Researchers

Simon Burgess, Ron Johnston, Carol Propper,  Kelvyn Jones, Anne Bolster, Rebecca Sarker
University of Bristol
1 October 2002 - 30 Sept 2005

Context

The project will investigate the impact of neighbourhood on individual’s life chances, attitudes and behaviour, in order to answer the question of whether locality matters. The findings will guide policy makers on whether to direct policies to individuals or areas.

Aims and Objectives

• Quantify impact of neighbourhood on individual;
• Estimate impact of local influences on outcomes;
• Utilise BHPS panel structure to exploit data on individual change of area and change in area characteristics;
• Address endogeneity of area in estimation of area effects;
• Overcome heterogeneity across individuals and sampling issues by using multilevel modelling;
• Issues addressed using a multidisciplinary approach.

Methodological aspects of importance

The techniques utilised in the project will advance methodological practice. The strength of neighbourhood effects operating at different scales will be measured. Additionally the issue of endogeneity will be addressed, by using matched house price data. More specifically, individuals who have very high fixed costs of moving, will be differentiated from others.

Research Design

The BHPS will be used, together with matched-in measures of area. For the purpose of the project area will be defined as ‘bespoke’ neighbourhood at a variety of scales, that is ever increasing (in terms of population and distance) concentric circles which are centred over postcodes of BHPS households. The problem of endogenous area characteristics will be addressed using institutional features of the UK housing market (i.e. social rented housing). Multilevel modelling, combined with econometric techniques will be used to model the data.

Outputs

Organise a conference to present findings and a workshop on data analysis and matching issues. Conferences planned: Geocomputation and EPOP