Local level database for socio-economic indicators

First, a literature review was conducted to assess the existing official data and socio-economic indicators, demonstrating the importance of the disaggregated analysis to explain territorial inequality trends and inform place-based policy initiatives. Second, a literature review on disaggregation methodologies of spatial data, notably, areal techniques and micro data-based techniques, and their weaknesses was conducted to justify the use of the Generalised Cross Entropy (GCE) method. Third, a literature review of the territorial inequalities in the EU to show that although there are many studies at the regional level an analysis at the local level was lacking.

Using the indicator of ‘household income’ is the common procedure when analysing income inequality. Income carries a large weight in the level of wellbeing and life satisfaction both at individual and also at the territorial level (either by locality or region), and is therefore considered a good indicator of economic welfare and development. However, when it comes to the spatial level, due to confidentiality reasons, in Europe in the best of the cases only information at the NUTS 2 region level of residence is provided. This implies that any study of European income inequality is spatially restricted to large, heterogeneous and, in many cases, geographically extensive regions. The broad NUTS 2 administrative regions do not capture the rural-urban divergence or core-periphery dynamics that might be present within the regions, nor do they allow the analysis of local development using this indicator.

The poverty indicator used is AROPE, a multidimensional single indicator, which measures the rate of people at risk of poverty or social exclusion.  

The AROPE indicator considers that an individual is at risk of poverty or social exclusion if he/she meets at least one of the following three criteria:  

Some additional local variables, related to the level of education, dependency or immigration of the localities, were collected for the EU countries where that information was available at such a level of spatial disaggregation. This completes the picture so that EU territorial inequalities in terms of income, poverty, education, immigration or ageing are given a new spatial perspective. Not only can spatial disparities between EU regions be analysed, but also, the territorial inequalities within the regions.

This methodology is based on the framework of Maximum Entropy. In this framework, the variable of interest has a probability distribution with an unknown probability for each value. The basic idea of this type of methodology is to obtain the estimation with the highest degree of uncertainty that at the same time is able to fulfil the conditions from observable data. The Generalised Maximum Entropy (GME) estimator has been applied to the estimation of linear regression equations. The GME estimator will choose the distribution of probabilities that deviate least from a situation of maximum uncertainty as the optimal solution.
By applying this novel methodology, it is possible to obtain consistent estimates of average household income at the local level and by natural intervals.