In 1964, Japanese officials expected 130,000 foreigners to visit during the Olympics, so they encouraged proprieters to get ready for the world to flock across the seas, not only to Tokyo, but to the beautiful vistas around Tokyo and beyond.

But alas, government projections proved to be overly optimistic as only 70,000 tourists were estimated to arrive that October. Millions of dollars were spent to accommodate more people and make the experience for non-Japanese tourists a good one in popular resorts like Atami and Hakone, but facilities never got close to capacity.

Kyoto hoteliers turned down reservations by Japanese wanting to see the old capitol in all its beauty during the refreshing Autumn season in anticipation of the busloads and trainloads of foreigners instead saw occupancy rates plummet, when instead they should have been near capacity.

When a country holds the Olympics, there is a promise that the tourists will come and the money will flow for hotels and restaurants. It is a promise that goes unchallenged, and proven time and time again to be baseless.

Economist Andrew Zimbalist explains in his fascinating book, Circus Maximus – The Economic Gamble Behind Hosting the Olympics and the World Cup,  that a country which hosts the Olympics may very well be welcoming a large number of foreign visitors (athletes, coaches, judges, media, family members, etc.), anywhere from 10 to 25,000 related to the Olympics. But rarely have there been instances where the number of tourists actually increase. Very often, it is the opposite – tourism decreases, sometimes significantly.

Zimbalist cited China, which dropped 6.8% in foreign visitors from 26.1 million in 2007 to 24.3 million in 2008 when the Olympics were hosted in Beijing. London experienced a year on year drop of 6.1% in overseas visitors from July and August 2011 to July August 2012, when the Summer Games were held there. Athens expected 105,000 foreign tourists per night during their Olympic Games in 2004, but in actually hosted only 14,000 per night.

Zimbalist does write that Sydney’s number did increase modestly from 2.5 million in 1999 to 2.7 million in 2000 when the Games were held in Australia, but that was more than a quarter less than projected, resulting in low occupancy rates on top of expansion hotel expansion in anticipation of higher numbers – just like in Tokyo in 1964.

So what’s happening?

A dynamic that is not understood is that Olympic traffic is not adding to overall numbers – it is merely replacing traffic that would have come. And yes, there are many tourists who avoid countries or cities hosting Olympics as they are not interested in sports, crowds and higher prices. Even worse, as has been proven, the overall numbers end up being worse than a normal, non-Olympic year.

Olympics and other big-tent events like the World Cup, or major international conventions for that matter are spike events. How does a city handle spikes in numbers? In 1964, luxury liners provided that short-term capacity. In 2016, Airbnb will help with the overflow in Rio. In 2020, where will the extra capacity come from. Got an extra bedroom to rent?