Researchers make Olympic predictions

Researchers, one of them an undergraduate, have used economic analysis to analyse and predict participation and medal outcomes for the upcoming Winter Olympics in Salt Lake City.

Daniel K.N. Johnson (visiting fellow at the Center for International Development) teamed up with Ayfer Ali ’02 to explain Olympic performance using economic and political variables.

This is their second Olympic collaboration. In 2000, they used the same method for the Sydney Summer Games, and found their predictions remarkably accurate. Their predictions had a correlation of 0.96 with actual participation, 0.95 with medal counts, and 0.96 for gold medals alone.

Johnson and Ali’s current paper compares the effects of income per capita and population on Summer versus Winter Games, finding that population matters more for summer and income matters more for winter results. Colder climates lead to more participation and more medals in both seasons. Nations with single-party or communist-like political structures win a surprisingly large number of medals in both seasons, considering the number of athletes they send.

No large or significant differences were found between events, although income is slightly more important for winning medals in equipment-intensive sports (e.g., luge or sailing) and population is slightly more important for stamina-intensive sports (e.g., nordic skiing or marathon) and team events.

For major participating nations, sending an extra athlete requires a $260 rise in income per capita. Similarly the “cost” of winning an extra medal is $1,700 per capita and $4,750 per capita for a gold medal.

Surprisingly, the researchers predict that although the United States will have the largest athletic contingent, Germany will win the most medals (31, 11 of them gold) and second place will go to Russia (21, 10 gold) as they edge out the United States (20, seven gold) and Norway (20, six gold).