Predictive Models for Welfare Reform
Posted on: 23 May 2014 by: Carl Haggerty - No Comments
Guest Post by Lucy Knight and Sue Rook
Over the summer, when most students would be taking a break from exams and deadlines, two young people will be deep in statistical regression modelling at County Hall. Sarah Blake and George Young are working on a project with Devon County Council and Exeter City Council to see whether the skills they are studying at university can help us plan our response to wide ranging welfare and benefit reforms set in motion by the Government.
The theory is fairly simple – linear regression takes measurements from something we can see, for instance demand on support services like Citizens Advice Bureaux or food banks, and works out whether it fits a formula. Even a complex event like the behaviour of a large population can be modelled if you have an idea what the different factors might be, and the advanced mathematical tools Sarah and George can command are able to test each factor in turn to work out what makes the most difference in the model and what creates the best ‘fit’.
Once we are confident about our model, we can use it to view possible future scenarios and decide how we will plan and allocate ever-reducing resources to support the people who will be most heavily affected by changes in the benefit system.