Victor Matheson, Thomas Peeters and Stefan Szymanski author a recent paper that attempts to measure, using monthly data of South African tourist arrivals, the impact on tourism from the 2010 FIFA World Cup. Matheson and Szymanski are two guru’s in the field of sports economics and are known for their pessimism of ex ante studies that attempt to estimate the potential economic impact of mega-events. And we had a lot of those in South Africa: already in 2003, a Grant Thornton study had optimistically predicted that more than half a million visitors from outside Africa would visit South Africa during the event, staying an average of 15 days.
The new paper confirms the authors’ pessimism about ex ante studies: it shows that only 193,323 additional, non-SADC tourists visited South Africa during the event, which is far below the government’s estimate of 309,554. Given that the event cost the South African government an estimated $3.9 billion (I’m not sure where the authors get this figure), they calculate that the average “World Cup” tourist was priced at a steep $19,500.
The authors use a standard methodology (similar to a paper that Karly Spronk and I used a while ago) and include several economic variables as controls. I liked the grouping of countries based on the size of their tourist arrivals. But they make a number of assumptions which I think does bias their results towards the more pessimistic perspective.
They exclude, for example, SADC visitors, based on the fact that “those travelling from SADC nations are generally not the high spending tourists that South Africa was seeking to attract during the World Cup”. This is not a great argument. Would we exclude the visitors from Latin America for the 2014 World Cup in Brazil simply because they do not spend as much as their European counterparts? I suspect not. With African countries growing at massive rates, tourist visits to South Africa are becoming important luxury consumption items, even if it is only for visiting friends and relatives, or for doing some shopping. Maria Santana-Gallego and I have made this same argument elsewhere. In fact, when including SADC arrivals, the estimated number of tourist arrivals during June/July increases to 271,000, a boost of nearly 80,000.
Because 2011 economic data was not yet available, the authors also restricted their analysis to only include 2010 and earlier years. Where they do include data for 2011, the results are significantly different. For example, I’ve reported above results from model 1. If model 2 is used, the 271,000 visitors rise to 391,000 visitors, an increase of 44%. Moreover, the authors consider both tourist arrivals during the duration of the event (June/July) and during the entire year. They find, for example, that while tourism increased by 271,000 visitors during the World Cup tournament, 627,000 total additional visitors arrived during the year (1,3 million according to their model 1). In the conclusion, the authors note that “we estimate the total impact of the World Cup on visitors to South Africa in 2010 to be as high as 490,135″, slightly more optimistic than the 8% average for mega-events Maria and I calculate in our Tourism Management paper.
I suspect the authors were surprised by these numbers, as identifying displacement – where tourists that would have visited the country in the absence of the event – is the Higgs Boson of the mega-event literature. (Spronk, Sieberts and I also attempted to find evidence of this with South African data.) But few tourists were displaced by the 2010 FIFA World Cup and there are several reasons for this: the event was held during our winter (off-peak) season, which meant that there were few constraints on our transport and accommodation infrastructure. Also, several countries that were not our traditional tourism markets participated in the event, drawing tourists that would not typically have considered South Africa as a tourist destination (the authors mention the rise of Latin American tourists, for example, and I think this is an important point when considering whether to host a mega-event). The luck of the qualification round, and the actual match outcomes, also play a role: had Russia qualified instead of tiny Slovenia (a difference of one away goal!), we might have seen a significant number of additional tourists. A similar case could be made for Egypt instead of Algeria, or China instead of North Korea. Also, had England and Germany progressed to the final, imagine the boost to tourist arrivals over the last few weeks of the tournament.
The paper makes a valuable, first contribution to measuring the ex post impact of hosting the 2010 FIFA World Cup on tourist arrivals to South Africa. Surely there will be many more. But rather than just counting tourists, we would do well to also focus on the underlying causes (policy or luck?) of these tourist arrival patterns.