The Olympics: Where communism wins

The Olympic motto, "Citius, Altius, Fortius" — faster, higher, stronger — might inspire athletes training for the Games, but for a nation looking for Olympic glory, a more useful dictum might be "Maior, Ditiores, Communistarum": bigger, richer, communist.

While upsets are always possible in any individual event, the factors that make a nation an Olympic powerhouse are pretty clear, and it's surprisingly easy to predict which countries will come out on top.

Shortly before the 2000 Sydney Olympics, two economic papers appeared within days of each other looking at the determinants of gold-medal success. Remarkably, both came to virtually the same conclusion about what makes a nation an Olympic champion. Ever since, the lead authors of each paper, Andrew Bernard of Dartmouth's Tuck School of Business and Daniel Johnson of Colorado College, have been using these factors to make predictions before each Olympics, sometimes with uncanny accuracy.

"When we compared the final medal count with the expected result in 2000, we thought we had made a horrendous error: It was like a .96 correlation," showing very high accuracy, Johnson said. "You don't get that sort of result in an economic model."

But Johnson soon realized this shouldn't be that surprising, as overall Olympic success is remarkably predictable. The most important factors by far are population size and per capita gross domestic product. The prediction models don't look at individual athletes or sports; they consider national team performance as a whole.

"Olympic athletes are like complex machines," said Emily Williams, a recent Tuck graduate who has taken over the work on Bernard's model in predicting the London Games. "The more people there are, the more machines there are. The more resources per person each country has, the more these machines can be invested in and turned into Olympic athletes."

GDP, in particular, is a large part of why the United States holds a nearly insurmountable lead in the all-time count, with 2,296 medals, whereas Russia and the Soviet Union come in a distant second, at 1,327. (Russia's historical performance is perhaps more impressive given its much lower GDP and population figures, and it holds a clue as to how a nation can punch above its weight.)

This may seem intuitive enough, but there are a couple of other ways a country can gain a leg up. One is to actually host the Games. This might be partly due to home-field advantage — less travel, familiarity with the facilities — and host countries' tendency to field more athletes, but it's also because countries tend to invest more in sports generally when they host the Games. Whether or not countries actually make money on the Olympics, they may want to send a signal about their place in the global economy by fielding competitive athletes; Greece, for instance, took home 16 medals in 2004, when Athens hosted, but only four in 2008.

One other factor identified by both papers is a bit more surprising: communism. Throughout the Cold War, when medal counts became a matter of not just national but ideological pride, communist governments like the Soviet Union and East Germany were able to allocate government resources much more efficiently to build sporting powerhouses. They consistently outperformed predictions based on size and GDP alone.

This wasn't true just for the Eastern Bloc. Cuba, for instance, has won more than twice as many Summer Olympics medals as Brazil, despite having only a fraction of its wealth and population.

Given these factors — size, wealth, host status, and communism — China's impressive performance at the 2008 Beijing Games — 100 medals, 51 gold — was hardly surprising. But China actually outperformed the models: Johnson's model predicted 79 medals for China in 2008, while Bernard's gave it 81. According to Johnson, China has been "historically the most difficult country to predict."

The only other country in China's heavyweight population division, India, has historically been one of the Olympics' great underperformers, with only 20 total summer medals. It's tied with Slovakia, which has 0.4 percent of India's population and wasn't even a country until 1993.

The world's largest democracy has never really made sports spending a priority. Plus, several of the sports that Indians excel at internationally, like cricket and squash, aren't Olympic events. Johnson's model predicts seven medals for India in the 2012, though it won only three in Beijing.

The country that most consistently outperforms expectations is Australia, which has the most medals per capita over the last three Games. Yes, Australia is traditionally a sporting culture, and the fact that 85 percent of the country lives within 30 miles of the ocean may contribute to its historic dominance in swimming.

But it also shouldn't be surprising to learn that after a disappointing zero-gold-medal performance in 1976, the Australian government embarked on a massive centralized training program to spur Olympic glory — modeled on the youth sports academies of the Eastern Bloc. Ahead of the 2000 Sydney Games, the country spent $20 million on scientific research aimed at improving its athletes' performance. Australia might be a democracy, but since the 1980s, it's taken an almost Soviet approach to preparing for the Olympics. And it's been among the top 10 nations in medal count since 1992, including an impressive 58-medal performance in Sydney.

So what do the numbers predict for London? Colorado's Johnson puts the United States in first place, with 99 medals. Russia, which underperformed its prediction by 11 medals last time, is still second, with 82. China is to come in third, with 67, though with more golds than Russia. And home-field advantage is expected to boost Britain into fourth place, with 45.

This year, Johnson is de-emphasizing the communist metric, because the sample size is so small in the post-Cold War world. So it may be interesting to watch what that means for countries like Cuba and Vietnam.

Using the formula developed by Dartmouth's Bernard — which has been more accurate because it includes by far the best indicator for medal count: who won last year — Williams has the same top four, but in a different order. She gives the United States 103 medals, followed by China with 94 and Russia with 67. Her projections are more optimistic for Britain, which is predicted to bag 62 medals, its highest count since the London Games of 1908, when it won nearly half the medals awarded.

There's also a new kid on the block in the prediction game. Using an expanded series of metrics that includes "political conditions, macroeconomic stability, macroeconomic conditions, human capital, technology, and the microeconomic environment," economists at Goldman Sachs have produced their own projected count, giving the United States 108 medals, China 98, Russia 74, and Britain 65.

Accurate as these models are, it's the deviations from the mean that make the Olympics fun. Watching Michael Phelps trounce the rest of the world in the pool might be fun for a while, but without the possibility of an upset, who would watch?

Williams predicts that the most interesting story in the next few Olympics may be not China, but smaller developing countries that are slowly increasing their share of the medal count.

"I'm glad the model's not always accurate," Johnson said. "Otherwise, we could just mail out the medals in advance."