Does hosting Olympic Games matter after all?

Is seems to be common knowledge that attracting big sports events is good for business and especially tourism. I have never found this argument particularly compelling, after all it is mostly local residents who attend such events. And previous research I reported on gives me right: in the case of the Atlanta Olympics, the impact was very short-lived and limited to the tourism industry. But maybe there is more and better evidence.

Markus Brückner and Evi Pappa take a different approach form the traditional impact study: they look at macroeconomic aggregates and focus on the anticipatory effect during the bidding process for the Olympic Games. The fact that a country is bidding gives people an indication that aggregate demand may increase in the future, especially if the country is selected into the last set of candidates. This anticipation can increase economic activity right now. Brückner and Pappa study a panel of 184 countries over 57 years. They find higher GDP growth during the five years before hosting, peaking at four years when the next host is announced. As expected, the impact fades quickly for unsuccessful bidders. And results are robust for World Exhibitions, but strangely reversed for Football World Cups. In all that, I wonder whether bidding for such large events is in fact exogenous. Indeed, you only want to bid if you have a healthy economy, especially if the event is large like the Olympic Games.