At a conference I went to last week, Ian Foster talked about advancements in computation and how today’s computing power combined with massive datasets are changing the way science is done in some disciplines. He gave a very interesting comparison — a global climate change simulation model versus a global economic simulation model. The climate model divided the world up into a grid of thousands of squares and carefully modelled many characterisitcs of each square and its interaction with the climate. The economic model divided the world up into 10 regions …

An audience member (the governor of the Reserve Bank of New Zealand, in fact) pointed out that the difficulty with doing such detailed modelling in economics is a lack of data. Send up a few satellites and you get a pretty good picture of the weather all over the globle, in real time. In economics, data at the sub-national level are generally unreliable and hard to come by, and we’re still trying to forecast the past six months, let alone know what’s going on in real time.

More on related topics here and here.