Published 14 July 2016
Spatial microsimulation (SMS) is a range of techniques for estimating the local distribution of a variable – here, household income – by combining social survey microdata with Census or administrative population totals. This paper makes a case for the value of these methods in social policy analysis of spatial economic differences because unlike other methods and sources, they permit distributional analysis of income, encompass both market outcomes and secondary distribution through taxes and transfers, and measure income poverty in standard national terms. As a demonstration of spatial microsimulation by iterative proportional fitting (IPF), the household income distribution in London's 33 boroughs in 2001/02 and 2011/12 is estimated in this paper. The coherence and plausibility of the results in comparison to other official statistics is examined in some detail. Two refinements to standard IPF methods are presented, including "multi-level IPF", which allows the use of both person- and household-level data; this is found to improve the estimation of poverty rates. The paper confirms the value of SMS for synchronic spatial analysis, and argues for its hitherto little-explored use in modelling spatial differences in the effects of fiscal and welfare policy changes.