skip to Main Content

Analysis of spatial inequalities at different geographical scales

This report provides empirical evidence on the relationship between local area income deprivation and individual socio-economic outcomes using a multi-scale approach. The main contribution of this work is its ability to define and measure neighbourhoods in a more precise or meaningful way to address issues of multiple scales and boundaries by using bespoke neighbourhood measures. In particular, in order to have a better understanding of the consequences of contextual local area income deprivation on individual’s outcomes, the suitability of different geographical units are considered both in terms of scale (i.e. from aggregate to very disaggregate) and type of boundary (i.e. administrative fixed boundaries vs more flexible boundaries), since different geographical scales and boundaries may lead to different results, with consequences on the design of public policies and their expected outcomes.

The analyses carried out in the report use geocoded longitudinal microdata for Sweden, the Netherlands and the UK, as well as longitudinal microdata from the EU-SILC for the RELOCAL partner countries with geographical identifiers for NUTS2 regions: Spain, France and Finland. Given the nature of the data available, different types of empirical analyses were developed with varying levels of methodological sophistication and spatial resolution.Regarding the former, the results from the analyses using longitudinal microdata for Sweden, the Netherlands and, to less extent, the UK on how contextual poverty, measured in terms of income deprivation, is related to individual labour income have shown that this relationship differs depending on the geographical scale at which contextual poverty is measured. The effect of contextual income deprivation appears to be most pronounced for lower spatial scales. Scaling up to larger geographical areas, the concentrations at micro scale are averaged out, resulting in less extremes of poverty concentration at these scales. As for the latter, the EU-SILC analyses are limited geographically to aggregate NUTS2 regions which hide substantial within-region variation in socio-economic conditions. Despite having strong limitations, it was possible to use the EU-SILC microdata to investigate patterns of income mobility, income inequality and inequality of opportunity across NUTS2 regions and by degree of urbanization of individual’s residential location for Spain, France and Finland. Overall, the results indicate that regional differences matter. In particular, the analyses provide some indication of a negative correlation between NUTS2 population size and the degree of upward mobility, in line with the results by degree of urbanization suggesting less income mobility for large urban areas. In addition, there is considerable variation in income inequality at the regional level, particularly if measured using different income share ratios as opposed to the more general measure of income inequality based on the Gini Index. This means that apparently similar levels of overall income inequality may hide variation in the more local profile of inequality between income shares in the top or bottom sides of the income distribution.

The work carried out in this report also has some limitations. Firstly, it measures residential poverty and deprivation in terms of low-income concentration, but income is only one dimension of poverty and while it would have been preferable to adopt a more multiple dimension definition this was not possible for data reasons. Furthermore, although the multi-scale approach shows that inequality is a multi-scale problem, on its own it cannot explain which mechanisms operate at different levels; achieving this would require combining them with detailed case study analysis.

One of the main conclusions from this report is that in order to have a better understanding of residential context, in particular area income deprivation, on individual socio-economic outcomes, it is important to measure and test the relationship at different geographical scales. However, the approach implemented in this report for Sweden, the Netherlands, and the UK can only be applied when geocoded data are available for very small spatial units and such data are still unavailable in many countries. Consequently, one very important conclusion and recommendation from the work carried out in this report is the need to improve the availability and access to socio-economic geocoded data at very low scale for more countries. Without this type of information it is not possible to provide guidance to policy makers on the more appropriate scales for public intervention.

 

This report provides empirical evidence on the relationship between local area income deprivation and individual socio-economic outcomes using a multi-scale approach. It uses interrelated data on individuals, place and time to investigate the influence of contextual local area income deprivation on individual labour income after controlling for individualscharacteristics and, where possible, family background.

Download the full report

Back To Top