Abstract

The democratic processes by which we elect politicians are problematic because we often have very incomplete information about the effectiveness of various political alternatives. This talk will consider how the political process could be redesigned to give voters much more accurate information, thus allowing them to make better choices. The talk will discuss the use of very simple mathematical modelling and statistical methods to test the proposed new political system, and to contrast it with the existing system used to select politicians in the UK, and elsewhere. The proposed system would almost certainly require changes and elaboration before any implementation. However, in modified form, it might prove very effective, in the long term, as a way of enhancing the quality of public policy.