The answer: Saint Lucia, with an astonishing 12 Nobel Laureates per 1,000,000 inhabitants, six times as many as the U.K. and twelve times as many as the U.S.
Which means, um, St. Lucia has two Nobel Prize winners, Sir Arthur Lewis, who shared the 1975 Prize in Economics, and Derek Walcott, who won the 1992 Prize in Literature. St. Lucia has 150,000 inhabitants.
Esoteric note: Lewis and Walcott were both born in St. Lucia, but spent most of their lives in the United Kingdom and Trinidad respectively. If they're counted wit the U.K. and Trinidad instead, then Iceland has the most Nobel Laureates per capita (3.5 per million), with one: Haldor Laxness (Literature, 1955), and a population of 300,000.
Congratulations to Anonymous and Belle Reve for winning the contest by providing the two possible right answers! Click to view a scan of your prize! If you E-mail me (email@example.com), I'll send you the actual prize in the mail.
So what did you imagine when I asked the question? Lots of doctors and scientists running around in lab coats? And what did you imagine for the nation? A rich, densely populated western nation with a large research budget? That's what I imagined when I heard the question at work. I was debating between the U.K. and Sweden, and I eventually answered Sweden.
Certainly I didn't expect the winner to be a tiny, relatively poor Caribbean island nation whose only major industry is tourism and whose main export is bananas.
This is a perfect example of how statistics that involve small numbers can be misleading when presented out of context. St. Lucia's Nobel Prizes are an interesting and benign example, but much scarier examples exist. For example, this crazy E-mail forward from a fictional Australian police officer claims that gun homicides in Victoria tripled after the government issued a gun buyback program in 1997. Sure they did: 7 gun homicides in 1997, 19 in 1998. Victoria had a population of 4.5 million in 1997. Woo.
And this is not even considering the problem of correlation and causation: these statistics don't prove that the gun buyback program had any effect whatsoever on gun crimes in Australia. Maybe in summer 1998, a crazed kangaroo stole a constable's gun and shot 12 people at a Kangaroos game.
So, with all this misinformation floating around, how to we educate the public on how to evaluate the statistics that other people present them?
Sounds like a job for public sociology.