans stayed away total arrivals were the same as the previous year. Some of those who pitch up for a sporting event have merely rescheduled a trip planned for another date, and should not count as extra arrivals. Mr Matheson concludes that, though the gross economic impact of big sporting events is large, these losses mean the net effect is negligible.

It may even be negative. Host governments spend vast sums on building stadiums and sprucing up nearby railways and roads. The Olympic authorities could pick host cities that already have the necessary infrastructure in place, and in such places the games might turn a profit. But the IOC likes host cities to erect grand edifices with the Olympic name on them.

The billions that Britain has spent on revamping bits of east London will generate benefits; but so would spending such sums on many other things. London's Olympic aquatic centre looks great, but it cost £269m—a great deal more than most public swimming pools. The roads built for the games may prove useful, but other projects might have done more good. They might have been cheaper, too; Olympic infrastructure tends to break budgets even more than infrastructure projects in general do. A recent working paper by Bent Flyvbjerg and Allison Stewart of the Saïd Business School at Oxford University found that every Olympiad since 1960 has gone over budget and that the average overrun, at 179%, was worse than for any other kind of mega-project.

When the consumer sits down to watch the games, though, he is not interested in efficient infrastructure investment. He wants to see something that lifts his spirits. This cannot be guaranteed—but unlike other purported benefits, it is at least not ruled out by the data. A study by Stefan Szymanski of the University of Michigan and Georgios Kavetsos of the Cass Business School in London finds no statistically significant upswing in national happiness attributable to hosting the Olympics (in this it comes a poor second to the football World Cup). But it doesn't seem to do any harm. And maybe this one will be particularly uplifting. After all, what could be more fun than watching athletes sweat while eating fish and chips?