A Review of Microsimulation for Policy Analysis

Abstract

There are a wide range of methodological frameworks and techniques for policy evaluation and socio-economic impact assessment. A useful distinction is to divide the literature on such models by the level of resolution adopted. It is then possible to identify macro, meso and micro approaches. Macro models, dealing with whole countries or nations, are most common in economics and social policy. Meso-scale models, where countries or nations are split into regional zones, have a longer tradition in regional science, planning and geography (McCann 2001; Stimson et al. 2006). For example, many former macro-scale models such as input–output techniques are now increasingly appearing in the literature at the regional scale. Quantitative geographers have tended to build meso-scale models for smaller geographical regions, such as small-area census-based zoning systems within cities (Wilson 1974; Foot 1981; Stillwell and Clarke 2004; Fotheringham and Rogerson 2009). These types of model have a long history of applied success but the complex dynamics which underlie social and economic change, as emphasised in Chap. 2, can produce very different results within different small-area localities and even within individual households (or firms). In particular, it is useful to be able to understand, estimate or predict which localities, households or individuals (given their demographic and socio-economic characteristics) are most likely to benefit from a change in socio-economic or environmental policies. Thus, it could be argued that policy relevant modelling is a challenging research area which is well suited to a modelling framework which emphasises household or individual-level processes at the local or micro level rather than aggregated processes at the macro/meso-level.