As part of her long paper for her honours degree, economics student Amy Power Jansen spent three weeks in Niger in Western Africa investigating the causes of the famine that plagued the country from 2005 to 2006.

In 2004, locust swarms and low rainfall meant that food production - primarily the cereal crop millet - dropped. There was a food crisis in 2005, but while when food production had dropped by even more in 2001, there was no crisis in 2002.

Jansen's research investigated why.

She found that although the national drop in production was sharper in 2001, some regions were affected more than others at a sub-national level. Areas near the Sahara, for instance, already had a higher variability in rainfall. This was also the main area affected by locusts, so the drop was more severe.

Additionally, farmers in the area tend to sell their harvest in October and re-buy what they need the following June to August. So with less to sell in October and higher prices in June, they had far less access to food than usual.