By building a predictive model of solvability for burglary using data from a U.K. police force, this paper demonstrates that it is possible to predict whether burglary offences will be solved from analysis of the evidence recovered during initial investigation. The model is then externally validated by being run on a test data set which is separate from the one used to build the model, and has been shown to be capable of reducing demand (for policing) by up to 42.2%, with a potential resource saving of up to £700,000.
In addition to building a predictive model for clearance of burglary, this research has evidenced that there is important information that is not being recorded electronically, and which is not therefore available for research, or for use in algorithmic solutions to policing issues. It is therefore recommended that policing agencies examine the methods by which they are recording data to ensure that it is recorded in a retrievable and useful manner, rather than in inaccessible free-text reports or paper records.
Bio:
Professor Anindya Banerjee is Professor of Econometrics in the Department of Economics in the University of Birmingham which he joined in January 2008. Before coming to Birmingham he was Professor at the European University Institute in Florence and Fellow of Wadham College, Oxford. Professor Banerjee received his Ph.D. from the University of Oxford. His interests lie in time series econometrics, including factor models, and the econometrics of integrated panel data. He has recently been using his expertise in econometric modelling to look at the use of algorithmic methods in augmenting police decision making.
Last modified:
This page is maintained by the seminar list administrator.