Hello, I'm Tim Harford and this is the shorter World Service Edition of More or Less, first

broadcast on the 5th of August 2016. This episode we cover Olympic medal tables and

maternal mortality rates.

Hello and welcome to More or Less on the BBC World Service, we're the programme which

looks at the numbers in the news and in life and I'm Tim Harford.

The Olympic Games are with us again and so we turn to statistics to see if we can

predict how many medals each nation will win.

The most obvious fact is that countries that are rich or populists tend to do rather

well, with the US and China at the top, but a statistical model using only GDP and population

is fairly crude. For example, it would put India at third in the

medal table and with all due respect, India's medal tally at the Olympic Games is usually

very modest indeed. So how can we build a cleverer statistical

forecast of the medals table and what would it tell us? Here to explain all is Jordan Dunbar.

We spoke to Dr Julia Bredman, who's an economist at the RWI Leibniz Institute for Economic

Research in Essen, Germany. With her colleagues, Carson Crete and Sebastian Norton, she's been

working on exactly that problem.

The most important factor that we have in the model is the number of medals won in the

last games. What can be shown is that there's a large persistence in Olympic success. There

are some countries that are always winning a lot of medals and some countries that will

never win a medal and that can not only be explained by observable factors. You can think

of it such as a sports culture or how important is Olympic success by considering the medals

won in the last games we can control for this consistency.

...